July 1, 2008

Google thinks you know what you mean, just not what you think you mean

OK, I'm going to lay off the "big bucket of bits is all you need" theory of science, computing and the future in a minute.

But not before this example of where simply using the relative frequency of words to perform spelling correction breaks down.

Posted by Chris at 4:47 PM | Comments (0) | TrackBack

June 29, 2008

A huge evergrowing pulsating brain that rules from the centre of the ultraworld*

Whenever I read something from the Cult of the Singularity, I find it hard to not conjure up the hectoring tones of Johnny from Mike Leigh's film Naked. You have to wonder how many spurious factoids David Thewlis had to commit to memory to get his improvised monologues* to work:

"And every barcode is divided into two sections by three markers and those markers are always represented by the number six. Six, six, six

"And what they’re planning to do, in order to eradicate all credit card fraud and in order to precipitate a totally cashless society…and they’ve already tested it on the American troops: they’re going to subcutaneously laser-tattoo that mark onto your right hand or onto your forehead. They’re going to replace plastic with flesh!

"Fact!"

I'll spare you the whole tirade but it leads up to the point where Johnny and the Singularists come together as one:

"And no, we’re not going to sprout extra limbs and wings and things because evolution itself is evolving. When it comes, the apocalypse itself will a part of the process of that leap of evolution. By the very definition of apocalypse, mankind must cease to exist, at least in a material form. We’ll have evolved into something that transcends matter, into a species of pure thought. Are you with me?"

And so, there I was reading Kevin Kelly's exposition of the OneMachine made out of old PCs yoked together that thinks with hyperlinks, mentally adding an extra "Fact!" at the end of every paragraph to complete the effect:

"Each new link wires up a subroutine, creates a loop, and unleashes a cascade of impulses. As waves of links surge around the world, they resemble the thought patterns of a very large brain."

Fact!

"By 2040, the planetary computer will attain as much processing power as all 7 billion human brains on Earth."

Fact!

And what do these computers actually do when harnessed as one? Some of them do something useful such as perform quantum mechanical calculations to predict protein folding. Unfortunately, they are more likely to be sending out tons of spam. But no mind, "we are headed toward a singular destiny: one vast computer composed of billions of chips and billions of brains, enveloping the planet in a single sphere of intelligence".

Fact! The techalypse is coming.

But there was one thing niggling at me: where were the figures coming from to support the contention that the One Machine rivals even one brain today? And this is assuming you accept Giulio Tononi's assertion that intelligence comes as a function of complexity, that you can just slam a bunch of circuits together and automatically get something that thinks. Towards the bottom of the page are some figures in a diagram.

By far the oddest one is the choice of 70MHz for the brain's operating frequency: "grey matter is about as speedy as an original Pentium". That sounds pretty quick to me given that the calcium induced cascade that triggers a neural response takes on the order of 200µs. That gives you a maximum frequency — even working on the basis that neurons switch like electronic transistors, which they don't — of tens of kilohertz. By that token, the human brain can barely keep up with a Sinclair ZX80. The actual frequency is probably way lower than that as neural signalling seems to rely on pulse trains that take tens of milliseconds to transmit from one neuron to another. The brain makes up for that sluggishness by not trying to work like an electronic computer. The transistor, as it turns out, is a pretty rotten analogue for a neuron, although maybe not nearly as bad as equating a hyperlink with a synapse.

But I'm really curious about the 70MHz. Where does that figure come from? Surely it can't be derived from Bruce Tainio who claimed in the early 1990s to have found a relationship between frequency and disease. According to Tainio's measurements, the brain has a 'bio-frequency' of 72MHz to 90MHz — genius intelligence is at the upper end, apparently. Fans of the woo business will be delighted to know that you can buy 'essential oils' that resonate in the same range and so help you get a better brain. And not those nasty gigahertz frequencies, like 2.4GHz, that mess your brain up. I can't find any paper from Tainio that explains his conclusions, just references on essential-oil websites, found courtesy of the resident Overmind otherwise known as Google. However, if I suspect my neurons to be running at 70MHz, I'm going to be ringing the doctor pronto, assuming that I'm actually able to.

* Sorry Orb fans, this post has approximately zero to do with Minnie Ripperton done ambient stylee, but here's a link to the video if that's all you wanted. But thanks to the Orb for sampling so much of Naked on S.A.L.T. (Orblivion) to save the aggro of fast-forwarding through the film to find the monologues.

Posted by Chris at 6:10 PM | Comments (0) | TrackBack

June 25, 2008

Scientific method's death a little premature

Chris Anderson of Wired has declared scientific method dead. And it's all thanks to Google, apparently, and the mass of data it is accummulating. Maybe Google really is making us stupid after all because the reasoning behind Anderson's conclusion is built on some shaky foundations.

Did Peter Norvig, Google's research director, really say: "All models are wrong, and increasingly you can succeed without them"? Because, if so, he seems to have misinterpreted what his own company has been doing. Yes, search and its related technologies do not rely on language models. But the core of all that Google does right now is based on a statistical approach that makes some basic assumptions about how language works. You might call it a model.

Anderson postulates a world based on machine learning, where the computer crunches through the data to come up with predictions.

"This is a world where massive amounts of data and applied mathematics replace every other tool that might be brought to bear...With enough data, the numbers speak for themselves."

Yet, machine-learning algorithms depend on the construction of some kind of model. It is not necessarily a deterministic model in the way that classical mechanics is, but just because it invokes statistics does not make it any less a model-based technique. What are models for? They allow you to make predictions about what will happen given some inputs.

OK, some branches of science are terrifyingly complex. Biology is the poster child for complexity. If you just take how DNA gets transcribed into RNA in a simple bacterium, there are thousands of potential interactions that get you to an RNA that will ultimately produce a protein. You get proteins sitting on the DNA that either encourage transcription or slow it down. Others bend the DNA round in weird shapes to activate a gene, but only when the conditions are just right. Yes, building a model of all these interactions is tough. But it is probably the only way of making sense of the processes and it is the way that biologists are making sense of the deluge of data. This is what systems biology is about.

They are using machine-learning and data-mining techniques to uncover patterns in the data. They are dredging through the seemingly countless genome and other 'ome databases to find data that they can plug into — yes, you guessed it — models.

Professor Jaroslav Stark of Imperial College sees modelling as a key to understanding what goes on inside living systems precisely because models are often inaccurate. For him, the fact that a model diverges from reality provides important clues to interactions that need to be taken into account. And they can provide a way to probe interactions where it is simply not possible to use traditional methods such as turning genes off selectively because that introduces other interactions.

The problem with Anderson's argument on this point is that, because what gets taught at school on biology has turned out to be inaccurate, we are getting further away from understanding through models. But that is what science is like: it finds new information, assimilates it and moves on. The biologists aren't finished yet, and aren't likely to be for another 30 years or so, even if they're lucky.

Anderson cites the work by J Craig Venter to sequence bacterial life in the oceans. A yacht is sailing around the world with a bucket to collect samples that get progressively filtered down until all you have left is bacterial DNA. This then gets dumped into an massive array of gene sequencers that randomly chop up the DNA with enzymes to produce fragments that can be separated to indicate the nucleic acids they contain. Computers then attempt to crunch through that data to reassemble the sequences into individual genomes. In practice, it's not possible to do that final step. At least, not right now. But, it is possible to see how much genes diverge among similar bacteria.

Venter has not really discovered unknown species of bacteria as Anderson writes because genetic sequence alone does not identify a species. Some of the putative genomes are very different to others, but Venter himself says that there is no percentage difference between genomes that will indicate a new species.

Basically, to identify a species, you have to go and look at how it lives and what it looks like. Maybe there is a shortcut to that process that involves the genome but until biologists fully understand the interplay between genes and the other bits of the genome, that is not going to be possible. It's probably easier with bacteria as they have comparatively little junk DNA, but it could still take some time. And the only way to build that model — even if it's a statistical one — is to assemble the genomes individually and examine the organisms. Not simply take a best guess as to how millions of fragments might match up in a genome.

What did Venter's team find? Based on predictions of the proteins that the assembled genomes produce, it seems that bacteria can have tuned versions of the light-sensitive protein proteorhodopsin. A single amino acid change in that sequence can alter the wavelength of light that the proteins absorbs and helps convert to energy. But that did not come just from blind number-crunching of the kind that Anderson suggests is the future. It was based on having a model of how rhodopsin works and then matching the gene data to it. Statistics helps, but there's still a model in there.

Big computers can certainly help with the creation and execution of models. But it seems unlikely that unleashing petaflops and petaflops on a problem blind is going to do much for machine learning.

Update: Now Kevin Kelly has chipped in, citing Google's translation system as evidence for the "stick it all in the Overmind/OneMachine" approach. Statistical language models have been kicking the structural models around the park for close to 40 years, and the techniques that work for search have some features in common with those that work in translation. What's happened with the web is that researchers have access to a huge corpus of text on which to train the systems. People are still working on the algorithms and they have to carefully pick the training corpus so as not to pollute the learning algorithm: the computers are just doing the boring legwork.

