First they began with a simple (for them) demonstration of serial operations - using a paintball gun wielding robot to draw a smiley face on a whiteboard. A hundred or so blue dots made the robot one of the slowest (and loudest) dot matrix printers we’ve seen.

Parallel operations would take something a little larger, and their 1100 paintball inkjet printer filled much of the stage. Powered up it would create a picture of the Mona Lisa in glorious 8-bit colour in a fraction of second. Huge air tanks held the compressed air the device needed to simultaneously launch all the paintballs in all the tubes.

The demonstration was certainly impressive, but it was more than a little misleading.

The type of data-centric work that CUDA GPUs handle is more about using parallel processes to handle lots of small pieces of data, not about building complex images from small pieces of data. With a parallel architecture like that you develop algorithms that break down big problems and big data sets into smaller, easier to work with, pieces. Farmed out across tens and hundreds of processors in a GPU, each data block can be processed, before being reassembled and the results delivered.

They’re not new techniques, either, for one thing the approach is at the heart of computational fluid dynamics and finite element analysis. The parallel techniques used in GPU computing are certainly impressive, and are already delivering supercomputing to the desktops of the scientists and engineers who need the power (an Nvision session on using GPU-based supercomputers to model the plasma dynamics around neutron stars and the black hole at the centre of the galaxy was particularly impressive). Low-cost high-performance computing is the GPU’s strength, especially when compared to the hefty power requirements of an equicalent array of traditional CPUs.

The Mythbusters’ demonstration was good (and an enjoyable piece of theatre), but it really told a different story. So how could the intrepid special effects team have told the real story of GPU computing?

How about one robot carrying a large, heavy cube across the stage? Suddenly it’s joined and over-taken by a swarm of smaller machines, all carrying smaller cubes - cubes that weigh as much as the single cube on the struggling robot. Or if paint is the preferred metaphor, a can of paint slowly emptying through a single pipe. Meanwhile another can empties through hundreds of holes in much less time.

So, how would you demonstrate it?

–S

Games may have made Nvidia, but it’s the rest of the graphics industry that keeps it going. Simulation and CAD drive much of today’s industrial design, while complex financial calculations can be run on GPU-powered parallel processors. It’s not just black hole plasma dynamics - it’s also the models that help calculate how a fusion reactor will operate. According to Nvidia GPU computing is bringing supercomputing to the desks of the people who need it the most - for just the cost of a video card.

One of the keynotes showcased a NASCAR simulator used by drivers to hone their skills. On stage we heard a populist story of what it was like to be a driver, and what it was like to use simulation tools. Off stage we heard a more interesting story about how the simulator developers were looking at using the latest generation of GPUs in their application. The ability to use a GPU for parallel processing - and the availability of powerful hardware physiscs engines - has made them completely rethink their next generation, as the new hardware features mean that they can now work on making the simulation more realistic.

That’s what the drivers want. Asked what he really wanted from a simulator, Kyle Busch didn’t talk about new high-resolution graphics or realtime ray tracing. What he wanted was more accurate physical behaviours. In the real world passing on the left is different from passing on the right, while slipstreaming another car can change the performance dramatically. A simulation may look real, but without the physics it’s not realistic at all.

One plan for the next generation is to move away from the current car model, with only 6-degrees of freedom. Instead, it really needs 72 degrees, for all the hinge and flex points - all of which are changing dynamically. That’s where parallel processing comes in, as it allows a car to be modelled in real time, taking advantage of physics engines to turn those model calculations into real world behaviours. Improving the simulation will mean more (and happier) customers - as well as a continually improving model that can be shared with vehicle manufacturers.

It’s an approach that requires specialist processing that goes beyond the traditional CPU. Don’t confuse it with the death of the CPU, though. There will always be a place for the traditional CPU - it’s just that silicon technology has become ubiquitous enough for specialist hardware to offload processor intensive functions.

Need to encrypt something? Just use the hardware cryptosystem built into a TPM. Need to do thread intensive Java? Hook up an Azul network processing appliance. Need to do complex vector calculations on large amounts of data? Use a GPU. Nvidia’s CEO Jen-Hsen Huang talks about it as heterogenous computing, where the CPU handles tasks, and more specialised hardware handle the complex tasks that tax general purpose silicon.

Intel and AMD may still say that general purpose processors are just what the world needs - but they’re still investing in HyperTransport and QuickPath, the fast buses that specialised silicon needs. I wonder why they’re doing that, if specialised silicon is the dead end they say it is. Is there something about Moore’s Law they’re not telling us?

