sorted by: recent | see : popular

Content Tagged with semiconductors + nvidia

After more than two years of pushing its scientific computing efforts, Nvidia’s graphics processors will be offered as an option in the newest line of Cray desktop supercomputers. The chipmaker plans to announce next week that its Tesla chips can be used in the $25,000 Cray desktop supercomputer, according to Nvidia spokesperson Andrew Humber. He said Nvidia has been in talks with Cray ever since the chipmaker announced its Tesla line of graphics processors in 2007, but that this is the first deal the two companies have inked.

The Cray CX1 computer launched today, with specs that include either 32 or 64 Intel cores and 4 terabytes of internal storage. The new machine, which runs a new version of Microsoft Windows, is a testament to both the demand for and the democratization of computing power. Indeed, people who earlier might have turned to grids or supercomputers for their problems are building powerful desktops with accelerator chips, while less scientifically minded folks, such as traders or product designers, who want to render things in 3-D are seeking more processing power.

Cray’s CX1 is the smallest supercomputer the venerable firm has ever built, but its downmarket shift is a response to both the needs of the market and the presence of accelerator chips trying to muscle in on its scientific computing turf in the high and low end. Chips such as IBM’s cell processor or GPUs from AMD or Nvidia are being dolled up with programming tools to take on scientific computing. The multiple cores in the Cell chip and GPUs are designed to parallelize tasks and execute them faster than a general purpose CPU, like the x86 processors offered by Intel.

At the desktop level, Nvidia has been touting stories such as the €4,000 (about $5,700 today) “supercomputer” built by scientists at the University of Antwerp creating 3-D images of internal organs that uses GPUs. With the CX1, Cray is acknowledging that trend and trying to ride it.

The effort to broaden its market comes as Cray sees it dominance in the supercomputing world waning. The top supercomputer in the world runs on a combined x86 and Cell processors. In the most recent list of the Top 500 supercomputers, Cray only made 16 of the machines for a 3.2 percent share of the fastest computers in the world. That’s quite a decline from when the Top 500 organization started tracking the data 15 years ago and Cray made 205 systems on the list. So Cray is thinking small to expand its market as the market demands more computing power.

Last week, Intel offered up a sneak peak of its Larrabee graphics processor, due out in 2009 or 2010 and guaranteed to raise the competitive pressure on graphics chip makers Nvidia and AMD. Unlike its existing integrated graphics chips, Larrabee will be a standalone processor, but don’t expect that it will be a success.

As computing has required faster chips, Intel and other chip makers have added more cores, a tactic that GPU makers have used for years in order to increase parallel processing. GPUs from Nvidia contain as many as 240 cores while those from AMD, which that company acquired when it purchased ATI Technologies in 2006, have hundreds. So they’re faster.

But they’re also harder to program, something Nvidia is trying to solve with more flexible chips and a new programming tool called CUDA. But most enterprise and consumer software runs on x86 chips and needs adaptations to take advantage of GPUs. Intel’s Larrabee chip has multiple cores, but is not a GPU. Intel claims this offers people the performance gains and ability to render graphics much like a GPU does, but that Larrabee’s x86 architecture allows for easier programming.

It’s nearly impossible to judge a chip until you’ve seen it in action and tracked whether OEMs want to put it in their devices, but my bet is that Intel won’t be able to compromise with a many-cored CPU and believe it will beat a GPU at its own game. Nvidia and AMD are hoping as much, especially Nvidia, which has the dominant GPU market share — right behind Intel’s integrated chips — and wages an almost constant battle again Intel’s PR on this front. As Nvidia’s small but fierce marketing team faces off against Intel’s Goliath, grab some popcorn, because it’ll be a graphics showdown worth watching.

Nvidia has plans for a mobile chipset that will change the look and functionality of smartphones when it hits in mid-to-late 2009. While many of the big chip vendors are placing bets on the concept of a mobile Internet device that’s larger than a smartphone, but smaller than a laptop, Nvidia’s APX 2500 chips could enable devices that are so sexy, they might render the need for an MID obsolete.

However, I’m told the company will announce an expansion of the APX chips into MIDs soon, so I could be wrong on that last point. Nvidia launched the chips that will make a smartphone function like a PC (or an iPhone) at the Mobile World Congress in February, and I can’t believe I missed it.

This is Nvidia’s first move into making the “brains” of a mobile device, and it’s using its graphics expertise to turn the devices containing the chips into portable media players that can play 10 hours of HD video (on an external screen) and 100 hours of MP3s on a single charge. All while the 750 MHz processor consumes less than a watt of power.

In a demo at Nvidia headquarters two weeks ago, I saw a device slightly larger than an iPhone power an HD rendering of a Pixar short called “For the Birds” on a big-screen TV. It was connected via an HDMI cable and it looked good at 720p. I get that some people don’t mind watching movies or TV on their cell phone or iPod screens, but if I’m able to download that content and plug it into a TV, that’s an entirely new ballgame for travel and sharing. I want that device.

The demo I saw was powered by Nvidia’s chipset running on Windows Mobile, creating a chip/OS combo that mimics some of the visual pizazz of the iPhone, but on a more business-friendly operating system. Sure, as far as mobile operating systems go, Windows Mobile isn’t exactly tearing it up, but the integration of business and pleasure could make the current angst of choosing between a BlackBerry or an iPhone a thing of the past.

The chipset will first appear at the end of this year in personal navigation devices and personal media players, with a smartphone due out in the middle of 2009. Unfortunately, the APX 2500 contains an HSDPA RF chip, so it won’t be deployed on my network, but TMobile subscribers should keep their eyes open. Like the iPhone, the APX is modem agnostic, which means it’s not tied to any particular cellular network. There’s plenty of room for Nvidia to stumble, since it doesn’t have the experience designing for the mobile space, but I’m hoping it can succeed right about the time my current mobile contract is up.