Kelly discounts idea of the approach killing scientific method. But dreams up a new term for it: "correlative analytics". This is hardly new. And questionably useful. As Robin comments below on the original version of this post, the finance community has been there, done that. Momentum trading is one 'algorithm' at the simple end of the spectrum. But it's basically taking outputs from a system and trying to use them as inputs. Not surprisingly, the results aren't all that spectacular.

However, if people want to believe that they can teach their computer biology by stuffing it full of all the genomics, proteomics, and other 'omics databases they can lay their hands on, I see no harm in letting them do it. However, the people doing real work on this stuff will be asking themselves: how was the data collected; what were the conditions? In short, while they may not read the data, they will attempt to understand how it came into being and then try to fit it into a model. It will get easier to automate some of those steps as labs adopt more standardised ways of generating the data, but we're still a long way from just stuffing bytes into a machine and let it figure it out for itself.

Posted by Chris at 12:04 PM | Comments (2) | TrackBack

And your effect on this company is...?

In 2003, Bill Gates channelled just about every user of Windows and its arcane ways in a memo dredged out of the antitrust actions by the Seattle PI. All he wanted to do was download Moviemaker but the Windows designers had other ideas:

"So I gave up and sent mail to Amir saying - where is this Moviemaker download? Does it exist?

So they told me that using the download page to download something was not something they anticipated."

It did not get better for Billg and his download past that point. However, Todd Bishop's post has a sting in the tail. He asked Gates on his departure about the email, sent almost five-and-a-half years ago:

As for the message, Gates smiled and said, "There's not a day that I don't send a piece of e-mail ... like that piece of e-mail. That's my job."

When people ask what Microsoft will be like now that Gates has left the building, this memo and the idea that Gates sent lots of them should be the clue. Nothing. Because if any of these memos had any effect, Windows would be a rather different piece of software. The structures that Microsoft built over the last 30 years effectively nullified any direct control that Gates had over software development. I'm sure people who weren't directly responsible for the problems Gates had with the download nodded and agreed with what he had to say, and they all listened intently to his speeches. But they then went on their way to product-planning meetings that not only created these hindrances but ossified them into place.

Posted by Chris at 11:00 AM | Comments (0) | TrackBack

June 24, 2008

Symbian's open road leads away from the smartphone

The obvious question when faced with today's decision by Nokia to buy out Symbian and release the software as open source was: if you have shipped 200 million handsets, what was the problem that forced you to do this? During the presentation that attempted to explain the move, executives such as Nokia executive vice president Kai Öistämö used the not-so-convincing argument that because Symbian has a 60 per cent share of the market, having charged up to $5 a handset to manufacturers, everything was going to be even better now that it is going to be free. Somehow, making it open source would dragoon in a bunch of application developers and convince everyone that Symbian is the only game in town in handsets. Forget Android, forget Limo and definitely don't bother about the closed-like-a-clam Apple iPhone.

Yet, despite having had ten years to build an unbeatable handset operating system, Symbian almost stumbled at the last hurdle. Nokia's majority ownership of the software maker has been a stumbling block with manufacturers, some of whom chose to build other user interfaces on top of the operating system to prevent Nokia from maintaining a stranglehold with the Series 60 environment. That is where environments such as UIQ and MOAP – used largely in Japan – have come in.

The situation has irritated operators such as Vodafone who find themselves having to deal with three different flavours of mobile phone built on ostensibly the same base when they have tried to tie back the number of platforms they support. Several years ago, Vodafone decided to try to restrict the amount of time it spent on software by picking three platforms: Limo; Microsoft; and Symbian. The idea of being able to bring Symbian back to one piece of software is far more attractive than the current situation.

The other problem for Symbian is that, since the iPhone, everyone has stopped caring about the operating system. All that mattered is who owned it and now, although Nokia takes on all the developers, the manufacturers seem happy with the source code going into a foundation. But, when it comes to phone design, it's now all about the user experience and having three or four companies slug it out over calendar applications and the like is not helping them sell more phones or more airtime. That is where Series 60, UIQ and MOAPS come in: they are where the action is for the handset makers and operators. The underlying operating system is simply a substrate.

This is where the picture of what actually happens with the Symbian Foundation gets hazy. Nokia is already some way down the road of deciding what software will be open-sourced, a process that could take two years to complete. The other players have agreed to sign up but it's unclear how much of UIQ and MOAPS will be in the final environment. Lee Williams, senior vice president for S60 software at Nokia said he reckoned the core environment will be based largely on S60 with components coming from the other guys. UIQ, for example, has a lot of touchscreen support that S60 does not.

Although the UIQ shareholders have agreed to back the Symbian Foundation by offering the environment up royalty free, some things have still to be decided, said Alain Mutricy, senior vice president of Motorola. "We are discussing right now with the UIQ management team how to restructure UIQ within the new ecosystem that is created by this move." He added later: "If we talk about two years down the road, we have to discuss with the shareholders and management team how their business model will adapt to the new ecosystem. But we will contribute the UIQ technology as and when the foundation is established."

For Vodafone, Symbian is still not the only game in town. Guido Arnone, director of Vodafone, said the company is continuing to work on the Linux-based Limo. "There is some competition but I believe it can be very complementary. Limo is relevant to higher end mobile touchscreen devices. Symbian is more lower tier. There is an air of competition for sure but there are areas where they are complementary to each other."

Tommi Uhari, executive vice president of STMicroelectronics, said he thought the removal of the royalty would help lower the point at which handset makers use Symbian. Although it was originally devised to drive the smartphone business, it now seems to be heading towards the featurephone. There is still the Linux option for handset makers and operators looking at the business Symbian might once have expected to command: the higher-end smartphones and mobile internet devices.

Posted by Chris at 4:30 PM | Comments (0) | TrackBack

June 20, 2008

A funny kernel panic happened on the way to the supermarket

People in the computer business just can't resist those Moore's Law versus the car analogies. Today's exhibit is Professor Steve Furber of the University of Manchester:

One litre of fuel would serve the UK for a year and oil reserves would last the expected lifetime of the solar system - if efficiency in the car industry had improved at the same rate as in the computer world - a leading computer scientist will tell an audience in Manchester, UK, on Friday 20 June 2008.

I bet he won't be telling them about the motorways clogged with automobiles stranded at odd angles as their drivers phone into call centres to be told: "Just try taking the battery out, then put it back in and start the car up. We can see if it happens again."*

Sorry, it's an old joke, but someone's got to do it.

* I once hired a Smart ForFour with an ECU that crashed so badly - in the middle of Wimbledon in rush hour - the only option was to reboot the car. When I next hired a car from them, I noticed that the ForFour was no longer on the list of vehicles.

Posted by Chris at 11:13 AM | Comments (0) | TrackBack

June 19, 2008

More Mentor

I've posted a couple of pieces on the attempt by Cadence Design Systems to buy Mentor Graphics at the Shrinking Violence blog, which I've set up to mainly cover the electronics business as silicon heads into its final decade of Moore's Law scaling.

The current design is temporary, which is why it's on a standard Movable Type template but that will change.

Posted by Chris at 6:09 PM | Comments (0) | TrackBack

Bioscience can't take on all the ethics issues

Last week, a group of social scientists from the University of Nottingham released their report on the ethical problems facing the technology of synthetic biology. Commissioned by the Biotechnology and Biological Sciences Research Council (BBSRC), the report called for a "thorough review of existing controls and safeguards" to extend them to synthetic biology.

Not just that. The public needs to be involved and may even be in the position to stop certain kinds of research: "It is vital to recognise the importance of maintaining public legitimacy and support. In order to achieve this, scientific research must not get too far ahead of public attitudes and potential applications should demonstrate clear social benefits."

This is from a different section but covers similar ground: "Partnership with civil society groups, social scientists and ethicists should be pursued as a highly effective way of understanding critical issues, engaging with publics and winning support for emerging scientific fields. However, at the same time it must be recognised that this is a two-way process and that some ethically problematic scientific projects and potentially controversial technologies may have to be abandoned in order to maintain trust."

This all sounds good in principle. But it is a process that could lead to some seriously strange decisions being made as to which branches of biological research are pursued and which are terminated. For a good many of the ethical issues that surround synthetic biology do not lie in the research but in the application. And in many cases, the economics of the application.

Take biofuels, for example. The lead story in today's Guardian by Julian Borger and John Vidal covers the contents of the Gallagher report, due to be published next week, on the role of biofuels in the current food shortage. Organisations are trading numbers as to what degree the production of biofuel feedstocks has had on the supply of food. But it seems the team led by Professor Ed Gallagher, head of the Renewable Fuels Agency, has concluded that EU governments were wrong to set targets for biofuel adoption without taking land use into account.

According to the story, the report draws a distinction between the first generation of biofuels - basically those in production today - and the second generation which will expand the amount of plant material that can be used to produce ethanol and other fuels. My piece in the technology section looks at the role that synthetic biology is likely to play in the development of those fuels as well as the following generations, which will attempt to substitute hydrogen for carbon-based fuels.

In principle, feedstock plants for the second-generation fuels will not displace food crops. In some cases, they could be one and the same. The stover from maize could go into fuel production with the main crop being used for animal feed, if not corn-on-the-cob. As some crops, such as the one planned by Agrivida, will modify the genes of a maize plant, the latter is less likely - most of today's genetically modified maize goes either into fuel or animal feed - but there are companies working on bugs that can digest leaves and stalks without needing an altered plant. Which direction the industry takes will depend on how efficient the processes are and it is way too early to decide on that one.