The reason you get free storage on Gmail and SkyDrive and Mozy and flickr and all the other Web 2.0 services isn’t just to keep up with all the other Web 2.0 services. It isn’t just to draw in visitors who can see and click ads. It’s because when you use hard drives, especially cheap, consumer grade hard drives, to run your search engine on, the only way to make that storage fast enough is to leave most of it empty.

The hard drives that Google stacks together are never more than 25% full, because any more than that and the latency to get the information back off is just too slow to make the search effective and slap the ads on it. The other 75% is just sitting there spinning around on the platter and making you happy by putting your files on there for free makes sense.

You don’t need access often enough to slow down Google’s own accesses and Google can queue your request up to retrieve when it’s convenient - your Internet connection isn’t fast enough for you to notice.

SSD is still expensive, but it could save a lot of money for enterprises and Web services because it’s lower power. For a notebook PC you care about 15% longer battery life - if Intel’s figures carry over to real PCs; for a data centre, you care about Watts/IOPS and Intel is claiming SSD can offer six times the read performance using 98% less power and needing only 75% of the space - because you can put cooler drives closer together without space for fans and AC. That means SSD should quickly start to become common. But that could also mean the end of ever-increasing free online storage.

On the one hand, SSD is fast enough that latency is much less of a problem, so you don’t have to leave most of it empty. And on the other, it’s expensive enough that you don’t want to leave any of it empty. 
-Mary

That’s the nickname for the way Intel tests how much of a difference SSD makes to user experience. It’s not just about how much extra battery life, although I’d like the expected 14 hours batter life I could get from the HP 2730p, the next version of my tablet, with SSD and the thin slab battery I already get 8 or 9 hours from.

The improved performance isn’t just for looking good in benchmarks or running video editing apps most people don’t use, it’s for stopping you sitting at the screen hollering “what are you doing!” as the hard drive light flashes on and off and Outlook sits there staring blankly. Not that it’s always Outlook; Acrobat is pretty good at sitting on its thumb, as are plenty of other applications. And as notebooks get smaller and lighter and 5400rpm is seen as something to aspire to, you can be left waiting far too often.

According to Intel’s cheekily named and possibly unscientific internal benchmark, you’ll be gnashing your teeth ten times less with an Intel SSD than a hard drive. They worked this out by asking a group of Intel employees to mark on a log sheet how often they got fed up enough with their computer to remember that they were keeping score. After two weeks they swapped them over to SSDs. And then after another two weeks, they made them go back to hard drives instead, sticking to show their frustration.

That frustration - and the tick marks - went down significantly Intel’s Principle Enginner for NAND Stephen Wells told me. “Not to zero; I’d still get annoyed if Windows blue-screened or something,” he said. But ten times less frustration was very noticeable. “And oh, the moaning and whining you got when we made people go back to the hard drive. I know - I was one of them. Do you want to get rid of your mouse? No.  Do you want to go back to DOS? No. In a few years will you want to get rid of your SSD? Absolutely not.”

 Not only is flash faster than hard drive, it’s more consistent. The 34 seconds it took to run through the photo and video tasks in one of Intel’s benchmarks always came out somewhere between 30 and 35 seconds, no matter how often the team ran it. But with the 5400rpm hard drive, Intel’s Chris Saleski told me the day before, the results were anywhere from one and a half minutes to two and a half minutes. 

Wells puts that down to the fact that data can be scattered anywhere around the disk and there’s an unknown latency in getting to it and getting it back that you don’t see with SSD - and he expects that to mean a more deterministic battery life with SSD as well. That way, when Windows says you have an hour of battery life left, you won’t find the machine hibernating to save your data fifteen minutes later. And that’s another thing I’d be ticking the frustration mark for…
-Mary
 

I’ve been swapping between the Zinc II, the TyTN II and my trusty HTC Excalibur (better known as the T-Mobile Dash) in an effort to fulfill my new year’s resolution about always having navigation with me. Google Maps does very well at location on some phones, but on my Excalibur it’s far from accurate so I’m looking for GPS. And while EDGE is OK for quick searches, I want 3G - mainly so I can use the phone as a modem with my laptop. I know built-in 3G is always better, to the tune of 25% better bandwidth, but not every laptop I use has it. I want Windows Mobile 6.1 for two things; threaded text messages and being able to search my email on Exchange Server from the phone inbox. I have a US Samsung BlackJack II which might be ideal - it’s the closest to the size of the Excalibur so far - but it’s very thoroughly locked to its US carrier.

