Intel gives update on Larrabee discrete graphics: still dead

Status
Not open for further replies.

Jos

Posts: 3,073   +97

For the last three years or so Intel had been working on an x86-derived graphics chip, dubbed Larrabee, that would supposedly shake up the discrete GPU market. After holding several demos, which unfortunately were far from being competitive, plans to market a consumer version of Larrabee were indefinitely put on hold last December, opting instead to release it as a software development platform for both graphic and high performance computing.

Stories posted since then have hinted at a comeback, but in a posting on the company's Technology@Intel blog yesterday, director of global communications Bill Kircos confirmed that Larrabee as a discrete graphics product for consumers is indeed dead -- "at least in the short-term." (Yes, he couldn't resist adding a bit of ambiguity). Intel said it plans to derive a business from Larrabee, however, using its multi-core GPU technology in the high performance computing market to process parallel code more efficiently -- much like Nvidia does with its CUDA programming model.

As far as consumer graphics are concerned, the company will continue pushing its integrated solutions for desktops, laptops and low-power handheld devices such as smartphones. Intel believes HD video and mobile computing are the most important areas to focus on moving forward when it comes to its graphics business.

Permalink to story.

 
"Intel believes HD video and mobile computing are the most important areas to focus on moving forward when it comes to its graphics business. "

Well they're *****s. Do they know how much more money they would bring in if they focused on their graphics solutions competing with ati and nvidia? Obviously not. Maybe its a good thing as them keeping their focus on cpu's is most likely a good thing. Focusing on sandy bridge is probably for the best. And slightly off topic. AMD should better get off their ****ing asses and start being competitive. Intel is dominating right now. AMD's 6 core is competing with the I5. This market has gone really weird. So little competition. AMD is behind with cpus but years ago they were dominating and nvidia is falling behind but they were previously dominating.
 
"Intel believes HD video and mobile computing are the most important areas to focus on moving forward when it comes to its graphics business."
Well they're *****s.

Since Intel only have 43.5% of the graphics market, yes, maybe they need to push just a little harder. While most of that share is from IGP based solutions (i.e. performance on par with this...) Intel probably figure that the majority of PC owners probably don't game at mainstream/enthusiast levels- a view seemingly backed up by this.
Add in a whole new extremely large (and mostly untapped) burgeoning market in handhelds, tablets, netbooks, smartphones ebook readers, digital tv, connected picture frames and home gateways (i.e multiple processors per potential customer), where ARM licensees including Qualcomm's Snapdragon, nVidia's Tegra2 and TI's OMAP are poised to cash in at Intel's expense and I can see how Intel might want to channel R&D funding to those areas at the expense of high-end graphics.
There is also the not-insignificant stumbling block that designing a successful graphics architecture is damn hard work. A fact seemingly born out by the entire market dominated by just two players.
Do they know how much more money they would bring in if they focused on their graphics solutions competing with ati and nvidia? Obviously not.
Given that Intel posted a CLEAN PROFIT of $US 2.4 billion for the first three months of this year, I would say that 1. Intel has fairly good vision when it comes to accumulation of revenue, and 2. I'd be quite happy if they didn't accumulate too much more market share.
Focusing on sandy bridge is probably for the best.
Sandy Bridge is a done deal, Intel are busy at work on it's replacements (Ivy Bridge - a die shrink from 32nm to 22nm, and probably the series after that, Haswell)
AMD should better get off their ****ing asses and start being competitive.
Unfortunately that requires significant R&D resources-Something in short supply at AMD, a few eureka moments from the design people and a change in public perception from that of AMD as Intel's ***** stepbrother.
 
Status
Not open for further replies.
Back