However, that "in principle" is a big assumption. There is no guarantee that a technique based on synthetic biology can guarantee that economic conditions will lead to an undesired effect. One, as yet unanswered question in the use of 'waste' biomass for producing fuel is how much of the plant do you need to lead behind in the ground. If you removed everything during the harvesting process, you would be creating the conditions for a dustbowl. How much biomass need to remain behind is currently an open question. In the case of corn stover, it is far from being all of it. But there is an upper limit. It's a similar situation for rice, rape, sugar and the forests that could potentially produce bug-ready biomass.

There is no-one involved in biofuel-technology research who wants to create a dustbowl. Or, if there is, I haven't found them yet. However, it is economic incentives and regulations such as those imposed by the EU on the current generation of biofuels that can distort a market effectively enough for the conditions to be right to wreak havoc on the agricultural system. It is practically impossible for any researcher to design out that possibility from what are going to be pretty fundamental technologies. And the more you look into an area such as synthetic biology, you realise how many different research themes are intertwined. Even if you thought banning one of them would prevent the dustbowl scenario, which one would you ban?

It is worth bearing in mind that previous interventions using food crops in the widest sense have caused more localised disruption. The World Bank and others spent years telling some countries they should plant more coffee, only for those robusta producers to find they were feeding their produce into a massive world glut, and not from potentially more lucrative cash food crops. It's hardly a surprise that similar things should have happened with the biofuels business.

Over-promise is a problem with any technology. LS9's contention that it could serve the US demand for petrol with a biorefinery the size of Chicago seems not unreasonable, especially when you divide that down into smaller producers, although perhaps not as many as Rob Carlson envisages - he is looking forward to the day when every home has its own refinery. However, what is not clear is just how much sugar or biomass has to go in through the gate to produce 140 million barrels a week. The US Department of Agriculture was a lot more circumspect about using biomass for fuel in its 2004 report: a third of the US fuel demand looked ambitious then and it still is, although it is not impossible as long as some technologies come good.

It's hard to argue against having a more open dialogue in and around science. But pushing the ethical debate too far upstream is not going to achieve the effect that anyone wants. Except for the fundamental ethical issues - things such as cloning - it can work. But, very often, the potential ethical nightmares do not reveal themselves until much later. Blaming basic science for them is not going to achieve better results.

Posted by Chris at 2:56 PM | Comments (0) | TrackBack

June 17, 2008

Mentor's big decision

There is clearly something in the water on the West Coast as hostile takeover fever is taking hold. Away from the Microsoft/Yahoo soap opera, another, somewhat smaller bid battle is gearing up. Mentor Graphics has rejected today's offer from Cadence Design Systems, citing antitrust issues among the reasons:

"As we recently indicated to Cadence, we reviewed Cadence's proposal and analyzed both the price proposed and the risks associated with obtaining antitrust approval for a combination between the companies,” said Walden C. Rhines, chairman and CEO of Mentor Graphics. "Following this review, we concluded that not only was the price insufficient to support a transaction but that the risks of not gaining regulatory approval were sufficiently high that the ability of the parties to consummate the transaction would be in jeopardy. For these and other reasons, our Board unanimously rejected the proposal."

On the conference call, Cadence did not distance itself from the idea that CEO Mike Fister could play the role of Steve Ballmer against who some analysts are setting up as the Jerry Yang in this battle, Mentor's CEO and chairman Wally Rhines. The script is similar: Mentor did not want to negotiate, and is not interested in providing value to shareholders.

However, Fister could equally be the loser in this particular contest as it's not entirely clear how much Cadence would gain from the $1.6bn deal against the risks of having to suffer scrutiny from either the Department of Justice or the Federal Trade Commission over antitrust issues. We are dealing here with two of the top three companies in the electronic design automation market. It's worth about $3bn, with almost all of the money going to just five companies.

Even if it succeeds, the risks to Cadence are huge. The deal is only 'accretive' — it boosts the earnings per share — if you assume pro-forma, non-GAAP accounting. The kind of accounting where the CFO decides which numbers are relevant rather than actual accounting standards. And it does not take into account the effect of having to report Mentor's numbers in a different way to which they are now.

Anyone who has followed EDA financials for any amount of time will curse the words 'ratable', 'revenue' and 'recognition'. These words define when the software companies decide to book the money they take for their tools. And they all do it differently. Not only that, every three or four years, they change the way they do it. This makes comparisons over any lengthy period of time a nightmare.

The upshot is that, if Mentor does succumb to a proxy war or decide to agree to a merger, the numbers will not quite be as good as everyone expects because Cadence will delay booking some of the money that the Mentor business units take in. Now, this is arguably a cleaner approach for a company that sells software licences, but it makes predictions of the value of the Mentor deal much harder to calculate.

Then you have the product risk, which is what most of the people watching this deal will care about. Gabe Moretti invokes the spectre of the Daisy/Cadnetix takeover from the early 1990s. This did not go well, although you could argue that these companies were on the way out anyway and they might as well have gone down together. This was the point where upstarts like Cadence and Synopsys — today's top two — were coming through. It's worth noting that Mentor stumbled at this point, reinventing itself under Rhines during the 1990s when he crossed over from chipmaker Texas Instruments.

Overlap is a big issue. On the Cadence conference call, Merrill Lynch senior analyst Jay Vleeschhouwer, who has covered EDA for some time, pointed out that Cadence has more than a little overlap with Mentor: "I can think of five areas where you and Mentor currently overlap, including verification and physical," he said and cited the example of Synopsys and Avant, where the two companies had comparatively complementary product lines. "It took Synopsys years to get the full benefit of Avant."

It's actually easier to count up the areas where Cadence and Mentor don't have competing product lines. It basically comes down to FPGA design and embedded software and, on the latter, Cadence has 'Project Sydney' coming up, which has a software verification component to it.

Here's Fister on the overlap problem raised by Vleeschhouwer: "I think the EDA market is full of subsegments. And certainly now many of the customers operate heterogeneous flows, a combination of points of light. That is where the extreme complement comes from. In my time I think we have demonstrated an ability to integrate not only the historic acquisitions but the ones we have embraced in the last four years. We developed a great capability to do that; the stucture of the holistic solutions engages in a collaborative fashion some of the rest of the industry and that is the muscle memory that we are going to use going forward." (Translations on a postcard please to...)

Bill Porter, chief administrative officer and formerly CFO, added: "Part of our strategy is to have a holistic solution, front and back. And part of Mentor's philosophy is to have very distinct points of technology throughout their product line. Our ability to integrate that will have benefit to our customers. We will be able to get on that very quickly....I think it has got some long-term benefits for our customers."

It's a fair point that Cadence has made acquisitions work for it: the company has a patchy history of turning internal R&D into top-notch tools. It has claimed, recently, that it is much better at it. However, the Mentor deal has demonstrated that Cadence believes Calibre is far superior to its own offering, Assura, which is a dramatic about-turn from a company that has busily been trying to push into the design-for-manufacturing (DFM) space. It is because of this that Cadence is now in the position that it cannot retreat from the takeover bid: it's signalled its weakness in certain key areas.

Is the deal worth it? For Cadence, the risks are huge and could turn out to be a 1+1=1 deal. Calibre will bring in the dough from day one but Cadence will have the issue of how it merges Assura with Mentor's tools. Mentor's PCB operation is a steady earner and the nature of that business means that Cadence would not have to spend much time merging tools. You just keep all of them that have a decent user base. Verification is far from clean-cut. It's hard to see the ModelSim team settling in well at Cadence. It took a while for them to settle in at Mentor and some of the other business units work because they have relative autonomy. That, historically, is not the way things have worked at Cadence.

I still need to do a full analysis of where the overlaps lie. But, trust me, the diagram is going to look messy. And, I believe that both Cadence and Synopsys are over-playing the concept of the one-stop-shop. The big chipmakers do want to buy a single design flow but there is a caveat: that flow has to have all the best tools in it. And none of the players can offer that. They won't be able to offer it even if Cadence succeeds in its purchase of Mentor. What they really want is some third-party to make the pain go away by taking the best tools around and glueing them together. This is not the same story that Fister is selling to Wall Street.

Mentor is also faced with a dilemma: a takeover might be bad for Cadence; it might be bad for EDA; but is it bad for shareholders? Let's face it, it's real money on the table. Last year, $16 was not a great offer. But, right now, it looks somewhat more attractive. Shareholders feeling a little rattled from the rollercoaster returns of the sector over the last ten years might feel like running, not walking, for the exit. Or, if they are feeling lucky, might want to take the cash and stick it in Synopsys, unless Aart de Geus has a rush of blood to the head and decides to bid for Magma Design Automation, while Cadence struggles to digest Mentor. A lot will depend on how Mentor's institutional shareholders fall on this deal.

Posted by Chris at 10:10 PM | Comments (4) | TrackBack

Turn the voltage down, you're stressing me out

Intel came close to giving the idea of having a fixed clock-speed rating on its upcoming Nehalem the heave-ho, according to Intel fellow Rajesh Kumar, speaking to journalists ahead of the VLSI Circuits Symposium in Hawaii this week. The people who were going to be putting the processor into PCs didn't care for the idea, it seems.