The TyTN II is a great phone - and as the Stella from O2 it comes with CoPilot, which is my favourite GPS tool - but it’s just a bit too big and slab-like for me personally. Plus the tilt action is great for viewing the screen, but it covers the two action buttons on the keyboard. The Zinc II is a little bit lighter, a little bit smaller and a lot sleeker, with a soft-touch easy-grip rubberized coating and a flush screen - it’s a very comfortable handful even for those of us with smaller hands. It also has a faster processor, which means the camera doesn’t make you wait an age to take your snap and it doesn’t get bogged down with lots of apps running in the background.

That’s handy as with the TouchFLO-style Quick Menu launching from what I expected to be the Start button, I found it easier to launch a new app than get back to the one I’d been using. Swiping your finger across the screen to turn between the pages of buttons and tapping to open apps is a good way to work in Windows Mobile Professional; my nails work pretty well instead of a stylus but menus are still pretty tiny. 

The keyboard isn’t going to suit everyone. The keys are almost flush and don’t click down very far, but they have enough action so you know you’ve actually hit the key and not having discrete keys means - practically - that you won’t get dirt, dust and sesame seeds creeping under them. I’m used to the square layout of the Excalibur (and every BlackBerry I’ve ever known) and having the wide rows of keys slows me down until I adjust, plus the central spacebar isn’t quite in the right place for me. As always, secondary keys are distributed around the keyboard seemingly at random so you’re hunting for the dash and the @ symbol; it really is time we had a standard for this. But each key is outlined in blue light which is one, rather cool and two, really helpful in dim light.

MWg used to be O2 Asia’s device arm; they’ve expanded out to the US and Europe, renamed the company as the Mobile World Group and teamed up with gadget specialists Expansys. You’re not going to see the Zinc II on the high street unless they get another distribution deal, but it’s well worth checking out.
-Mary

I looked at her in surpise. “What do you mean? It’s virtually silent.”

“So what are all those fans?”

“That’s my desktop…”

That was when I realised it was time to change the machine I used every day. Bought over five years ago, it was starting to struggle with the processing and graphics requirements of today’s desktop applications. I’d got used to the roar of the fans - but throwing more and more cooling at yesterday’s technology really wasn’t the answer. After all, it would just make the office noisier!

A little web research,and I’d found that my usual hardware component supplier was selling very nice looking PCs - with most of what I needed. I decided to be as future proof as possible, and ordered a quad-core Intel box, with plenty of USB ports, 750GB of hard disk, and 4GB of RAM. I picked up a hefty graphics card as well, all for a third of what I’d spent five years ago.

Setting up the machine was easy enough. It had come with XP Home, but I blew that away and went with a Vista Ultimate install. It wasn’t very long before I had the new box online, and hooked up to our office domain. All-in-all it was relatively painless, though I still miss the option of having an extended desktop rather than the traditional dual monitor approach.

It took me a couple of days to install all the applications I needed - with a couople of caveats. It’s important to make sure that you deactivate applications like Adobe’s CS3 or Apple’s iTunes (and that you’re careful to make sure you import all iPhone applications before doing a first sync on a new PC).

So what are my key applications? I keep a list in OneNote, so I don’t forget anything - and here a few key applications:

• Microsoft Office 2007 - I live in Outlook, OneNote and Word
• Visual Studio 2008 - My usual development tool
• Firefox 3 - What else for the web?
• Xobni - Simplifying my inbox and my correspondence
• Clipmate - Managing the Windows clipboard
• Paint.Net - Image editing for free
• Cardscan - I get a lot of business cards, and this gets them into theOutlook address book easily
• Avast! AV - One of the best free AV tools around, and my recommendation
• Adobe CS3 Web - Web design and image manipulation
• Alzip - A good, fast, free archive management tool
• Filezilla - The best free FTP tool around
• Multiplicity Pro - Controlling my laptop from my desktop keyboard and mouse
• Feed Demon - RSS reader
• Aptana Studio - A powerful (and free) JavaScript and AJAX development tool

Of course there’s more - there are clients for social media networks, and tools to manage files between desktop and server.

My files moved across quickly, and I’ve been using the new machine since Monday - and I turned the old desktop off at the end of Monday, and it’s not been on since. Four cores and a 512MB NVidia 9600GT are an ideal Vista platform, and the OS is running smoothly - and extremely fast.