The company has radically altered the way that Nehalem is clocked compared with its predecessors in order to improve both memory bandwidth and power consumption. It means that the core, memory buses and I/O run almost independently.

The bigger change is internal, where it seems that the concept of a fixed clock running at several gigahertz has been discarded in favour of letting the logic run at its own speed. This is something that people such as former ARM architect Professor Steve Furber have been advocating for years. The concept of a system clock is entirely artificial and exists largely to make life easy for chip designers and simplify the job of testing chips as they come off the production line. Chips such as the Amulet don't run off any kind of clock: the logic inside finds its own speed.

"The idea is not new," said Kumar. "But the implementation is new."

With a conventional clocked design, the architects work out how much logic each part of the pipeline can perform within a cycle. They then add on some slack to cope with the vagaries of manufacturing. For decades, this has worked pretty well although knocking some bits of the pipeline into line can have engineers tearing their hair out. "Negative slack" is not a term they like to hear.

Now, everybody cares about power consumption and the synchronous techniques of the past don't look so attractive. One problem is that having to add the guard bands increases power consumption because some part of the system are struggling to keep up. Others produce a result and sit around twiddling their thumbs waiting for the end of the clock cycle. The first circuit needs as much voltage as you can feed it — the higher the voltage, the faster transistors will switch. The second one you could afford to feed with a lower voltage, so that it slows down to the point where it comes up with result just ahead of the next clock cycle.

Where it gets worse is that, in modern processes, every transistor works slightly differently to its neighbours. Some are faster than expected; others are slower. To make sure everything can keep up, you have to increase the guard bands. And that pushes the overall power consumption up.

Kumar claimed it's now different with the Nehalem: "We have introduced a chip that adapts every cycle to the dynamic power. We can get higher frequency and lower voltage."

According to one of Intel's slides, the "duty cycle adapts to transistor variation and lifetime stress". Yes, not only do transistors come out of the fab with something of a spread in terms of performance, they change as they get used. Some will fare much worse than others in usage, slowing down over time. Higher voltages do not do sub-micron transistors a lot of good. So, being able to adapt to that change in speed is important.

However, the idea of a sort-of-3GHz processor did not appeal to Intel's customers. "We debated this quite a bit while doing the implementation," said Kumar. A less clock-centric approach was "the obvious way" to go, he added. But, "We realised quickly that people did not want that. They hated the idea of asynchronicity and indeterminism. So a tremendous amount of innovation has gone into avoiding that. We spent a lot of time working on that. Internally, the chip is adapting but, from the outside, it is deterministic."

In effect, there is an averaging process that goes on to ensure that people do not wind up with processors that run at subtly different speeds long term. "Every few seconds, it is averaging so that, from the outside, it is running at a fixed frequency. All the time. There is no indeterminism."

It's important for Intel to hide a variable clock cycle. But the idea of a stretchy clock cycle is something we may see from other quarters. ARM's director of R&D Krisztián Flautner studied at the University of Michigan where the Razor concept was developed.

Razor works on the assumption that circuits run at maximum speed for a given supply voltage and that, if they cannot meet timing, the calculation will seem to fail. But the correct result will become available at some point once the circuit has stabilised. The trick is knowing when this happens — so you add some logic to watch for this. The additional Razor logic cancels the false result and fetch the correct result from a shadow register. In effect, the circuit runs speculatively but can be corrected after the fact.

“The concept here is that we treat part of the cycle time as the error region; the place where we might screw up,” said Flautner last year.

The team has developed a self-correcting flip-flop and a memory cell, where a second sense amplifier detects the error. One of the issues is that the circuit is no longer deterministic — the problem that Intel wanted to avoid — as errors need to be fixed after the fact.

However, with this kind of approach, chips could be 40 per cent more energy efficient than they are today, according to Flautner. But runtime approaches such as Razor incur their own energy and area overhead. He said the area overhead was relatively unimportant: the question is whether the power drawn by the additional Razor logic outweighs the potential saving from being able to run circuits much closer to their timing margin by reducing their supply voltages.

Does the circuit needs to be deterministic? Flautner said this issue was confronted by software engineers when caches were introduced on embedded processors.

Posted by Chris at 6:28 AM | Comments (0) | TrackBack

June 12, 2008

You wait ages for a GPU programming environment to come along and then...

A couple of months ago, nVidia's Jen-Hsun Huang decided to stick his head out of the window and shout he wasn't going to take it anymore. Or at least, gather a bunch of analysts together at the graphics chipmaker's HQ and tell them he wasn't going to take it anymore. The trigger was Intel's developer forum in China where Intel's Pat Gelsinger declared the death of today's graphics processor (GPU). Curiously, Gelsinger claimed that just ahead of talking about Larrabee: Intel's latest foray into the GPU business (it's a different kind of GPU, you understand).

The argument from the Intel side was that traditional processors would take over many of the rendering functions in 3D graphics, largely because there are going to be so many of them. Huang had the opposite argument: GPUs already have lots of processors on them, why not use them for offloading software from the host processor?

And so the stage is set for a new kind of architecture war in which you have different kinds of microprocessor fighting over the same ground.

At the analysts meeting, nVidia lined up a bunch of demoes from people operating in supercomputer-land who have tried plugging together thousands of Xeons and Athlons and decided they have had enough of it. They want something new. IBM's Cell looked promising for a while but a number of supercomputer users have decided that it does not deliver quite what they want, despite the headline performance claims over Roadrunner.

Their main options are GPUs and field-programmable gate arrays (FPGAs), chips that let you define whatever hardware circuitry you like. FPGAs are not that great on the kind of floating-point code that host processors can handle but they tear through things like genetics and chemical-matching programs.

The problem is finding a programming environment that will handle all of the above. The supercomputer users want to be able to work will a combination of x86, GPU and FPGA-based processors. There isn't anything out there that will work across more than one type of accelerator. Got an nVidia card? Use CUDA. But that's no good if you happen to have an ATI GPU. For that, your only option is the long-in-the-tooth Brook environment. Intel is lining up its own architecture under the banner of QuickAssist. Its current implementation is oriented towards FPGAs but the scope is likely to widen as Larrabee gets closer to shipping.

There is an open effort under the OpenFPGA banner. But, again, it is, as its name suggests, focused on FPGAs. A number of the specialist vendors in the supercomputer world like the idea of OpenFPGA because it has potential as an independent standard. However, the amount of money behind the other players suggests that we are likely to be looking at a vendor-derived standard emerging.

But the GPU programming environments will, with the exception of the FPGAs handled by QuickAssist - as Intel hasn't made FPGA since the mid-1990s - be oriented to vendors's own hardware.

Then there are the unplayed hands of Apple and Microsoft. Apple provided hints of what it is doing in this area on Monday in talking briefly about the Snow Leopard release of Mac OS X:

Snow Leopard further extends support for modern hardware with Open Computing Language (OpenCL), which lets any application tap into the vast gigaflops of GPU computing power previously available only to graphics applications. OpenCL is based on the C programming language and has been proposed as an open standard.

This, potentially, is a smart move on Apple's part. The company would benefit not just from having one language to work with all the GPUs it could buy in from AMD, Intel or nVidia but gets a chance to wrong-foot Microsoft. The software giant has projects to look into acceleration using GPUs and FPGAs - a lot of the work is being carried out by people such as Satnam Singh at Cambridge - but has very little in the way of product plans. Apple gets the chance to define the programming model of the future, giving it a lot more architectural control than it has now, and come out looking like the good guy by providing a software layer that works with a much wider range of hardware than anything that the chipmakers plan to offer.

At this stage of the game, unless Microsoft is genuinely ready to go public with an API, the only realistic counter-proposal would be for nVidia to port CUDA to other architectures. True, AMD could propose the same thing but it is coming from further behind. It's hard to see anyone accepting an Intel-proposed API unless they really have no other choice.

There is plenty that can go wrong with OpenCL. Programming GPUs to run code is one thing. Getting the code to run as fast as you'd expect 128 processors to do it is another matter, according to supercomputer users, some of whom turned up at the MRSC conference in Belfast a couple of months ago to talk about their experiences. But, unless Microsoft can come out with a real API and library first, Apple is in a good position simply by virtue of being in control of a desktop computing environment rather than having to sell chips into existing platforms.

Posted by Chris at 9:42 PM | Comments (0) | TrackBack

June 4, 2008

Dear Tom Tom, this is not a road

In Sicily for a holiday in the second half of May, my girlfriend and I decided to go to Pantalica. It sounds as though it ought to be a South American heavy metal act but is an enormous, sprawling necropolis that dates back to the Bronze Age. From about 1300 BC, the inhabitants buried their dead in caves cut into the sides of the gorge cut by the Anapo river. They cut thousands of square holes in the cliffs and dragged the bodies of their relatives up to them, ultimately to be uncovered and shipped off to museums by archaeologists.

You can get to Pantalica from two directions: Ferla to the west and Sortino to the northeast. The roads almost meet, but not quite. However, the Tom Tom satnav shows one stretch of road joining Ferla and Sortino by way of Pantalica. Before we got there, I assumed that there was a road there but it was no more than a dirt track for the section that ran down into the gorge and up the other side, as local maps show a break between the two sections of tarmac. This was on the basis that in all the stories of satnavs going wrong, most of the time the road actually existed. It just wasn't all that useful to regular motor vehicles.

Not so this time. The path down into the gorge pretty much dates back to the Bronze Age. You have steep steps cut into the stone that have been there so long they have, in some cases, been eroded into rock pools. No-one's going be off-roading down in the gorge unless they've given their 4x4 mechanical legs.

For a while, I thought Tom Tom's fake road might follow the path. But, having had the chance to compare it against the satellite images in Google Maps, it seems the connecting bit of road is total fantasy. If it existed, it would be quite a bridge. But there is nothing there other than a deep, tree-filled gorge.

The bright patch in the lower half of the zoomed image is the Ferla-side car park. The road from Sortino runs out of tarmac as it comes in from the right. It turns into a dirt track that kinks up. Then, in real life, it stops: the way down is a narrow path that is almost invisible on the satellite image. You can see the path that snakes up and to the left from the Ferla side. This crosses the river, eventually, towards the top where the gorge bends round. Shown as an overlay on the top image, after a bit of Photoshop work, Tom Tom's road just seems to be a spline that joins the Sortino half from the end of the kink directly to the bit where it dissolves into a sand-covered car park.

From the looks of it, I'd guess a little bit of over-enthusiastic error correction has been going here. I can't help feeling that someone looked at the map data they bought in, saw a break in the road and thought: "That can't be right." So they 'fixed' it.

Posted by Chris at 5:09 PM | Comments (0) | TrackBack

May 16, 2008

The pain of creeping featuritis

Apple's Mail.app is beginning to drive me up the wall. Thanks to a bone-headed decision made in the Leopard update, a nearly useless feature has rendered a very useful addon almost as redundant.

For some reason, a UI genius at Apple decided that it would be just spiffing to have the program jump to an email when you hit a key, using the starting letter of the email's subject line as the destination. Unfortunately, the code conflicts with the handy Mail Act-on from Indev. This piece of donationware which comes from the same company that provides the also-handy tool MailTags, lets you assign macros to keystrokes.

I have three key-commands that I use all the time. One shoves press releases into the Releases bucket folder; one puts emails into the Invitations folder (which is slightly misnamed - it's the folder for everything related to interviews and meetings); and the third does the same but sets a @Waiting tag.

However, since Leopard came out, Mail Act-on now has to fight with the built-in keystroke detector so it works only about 50-60 per cent of the time. On an operating system where you find a lot of add-on software that adheres to the 'wei wu wei' philosophy, this is a real problem. It means having to go back to dragging releases with the mouse or watch the fight between a great little piece of software and some weekend hack that has almost no reason for existing.

You want to see how Mail deals with folders that contain emails from a mailing list - most will start with the same letter. I'm not sure I've figured out the logic of what it does yet.

Posted by Chris at 2:02 PM | Comments (3) | TrackBack

May 14, 2008

Infineon and the nuclear option

Steering a company into a near-suicidal megamerger has to be one of the more creative ideas to get rid of a chief executive that a chairman has ever had. But that seems to be the upshot of the story reported by the Financial Times Deutschland earlier today. The story has no named sources. But, at the same time, Reinhard Ploss, head of operations at Infineon Technologies, had to recruit vice president Eric Mayer to stand in for him at the IET/GSA Semiconductor Forum today while Ploss apparently dealt with things back at base.

According to the story, Infineon chairman Max Dietrich Kley thinks buying NXP Semiconductor, currently owned by private-equity KKR, is a good idea. Chief executive Wolfgang Ziebart disagrees strongly. There is no real way to reconcile these differences so one of them will have to go, and it will probably be Ziebart. As the story in FTD sums up:

Despite Ziebart's competence, "in a shark-tank like Infineon he is out of place," said a senior manager.

Joining from tyre-maker Continental in 2004, after former Infineon chief executive Ulrich Schumacher resigned, Ziebart moved to rid the chipmaker of what would prove to be his albatross: the memory maker Qimonda. However, he didn't move quickly enough. A few years earlier and it might have been possible to flog Qimonda on the public markets. However, under Schumacher, floating off the memory operation was never going to be a starter. The result was that Qimonda emerged into a crunching recession in the DRAM business. Ziebart was left to explain to shareholders why Qimonda was still around and dragging down Infineon's numbers.

Kley, it seems, is keen to blame the situation on Ziebart. But, in case anyone thought Kley might have a handle on the business, apparently has floated the idea of a purchase or merger with NXP. It's not hard to see why Infineon executives would be keen to leak this one: they know a calamity when they see it. The only winner in such a deal would be KKR, which bought NXP when times were good and has now found that any chipmaker would be tough to offload.

The problem with such a combination is not so much that NXP is in dire trouble but that any attempt to take on a merger of this kind would cripple both companies. Not only would you have the problem of resolving huge product overlaps, there would be the slightly crazy situation of Infineon's wireless chip business competing with the ST/NXP joint venture. It's not beyond the bounds of possibility that ST would wind up taking on Infineon's communications-chip operation. But, if you were an Infineon shareholder, you would wonder whether handing Carlo Bozotti a further boost in wireless is a good use of your money.

People such as NXP's Frans van Houten and Ziebart seem happier doing much smaller slice-and-dice deals that add a business unit here, lop one off there as they try to find markets they can lead. This is the sensible route in the fab-light world. Slamming two top-20 chipmakers together is an strategy from the last decade.

Posted by Chris at 9:53 PM | Comments (0) | TrackBack

May 13, 2008

Big is the new small

Last week, three of the world's biggest chipmakers decided that they were going to try to move to bigger wafers in the next decade. Today's wafers are the size of vinyl LPs: 300mm across. The next step up is that of a family sized pizza: an impressive 450mm in diameter.

The people who the chipmakers expect to build the equipment to handle these things greeted the news with...very little. They certainly didn't jump up and down, waving to Intel, Samsung and TSMC, shouting out: "Me, pick me!" They might consider investigating the feasibility of boosting wafer size to 450mm if these three were going to pay all of the development costs. But, they had played this game before, with the transition from 200mm to 300mm at the start of this decade and they weren't going to get fooled again.

Last time around, the big chipmakers - and there were about three times as many in the club as this time - said they wanted to do the move early, during the late 1990s. Then there was the Asian crisis and the chipmakers decided they didn't want to move after all. Instead, they decided to speed up the development of the processes rather than use bigger wafers, at least until around 2002. And it was the equipment suppliers left wondering how they were going to explain hundreds of millions of dollars in R&D for products that would not be needed for several years.

Since the end of the 1980s, the industry has not liked moving to larger wafers. Back in 1975, Gordon Moore used historical data to predict how big wafers would get - they were about 75mm across then. At the time, wafer size increased at roughly the same speed as the increase in average chip size. Had this trend continued, fabs would be working with wafers 1.5m across by now. In practice, die size increased faster than wafer size for about 20 years then manufacturers worked out that they could not push die size any further, and stopped.

There is still a good reason for going to larger wafers: you can get more chips onto each one. In principle, that cuts costs, especially for the chemical processes, such as etching stuff off and laying down metals. For the photolithography steps - the most expensive parts of the process - there is not that much of an advantage. However, moving to bigger wafers here does not have so much of an effect on equipment cost. With the other steps, because they need carefully controlled conditions, big wafers are a headache. But, if you can crack the problems, you wind up with cheaper chips at the end of the process.

For Intel, Samsung and TSMC the move is potentially a good way to gain a lead over competitors who cannot justify the move to 450mm. You can get 1500 chips, each measuring 1cm-square, on a 450mm wafer, versus around 700 on a 300mm wafer. That favours companies making chips in very high volumes. For Intel and Samsung, which ship bulk processors and memories, respectively, anything that lowers cost in volume is good. For TSMC, the situation is more complex as it serves a wide variety of customers. However, as the world's largest foundry, it has its eyes on the business of the other companies in the list of top-25 chipmakers. With 450mm, it could undercut smaller rivals such as UMC, SMIC and Chartered without denting margins. It's main competition would probably, by around 2012 when the transition is supposed to start, be Samsung.

However, they will have to do a lot of sweet-talking to the equipment makers or simply do a lot of the work themselves - it may be no coincidence that equipment makers' group SEMI just put out a report about getting their technology ripped off by customers. It is tempting to think that consolidation is concentrating manufacturing capacity in a few hands. In fact, Intel and Samsung have lost share since the move to 300mm.

In 2001, the two companies accounted for almost 22 per cent of chip sales. In 2007, that slipped to just under 20 per cent. If you tot up the companies who decided to install 300mm equipment this decade, they accounted for close to 60 per cent of the market in 2001. Estimating TSMC's share of manufacturing is tough, because you have to extrapolate it from what its customers sell. But it's probably close to Samsung's share today. Let's be generous and call it 10 per cent of the market, just for fun. That's a total of 30 per cent of the chip market.