One thing I’ve done, to make sure I use one of Vista’s best features, is turn off the Quick Launch tool bar. It’s making me use the search word wheel on the start menu a lot more - and that’s good.

The office? A lot quieter. I can now hear the fans on the NAS across the room.

–Simon

If you actually use any Java applets, you may also get an error telling you there are several Java Virtual Machines running. 

It’s bad enough that Sun has, for at least the second time, put out software without a digital signature proving where it comes from, the most basic security check on code for the end user. It’s equally annoying that the suggestion from Sun is that you just click ‘Allow’ every time until the bug gets fixed in Java 6 Update 10 (’officially released later this summer’) and that Internet Explorer doesn’t let me say ‘Don’t ever allow this to run’.

But how about an update agent that runs every time you run your browser? That’s not very respectful of my resources, or my bandwidth. Other applications have periodic checks for updates and they only run when I’m not buys doing other things (Vista has an API for this, so even if you have umpteen different notification systems running, they can all find out when you’ve stopped to think or turned away to pick up the phone and do their updates, checks and maintenance without slowing you down). Why does Java need to check for updates so obsessively?

The Java control panel doesn’t think it needs to check that often; the default setting appears to be check monthly. So why does it hook into Internet Explorer to run the update agent all the time? Personally, I’m turning off the updater altogether, although that’s not a decision you’ll want all your users to take.

I can’t tell you exactly where the Java control panel hides itself; I couldn’t find anything in the All Programs list so I typed ‘Java’ into the search bar on the Vista Start menu and it offered me the Java control panel without having to dig for it. On the Update tab clear the check box for ‘Check for updates automatically’ and stick to your decision when Java asks if you won’t reconsider and click ‘Check Monthly’ instead because that’s the setting you started with. You may have to quit and restart Internet Explorer to prise Java’s hook out of the code and then you can go back to having browser windows open without a security warning that you train yourself to ignore.

That’s the problem with dialog boxes where it’s OK to just click yes, and one of the interface issues with Vista’s User Account Control. Any time there’s a dialog that’s in your way, the temptation is to click Yes just to get rid of it. Ask users if they want to do this unsafe thing, if they really want to, if they really really really want to and they’ll click Yes with less and less hesitation. Years of popups and confirmation dialogs have trained the user like a monkey in an experiment; click here and get what you want.

But you have to have confirmation for some things (Format C:? Record Battlefield Earth? Delete your wedding photos? Install an application just because you clicked on a URL?).

The real problem is that the PC has no idea of context or common sense; I navigated to the home page for the Kevtris game by typing in the URL, so when I click the download link and then click Run I really do want to install the game, but if I clicked an ad link in my email and it goes straight to installing a Trojan I really don’t want to. The PC has to leave intelligent decisions up to the user, and that means dialog boxes and confirmations when there’s anything that could be suspicious. Not remembering to sign the code for your application? That’s either suspicious, downright penny-pinching ($25 for a certificate) or shows you don’t have a good sign-off process for your developers. Either way, yes, I do want my browser to warn me about you.
-Mary

It turns out it isn’t - and that the built-in tools are effective spam blockers.

If you’re not using Exchange 2007 Content Filter (or Exchange 2007’s Intelligent Message Filter) turn them on. This is one of the most effective weapons in your arsenal. It’s regularly updated, and it scans messages for common spam formats. Mesages are categorised and given spam ratings, which you can use to reject, quarantine, or file messages in users’ Junk Mail folders. CF is surprisingly easy to use - set it up, set the basic filtering rules, and then occasionaly check your quarantine mail box for false positives.

Exchange 2007 has even added whitelisting for persistently filtered false positives. Once a domain is whitelisted, there’s no more delving in the spam folders for Twitter invites or press releases from Kaspersky and Sophos.

I’d been running my server like that for some time, when I discovered another trick that turned out to make a huge difference. Exchange actually supports using real-time block lists (RBLs), which are lists of spam IP addresses hosted by services like SpamCop and Spamhaus. It’s trivially easy to add new block lists to Exchange - just find the lookup address on the block list site (Spamhaus’ is zen.spamhaus.org), and add it and the provider name in the Block List Provider section of Exchange’s anti-spam tools.