As an equipment supplier, are you going to go with what 30 per cent of the market wants, knowing that it will damage the ambitions of any other companies that want to do their own manufacturing come 2012? If you pretty much want to guarantee that there will only be two foundries by the end of the next decade, you are going to bet on 450mm. But, knowing that a larger, more varied customer base is good for business, you are not going to bet in that direction.

There is a further problem for the big three. Some of the equipment suppliers have decided that markets such as solar power and displays look a lot more attractive right now than waiting for Intel to tell them how much they will get paid. Applied Materials' latest results, for example, show that display and photovoltaic sales have grown in the last year, while sales to their traditional silicon customers have slipped badly.

Posted by Chris at 10:04 PM | Comments (4) | TrackBack

May 6, 2008

The trouble with wireless power

Splashpower, the UK company that launched several years back with a popular promise - to banish the proliferation of power adapters in business baggage - is no more. Having gone into liquidation just over a month ago, its patents have been bought on behalf of competitor Fulton Innovation by its parent Alticor.

Fulton has its own wireless charging technology that it calls eCoupled. The Splashpower purchase means it gets a second tranche of patents, just in case this business ever takes off.

The problem that Splashpower faced, and which Fulton still has, is convincing electronics OEMs that they should dump a cheap charger, that uses cables, and replace it with something more expensive. Yes, there is the vague promise that, if lots of manufacturers go with inductive charging, you need only take one charger and its associated power pad with you on a trip to keep music players, cellphones, laptops and PDAs all nicely topped up. Just as long as all the manufacturers sign up to the same system.

The problem with this kind of charger is that all the real IP is in the protocol that gets the charger to talk to the gadget. With one less player in the market, that is more likely. But it's not guaranteed. And the promise of fewer chargers only works if lots of OEMs all sign up for the same system. When you consider that the main component of the IP is a communications protocol, that's a lot of power you give to the power supplier. And it's a big shift from today's market where you have 500 different Chinese manufacturers all happy to give you a very good price on a custom power adapter.

The Register is right on the money here talking about USB as a more likely unifier for the one charger to rule them all. For starters, there is already a burgeoning aftermarket in USB charging adapters. And, even if gadget makers can't agree on which USB connector to use, it's a lot more convenient to take a few small adapters on a trip than a bag full of wall warts.

So, what happens to wireless power? Stays right where it is: in situations where you don't want exposed power connector pads. Fulton's eCoupled technology started off in the water treatment business where corrosion would quickly destroy electrical contacts. And there are plenty of electric toothbrushes on sale. Why? Because they get wet too.

The one wild card in this business is the idea of microwave power transmission: beaming electrical power and signals to things like bookshelf speakers and light fittings from a transmitter in your living room. It sounds like a great way to get rid of wiring clutter. But just wait until the electrosensitives get stuck into that one.

Posted by Chris at 3:36 PM | Comments (3) | TrackBack

May 2, 2008

Toys for the front line

It was a with a hint of disdain that Raglan Tribe nodded across at the teams placed next to his company’s table. "You see some of the other teams? You have one right here. They have all this equipment. We have just got this one vehicle," he said, pointing to the small battery-powered car sitting beside him.

Until Tribe and his colleagues at consultancy Mindsheet stripped it down and gave it a new electronic brain, it was a regular toy car. Now, it’s a wheeled robot that is meant to find guerilla marksmen hiding in alleyways and buildings. But Mindsheet was flanked by two teams that are not taking any chances in the UK Ministry of Defence Grand Challenge final in the summer: they have one-time nuclear-waste inspection and bomb-disposal robots that are meant to work alongside purpose-made aircraft and off-the-shelf gliders. The venue was a conference to show off what Grand Challenge teams had developed so far to visiting military chiefs and researchers.

The Mindsheeet vehicle was meant to be small and easily portable, not like some of the vehicles sitting nearby, including a UAV with a 1.5m wingspan. Tribe said his team talked to soldiers who fought in conflicts such as Afghanistan to find out what they needed. Soldiers are pretty consistent in what they want: something light, easy to assemble and reliable.

At a seminar last year to kick off the competition, which is meant to uncover robotic technologies that can detect threats in the kind of wars that the British Army is now fighting, in countries such as Afghanistan, soldiers talked about the problem. They want to be able to see a lot more when on patrol. But, when your pack already weighs 70kg, anything else that goes in has to earn its keep.

The armed forces found that the large UAVs they have used are not always that effective. It is often difficult to book time on them and they can't see everything. The problem for the army in Afghanistan is that it is often working around the walled compounds found in the country's towns.

The current UAVs work best in rural areas. Major Giles Timms, who commanded B Company on a tour in Afghanistan, said the UAVs they had access to could not see through the tree cover present in many of the environments his soldiers had to patrol in Sangin Valley.

Timms said the company had used the Desert Hawk UAV made by Lockheed Martin but that it was "a bit hit and miss with us. It was quite difficult in that terrain. With big, thick heavy canopy, the enemy can hide easily. It has utility but it has limitations".

In towns, soldiers want roving eyes that are smaller and more agile, able to see over walls but keep below the trees and get closer. "How do I tell the difference between a gunman, someone with an RPG or someone carrying scaffolding poles. Very difficult to do with [today's UAVs]," said Lt Col Ian Thompson of 3Para last year.

That is where the Grand Challenge comes in: find companies with designs for much smaller, cheaper and manoeuvrable robot vehicles. However, the reality is that any robot vehicle that goes out on patrol probably won't have to displace equipment in a backpack. "Where I believe we will end up is with a piece of equipment that [is carried around] inside a helicopter," Thompson said.

At the end is a trophy, cast from metal taken from a Spitfire. Unlike the US Grand Challenge competitions there is no cash prize: HM Treasury rules don't allow it. But the aim of most of the 11 teams taking part is not to win the trophy but to pick up the research-and-development contracts that will result from any piece of technology that piques the interest of defence users. The Grand Challenge is not so much a competition as a showcase for technology the MoD might not otherwise find. And some teams want to make sure they have all their bases covered.

The two teams that flanked Mindsheet had at their disposal ground robots and gliders. The Stellar team’s entry will use a land robot built on the chassis of a bomb-disposal vehicle that will work with not one but two different UAVs. The Silicon Valley team has two robotic vehicles from Moonbuggy - one of them used for surveying radiation-contamined land - and an regular model glider fitted with cameras under the wings and GPS receiver.

"We have got our eyes on what comes after the competition and designed our whole architecture around that. The system has got a lot more built-in functionality that would be used in more representative scenarios," said Steve Fernandes of Selex Galileo, which is part-funding the Stellar team. "What we have been doing is engaging the military customer to look at the scenarios and how the system could be deployed. We have tried to look across all the lines of development. We have tried to look at logistics and how the system fits in with other military equipment. And, most importantly, how it would link into the large systems currently deployed, such as Bowman."

But the link into the military networks has not escaped Mindsheet, as some companies involved with the Grand Challenge also want to see what small robots can do. Two teams, and Mindsheet is one of them, took up an offer from MBDA to have its cars hook into the defence-technology supplier’s network to download the data they capture.

Many of the teams have adapted off-the-shelf hardware. Some, like Mindsheet are using adapted toys. Model helicopters are particularly popular, although their petrol-driven engines are noisier and will attract more attention from a marksman than the custom-designed battery-powered aircraft that were on show. Teams went with the readymade hardware because time was tight – the competition did not launch officially until the spring of last year – but also because a lot of the focus is on sensors and getting the vehicles to work autonomously. How much time vehicles can keep running is a problem – many of them will run for no more than a couple of hours on a full charge – but teams believe they can work on that later.

Bill Bailey, a consultant to Selex Galileo and former head of intelligence in Afghanistan, explained: "They are just platforms. If someone comes out with a better platform, we can use that. And they have to be cheap because they will need to be replaced."

August is when these machines will be put through their paces in what is something of an experiment for the MoD as it tries to get smaller companies and universities to supply it. A total of 11 teams have made it to the final. Some will be happy to compete for just the trophy, but most are hoping that what they really get is an R&D contract.

Posted by Chris at 11:55 AM | Comments (0) | TrackBack

May 1, 2008

Resistance to memory

It's taken more than 35 years to find, but it looks as though HP Labs has found a cousin to the resistor and the capacitor hiding in the delicate thin films of metal oxides.

Naturally, HP Labs talks up the prospects of the memristor. One application the researchers have put forward is as a potential successor to devices such as the venerable DRAM. On the face of it, this is arguably the worst place to go. The industry is littered with 'nearly there' memories that have better properties than those apparently on offer with the memresistive approach. The HP labs team namechecks a bunch of materials in the Nature paper that show memresistive-like effects. However, at least three major categories are already in production or have multiple teams working on them, with varying degrees of success.

Chalcogenides are the materials that go into phase-change memories of the kind being pushed by Numonyx - the JV formed by Intel and STMicroelectronics. To give you an idea of how long it can take to get a memory technology off the ground, phase-change memories have been around about as long as Leon Chua's theory of the memristor. And you still can't buy one in the shops. On top of that, the phase-change memory is meant to be non-volatile: it doesn't forget stuff when you take out the battery.

The memristor will not be a non-volatile memory but only semi-non-volatile, according to the researchers. It seems that, like a capacitor, these things 'leak' a little. Leave it too long, and it will have forgotten what you told it.

This is a problem that afflicts the latest new memory technology: metal oxide, which is also being touted as a possible candidate for the memristor treatment. Metal-oxide memories are programmed by heating. Unfortunately, right now, just storing them at room temperature provides enough energy after a few days, or even hours, for them to reset themselves.

Then you have the perovskites, such as barium titanate. These are already in use in ferroelectric memories. You can go out and buy these but it's another memory technology that never quite made it to the mainstream.

However, it seems that something like memristor behaviour has been seen in organic materials. This may be the way forward as it points to the possibility of being able to print memory devices using organic chemicals. These kinds of material make pretty rubbishy transistors, but they might perform better as memristors.

The part that might lead to radical changes in computer design is the observation that memristors work in a similar way to the Hodgkin-Huxley model of the neuron.

One big problem with nanoscale electronics is variability: these things are so small that there's way too much of it. This makes it tough to build reliable binary switches: the primary use of a conventional transistor. But, what if you don't want to make a switch? This is the kind of work being performed by researchers such as Professor Steve Furber's group at the University of Manchester with the EPSRC-funded Spinnaker project. The idea behind the project is that you dump binary logic in favour of a system that lies on statistics. In that kind of environment, manufacturing variability is not necessarily your friend, but it's way less of an enemy.

The inspiration for the work is the brain and the way that neurons communicate with each other. In Prof Furber's model, you use a bunch of them together to effectively vote on a calculation. The overall elements wind up bigger but you use the elements to process more information than just binary bits. Right now, the team is using arrays of ARM processors to model neurons. However, if the work pays off, it might point to a simplified system that could be implemented using either nanonscale transistors or elements such as memristors, which have the advantage of working more like a neuron out of the box, as it were.

Posted by Chris at 9:12 AM | Comments (0) | TrackBack

April 29, 2008

It's engineering, Jim, but not as we know it

When synthetic biologists talk about what they are doing, they often point to the analogies between their work and what happens in engineering, particularly electronics engineering. You can point to some processes in living cells and describe them in the same terms as digital logic or oscillators - the kind of functions you find in a lot of electronic circuits.

The analogies don't stop there: the aim of synthetic biology is to develop a kit of parts from which you can build organic systems able to make fuels, drugs and chemical sensors. What are the parts? Professor Richard Kitney of Imperial College, London says: "We mean encoded biological functions: usually we mean modified bacterial DNA."

That modified DNA is injected into bacteria which has the machinery already in place to do the next bit, which is to make the parts work together to create simple circuits and, ultimately, create a system that does something. The annual iGEM competition, where undergraduate teams cook up modified bacteria to do unusual things, shows what can be done even at this stage.

You can have bacteria that smell differently based on whether their cultures are growing or have run out of room. The MIT team that did that put the genes for wintergreen - for a minty smell - and the chemical that gives bananas their characteristic scent in the bacterial chromosome and wired them to built-in sensors. The result was a rudimentary computer that looked at the state of the bacteria and reported what it found as smell.

From this, it sounds as though science is well along the road to being able to design biosystems. The relationship between parts of the bacterial genome seem so well understood that people have started to build tools to assemble the designer sequences: CAD for the genome.

It turns out that the technology to generate genetic sequences is outpacing bioengineers' ability to define them. Patrick Cai of the Virginia Bioinformatics Institute said at the BioSysBio conference last week: “We have more or less followed Moore’s Law for DNA sequencing.”

The technology has reached the point where sequencing company Blue Heron Biotechnology was able to produce a series of DNA chains that were combined to form a genome of 580,000 nucleic acids (bases) for the ‘artificial’ bacterium developed by the J Craig Venter Institute. However, this genome is based on a naturally occurring sequence.

The problem is that the ability to design novel sequences has failed to keep pace. Cai pointed to a competition run by Blue Heron, which would have provided the winner with $250,000 of sequencing in the form of a 40,000 base-long string of DNA. The company, apparently, did not receive a single entry.

The Virginia group has come up with Genocad: a tool that checks artificial gene sequences to see if they obey the 'grammar' of bacterial genetics. Jean Peccoud, associate professor at the institute wants to go further and build a computer language for defining DNA: "We would like to have something like the C programming language or Visual Basic," he says. You would define switches, oscillators, and AND and or OR gates using functions that look like this:

switch (ligand x, ligand y, reporter g)

A compiler would then take all those statements and compile them into a DNA sequence that would, when inserted into a cell, start to perform all the functions you defined. This language, which Peccoud calls XDL, is a long way from reality: "What I showed is only one part of what such a language would look like."

It will probably look less like C than the electronics hardware descriptions in use today, as they handle concurrency, which C certainly doesn't in its native form. Peccoud says: "We are looking at Verilog and VHDL, as used in electronics. We are looking with people who have an EE background to see if they are suitable."

However, this is where the analogies between electronics and IT begin to break down. If you look at a circuit diagram, you have lots of things like AND gates scattered around. In biology, as it stands today, you only get to use one. If you want to have two AND gates in the same biosystem, you need to find another gene that performs the same job but in a different way.

The one thing that electronics has over biology is the ability to define connections between parts. In bacterial cells, at least, you don't get that. Everything happens in what is effectively a little bag of soup in which proteins and molecules only get together by bumping into each other. And just because they bump into each other, it doesn't mean anything will happen: all the processes are statistical. It turns out that electronics is getting a lot more statistical, but it's a convenient abstraction to think as the processes as being deterministic.

Now, a biolanguage compiler could deal with a lot of that, simply picking compatible parts from a big list, such as the MIT Registry of chunks of DNA. However, as you add more genes into a system, you start to introduce odd little dependencies which means, in most cases, the thing doesn't work. There seem to be ways around this. One is to simplify the bacterial DNA chassis to the bare minimum. Another is to focus on doing more design downstream of the gene - it seems that nature makes a lot of use of what are called protein scaffolds. With these, you bring a bunch of functions together in one protein or design the proteins in such a way that only when certain proteins stick together does something happen.

Some insight into this has come from a group that focuses not on turning bacteria into factories but trying to work out why evolution has produced certain things. This is what is happening at the University of California at San Francisco.

"It turns out that signalling proteins are highly structurally modular," said Caleb Bashor, who is based at UCSF, at a meeting on synthetic biology late last year in Cambridge. Some protein enzymes in natural cells sit on a scaffold. Simply replacing and rearranging elements on one scaffold can ‘reprogram’ the protein. "It suggests that the complexity of networks is due to the interactions [between elements on a scaffold].

"Scaffolds appear to provide a powerful platform for exploring the plasticity of [cell] pathway signal processing: a tool for understanding circuit-design rules. Using them, we could rewire cells with new or modified behaviours," Bashor claims.

The problem that faces the synthetic biology community is working out just how concepts such as modularity fit into the discipline. It may be that engineers can only design these systems with a lot of computer support - with statistical tools working out what the potential interactions are between genes and other parts of the cell. It will be very different to what most people think of engineering. But the concentration on statistics might have other spinoff benefits - the inability to deal with random, infrequent events is so often what brings IT to its knees.

Posted by Chris at 11:28 AM | Comments (1) | TrackBack

April 28, 2008

Tech's forgotten takeover attempts

Marc Andreessen has a good dissection of the strategies that Yahoo could deploy to try to fend off a possible hostile takeover by Microsoft. However, most of it is of the form, "I wouldn't start from here if I were you". If Microsoft presses ahead, it seems likely that Microhoo is not far away. And, given the state of the markets, Yahoo's board may not be alone in looking over their collective shoulders.

Andreesen contends that Oracle-Peoplesoft was the first major hostile takeover in the tech sector: the argument being that tech companies depend far too much on soft assets, which are wont to leave even if the takeover goes through.

It should really be successful bids A number of attempts took place about ten years ago and a couple of years after the one that surprised everyone but which ended with an agreed deal: IBM's bid for Lotus Development.

The takeover spree of the late 1990s was fairly shortlived and, of close to ten big attempts, most of the bids were withdrawn. There was a lull and along came two more, away from the software business. At the time, financial analysts believed that a feeding frenzy was on the way and boards of directors queued up to put poison pills in place. Analog Devices only recently cancelled its - maybe it will start to reconsider its move. And yes, ADI's poison pill entered the company's byelaws in 1998. In the event, the feeding frenzy never took place.

However, the reasons as to why hostile takeovers are soon to be in fashion are similar: Andreessen is right to predict more of the same happening. And we already have interventionist investors trying to unseat boards - that is, doing takeovers the cheap way. Obrem Capital is continuing with its attempt to oust Ray Zinn and co from Micrel Semiconductor.

In the late 1990s, the big deal was AlliedSignal's takeover of connector maker AMP for close to $10bn (in 1998's money) - that's plenty major enough as Oracle picked up Peoplesoft for just over $10bn in 2005. It was a bitter campaign, as many of these things are, with the companies suing each other over poison pills and the like.

Why 1998? Basically, although the Internet revolution was picking up steam, the electronics companies were suffering from the hangover of the Asian crisis. Software companies had never had it so good. The people making the bits that went inside the computers they had to buy were facing dramatic slides in share price: making them juicy targets for anyone with a bit of spare cash. However, the slide was comparatively shortlived as electronics companies managed to get themselves back on the Internet bandwagon, at least for as long as it lasted.