Without RBL support turned on I was getting 500 or so spam messages in my quarantine a day, making it hard to filter out the few false positives. With it on, I’m down to less than 100. Managing my spam is a lot easier - and with whitelisting, I’m having to look in the spam folder a lot less often…

–Simon

I recently put the largest machine we’ve ever had onto our office network. A 64-bit server, with 6GB of RAM and 1TB of disk, it’s taken on the role of handling all our mail and files. When a new desktop PC arrives later this week, there’ll be over 7TB of storage on this small office network.

Just a couple of years ago we were surprised if we had more than 500GB of storage in an office. Turning back the clock still further, I helped design a UK-scale photo storage service, where we had a hierarchical storage system with a whole 3TB of spinning disk and 30TB of fast tape. Today’s fast external drives are making that architecture obsolete - our new server is using eSATA to do a whole server backup onto a 500GB drive. That back up? It only takes 30 minutes…

Outside out office the usual run of press releases seem focused on delivering larger and larger numbers. Cuil’s leaked launch (and the claimed size of its index) led Google to claim that it had indexed over 1 trillion web pages. That’s a pretty big number - 10 to the power of 12. It’s also the approximate number of bacteria living on the human body.

Closer to home, BT is claiming that it’s hooked up just under 17 million homes with broadband connections. It turns out that there aren’t many more homes to be connected, with broadband analysts Point Topic suggesting there are only around a million households that can migrate to broadband left - and around 9.6 million that don’t have internet access at all. The days of massive growth in broadband are behind us now, and what was a luxury is rapidly becoming a commodity. It would be interesting to see the spreadhseets at BT, as the company juggles the numbers to see how it can make money from running the data pipes.

After all, there’s plenty of scope for bringing the world online. Gartner recently suggested that there were over a billion PCs in use around the world (and soon there’ll be a billion transistors in each processor, thanks to Intel). While getting to a billion PCs in 30 years may seem a lot, there’ll be another billion in just 6 years, thanks to 12% annual growth. There
’s a lot of scope for significant social change here, as the emerging world (and especially the BRIC nations) start coming on line. The anglophile web will become just a part of a global, multi-lingual web - after all, even without the iPhone, China Mobile subscribers use more mobile data than any other network.

With all those machines, and all that information out there, there’s an issue of managing the information - and manging the storage it requires. The BBC has just such a problem.

The archive is currently managing about 700,000 digital items, with most of it still on discrete media (digital video tape, CDs, DVDs). There are about 280,000 actual master files, digitised from U-Matic video and 1/4″ audio originals, and from magnetic sound tracks. Then there are 60,000 viewing-quality video files, but these are held on CD-ROM in anticipation of a mass storage system. Overall they’re managing 12 petabytes, mainly on digital videotape - with a growth rate of about 400 terabytes a year, mainly on digital videotape.

If the numbers in your office are getting to big, be glad you’re not dealing with any of the really big numbers out there!

I noticed the other day that the market didn’t take well to the news that BT is really moving forward on plans to roll out fibre across to the UK to drag broadband speeds into the 21st century (think 8Mbps DSL is fast? - check out Korea, or Paris where they’re laying 30Mbps fibre). Cable coverage in the UK is a joke (NTL bought the cheapest demographic data it could find for high population density and ended up cabling multiple occupancy council estates where it couldn’t get licenses to offer a service and running out of money before it got round all the consumers and small businesses that actually wanted cable modems).

Now the analysts at Point Topic have done some interesting sums. BT’s proposal to cover 40% of the homes in the UK for £1.5bn works out at £150 per household - a lot less than the £800 each in previous calculations for doing all 25 million households. And making that pay dividends to all those worried shareholders will only take about £3 per household, according to Point Topic, because BT will be making savings on operating costs. Fibre means new services to sell; we might finally be able to get seamless roaming between landline calls, mobile calls and VOIP - it’s all IP underneath, after all. Some of the bandwidth will doubtless get eaten up by pay-for IP TV services.

And the regulator will need to keep an eye on who you can buy fibre from or we’ll be back to a monopoly faster than you can tell Sid pirated content isn’t the only reason anyone wants a fast connection (when did you last use an MSDN CD instead of downloading the ISO?). The industry has been asking OFCOM to promise it will be able to make money out of fibre as if it was something new and different. There may more trenches to dig in remote areas - although you can blow fibre down an existing conduit with compressed air - and you have to get the termination right, but it’s not rocket science. As Tim Johnson at Point Topic puts it, “by and large BT’s shareholders should be able to finance the investment, carry the risk and reap a good profit in return.”

Bandwidth; it’s a business, not a right, but it should be good business all round.
-Mary