In the end, in 1999, AMP was 'rescued', for the sum of $11bn, by Tyco International - yes, that Tyco. It all happened before Tyco boss, Dennis Koslowski, who liked to buy things - oh, he really liked buying things - was ejected from the company and ultimately tried for larceny.

When ADI cancelled its poison pill, I criticised the naming of them as "shareholder rights plans" as a misleading euphemism. They are surely there for the management. But there is an interesting paper that looks at the payoffs for companies that adopt them that concludes that poison pills can lead to bigger payoffs for shareholders.

AlliedSignal's move on AMP was mostly about money. The attempt by Mentor Graphics to take over Quickturn Design Systems was about putting an end to a bitter legal patent dispute - a fight that would lead to the boss of another company winding up with a long prison sentence for, among other things, trying to arrange the murder of a judge. Mentor was none too fussed if Quickturn people left, the aim was simply to stop the litigation.

Like AMP, Quickturn found a rescuing company: Cadence Design Systems. The patent war didn't end until the autumn of 2003, when Cadence and Mentor signed a deal, with Mentor paying $18m. It was more than ten times less than it wanted to pay for Quickturn, but also after a lot of legal fees had been paid.

Posted by Chris at 6:40 PM | Comments (0) | TrackBack

April 27, 2008

Talkin' bout a revolution

Last week, two events in London showed how far apart the views can be on what, to some, is the beginning of the future of chemical engineering and what to others is simply the beginning of the end.

At the IET's BioSysBio conference, which kicked off last Sunday, Professor Richard Kitney of Imperial College, London, argued that synthetic biology is the engine of a third industrial revolution. He pointed to the discoveries of the mid-19th Century and how they drove the rise of the synthetic chemical industry.

The problem that conventional synthetic chemistry has is that it is a brute-force process. It excels at producing simple molecules in high volume. But complex chemicals, particularly those needed for drugs, are expensive to manufacture. And it is no good for producing fuels because you have to put more energy in than you will ever get out.

Cells are chemical factories in miniature that are very good at producing complex chemical structures. Unfortunately, nature has not seen fit to evolve a petrol-producing bacterium. Synthetic biology opens up a future when gene reprogramming will make it possible to develop a bacterium that can turn sunlight and excess carbon dioxide into petrol or ethanol.

However, it does not take long for the ethical issues to surface. Opponents to this kind of technology worry the world will end up covered in toxic green goo pumped out by bugs gone bad.

At BioSysBio, scientists were concerned that synthetic biology could go the way of genetically modified (GM) crops, with stifling restrictions put on research and exploitation of the technology.

A few days later, it was easy to see why. Prof Kitney was also one of the speakers at the Royal Society of Chemistry's Engineering Life seminar. Biosafety was not meant to be top of the agenda, but it ended up that way. The first person I met, just before it kicked off, was a health and safety specialist from one of the London colleges.

During the seminar, Professor John McCarthy of the University of Manchester's Interdisciplinary Biocentre, wanted to take questions on the science before going on to the ethics and safety issues. However, the Q&A bypassed the science entirely and went straight for the safeguards.

As Baron Patrick Jenkin of Roding put it from the audience side: "We have moved immediately into this area. And that is why I am here...I don't think the public has begun to hoist onboard the huge potential risks and benefits."

Lord Jenkin said he planned to table a question in the House of Lords on biosafety regulation: "Is the current system of regulation sufficient to cover synthetic biology? If the public are going to be satisfied they need to be confident that the system of regulation is keeping an eye on it."

Although the Parliamentary Office of Science and Technology (POST) put out a report on synthetic biology at the beginning of the year, the debate at political level has barely started in the UK. It's not exactly racing ahead in other countries. Researchers from the J Craig Venter Institute, which is pushing ahead with research into synthetic genomes, co-authored a report on governance which largely concentrated on bioterrorism threats.

"It was an interesting assumption that this seemed to cover all the social and ethical issues," said Brian Wynne, associate director of the ESRC Centre for Economic and Social Aspects of Genomics at the Engineering Life seminar. He wryly quoted from one of the papers published by the JCVI last year, which described the transplantation of a genome from one strain of bacteria into the cells of another. One line said: "The societal and ethical implications have been explored". It referred to a paper from Science written in 1999 and the Options for Governance.

Regulation may be the least of anyone's worries. Personally, I think the framework will not change all that much from what we have today, other than some additional safeguards that focus on the companies that produce DNA sequences based on recipes provided by researchers. That is not because of indolence but because most of the core technologies needed for synthetic biology were introduced in the early 1970s. The two changes since then have been the rise of DNA sequencing as an industry and an improved understanding of what genes do. It is worth remembering that, in the mid-1970s, scientists accepted tight rules over what could be done with the then novel technique of recombinant DNA which were then loosened slightly as it became apparent that some risks had been over-estimated.

The problem is the role of industry. If synthetic biology works - and there are few reasons to believe that it won't, even if the reality falls short of what the JCVI is promising - industrial exploitation is inevitable. But the antics of companies such as Monsanto did as much as anything to turn the public off even the idea of GM. A rush toward exploitation is likely to trigger a similar reaction - the antibodies are now in the system.

Posted by Chris at 10:23 PM | Comments (0) | TrackBack

April 25, 2008

"Words are there for themselves"

Today's Graph of Nonsense Award goes to the one reproduced by Erick Schonfeld at Techcrunch to try to explain the chocolately goodness that lies behind Radar Networks' plan to apply Semantic Web technologies to search. It's one of those classic startup-generated graphs that purport to show how existing, in their view clapped-out, technology is going to give way to the shiny new stuff.

The argument relayed to Schonfeld by Radar's Nova Spivack is that today's search engines, which are based on keyword searching, are running out of steam. Spivack's contention is that the volume of data is going to overwhelm keyword-based technologies soon and that what you need is to add meaning to the underlying text to help the poor old search engines out. And, in startup style, the argument runs that the established search engines cannot deal with a root-and-branch reworking of their algorithms.

For some bizarre reason, Spivack puts down tagging and natural language search as points on the way to semantic search, rather than having semantic search before natural language, which seems to be his argument. And then we get to "reasoning": presumably at the point where we hit the Singularity or something.

Before everyone gets excited about Google going the way of AltaVista, we should take a step back and have a look at what goes on with keyword searching and then listen to what one of the technique's creators had to say on the subject of the Semantic Web.

One striking aspect of search technology is the resilience of that old-hat keyword technology. The core of most search engines is an algorithm developed in the early 1970s, primarily by the late Karen Spärck Jones and Stephen Robertson, who is now at Microsoft Labs in Cambridge. At the time, people believed that computers would have to understand grammar. The breakthrough made by Spärck Jones and Robertson, among others, was that you didn't need to force the computer to process grammar. The world is just too complicated for a machine to handle more than simple phrases.

But statistical processing works amazingly well. The algorithm that resulted from this work is amazingly simple. You calculate the frequency of each word in a document but knock out all the words that are found in most documents. Documents only score highly for words that are found in a small subset of files. It's why names and specific terms work so well in locating documents when you are using Google.

Spivack contends: "Keyword search engines return haystacks, but what we really are looking for are the needles."

No, the TF-DF algorithm does indeed return needles, almost by definition, because you don't bother indexing most of the haystack. The failing of the technology is that you have to know what the needles are called. But there are ways around the problem that do not demand the introduction of an additional layer of data.

Spivack points to the problem of citation in driving the results provided by Google as a problem of keyword searching. It has very little to do with keyword search - that is Google PageRank at work, which is a technique that gets overlaid onto the other statistical mechanisms to provide the search engine with a bit more guidance as to which pages most people find useful. Search-engine researchers are sceptical as to how much even Google uses PageRank to rank pages these days. As spam and SEO techniques get more intrusive, I can see PageRank getting dumped entirely.

The important thing to bear in mind about today's search is not that it works using keywords: it's that it uses statistics, not any form of what you might call deterministic processing. The mechanisms have been augmented over the years into what the head of search specialist Autonomy, Dr Mike Lynch, calls "keyword plus".

Lynch contents that the Autonomy engine has gone beyond the TF-IDF algorithm of Spärck Jones and Robertson. But it still, very much, uses statistics. Autonomy uses Bayesian inference which dispenses with the concept of word frequency. Autonomy replaces it with an approach that uses the probability of finding certain words within a portion of the document. One of the problems with Bayesian inference is that the number of calculations you need to make is much higher than with keyword indexing if you are not careful. But there some tricks you can pull to reduce the workload.

The important thing is that every successful search engine in use today doesn't have to understand language by attempting to deconstruct sentences using grammar rules or trying to extract meaning: it just has to use statistics. Spärck Jones' final lecture, which was recorded about a month before her death from cancer at the age of 71, describes how successful statistical processing has been when dealing with language. In fact, even language researchers have found statistical models more successful than deterministic grammar models in recent years.

As Robertson put it to me when I interviewed him a week ago, there is plenty of scope for statistical processing in search and similar applications.

Statistical methods that evolve from the ones in use today probably have a lot brighter future than any kind of semantic searc