SPEAKING TO JOURNALISTS after his keynote at Nvision 2008 today, Nvidia CEO Jen Hsun Huang answered questions about his passion for visual computing and his feelings about x86's.
Huang started off a bit evangelistically, claiming his main message was "on behalf of the visual computing ecosystem," adding that Nvidia would be working towards "solving some of the worlds most complex visual computing challenges."
Denying that Nvidia had put on Nvision because it was jealous of Intel's developer forum conference, Huang remarked that he wasn't "quite that unambitious," adding "IDF's not that big."
Puffing out his chest, Huang continued "IDF is really for the PC ecosystem, Nvision is really for the visual computing ecosystem. It's not about us, its about our ecosystem."
"We also wanted to define the industry, like Cisco defined the networking industry", noted Huang, who reckoned "Every important industry needs to have a conference and our industry doesn't have one."
Probably having picked up on the grumblings after his key note that there had been nothing new, Huang went on the defensive, claiming "we made it a point not to announce products here, this is not our platform to announce products, this is a platform for the ecosystem."
Of course, it wasn't long before the discussion got around to Larabee. "We don't know what Larrabee is because it hasn’t shipped yet, so we can’t really talk about it," said Huan, who boastfully (and some would say arrogantly) added "also by the time that Larrabee ships, Nvidia's technology will be so much more advanced."
According to the Nvidia chief, "The reason why Larrabee is so important to Intel is because GPU computing is really important. Our GPUs today are completely programmable, completely general purpose and they support C – the modern computing language."
Like a dog with a bone, Huang continued, "What is the number one benefit of x86?" Answering his own question, Huang said "x86 is all about binary compatibility. Nobody ever said that if I built a new computer architecture, the one I would create is x86 unless I wanted binary compatibility."
This, said Huang, led to the question "is Larrabee binary compatible with Windows? Is Larrabee binary compatible with x86 and 64-bit x86? Is Larrabee binary compatible with SSE 3, 4, 5, 6?." Again choosing to respond to his own question, Huang declared, "The answer is no."
So what is the point of an X86 according to Huang? "It has tools and so does every other CPU. ARM, PowerPC, Cell – they all have tools. The reason why these tools exist is because most of them are high-level language tools. Therefore, I think that part of is a bit of a distraction and a bit of a smokescreen."
But Huang didn't completely dismiss the x86. "We believe in heterogeneous computing. The CPU and GPU should work together and the CPU is x86, so I’m x86 as well," said Huang.
"Competition frames an industry" gushed Huang, who claimed Nvidia had invited both AMD and Intel to take part in the conference. "We wanted them to be here," he claimed, reckoning that changing the name from Nvision to something different in the future shouldn't be an issue. But, reckoned Huang, " Nvision isn't a bad name," adding that at least he hadn't named it the "Nvidia development conference."
Briefly touching on Nvidia's plans to make headway in the mobile market, Huang commented that Nvidia's "strategy with mobile is to, number one, completely focus on smart phones. We believe that the phone, the mobile device will become the next personal computer."
Huang also discussed Gforce and Cuda briefly, noting towards the end, "We have this carrier, it's called Gforce and we invest about a billion dollars a year on Gforce."
Huang pressed on, boasting, "Cuda on Geforce is so revolutionary. We could take Cuda into the world without a single person paying for Cuda. Only I pay for Cuda."
We reckon Huang cuda elaborated on the whole subject just a wee bit more though. Oh well, watch this space in case he changes his mind over the next two days.µ
Jen Hsun Huang is such a bloviating phoney! He should stop with his drooling utterances and learn how to start manufacturing decent, i.e., non-defective, parts.
He defends his company against storms and falling meteors.

Kind of reminds me of that guy in iraq during the first part of the war. He'd wave both arms open and say "We're deterring the invasor" now that's the spirit!
"Every important industry needs to have a conference and our industry doesn't have one."

Ummm.... SIGGRAPH?
"We believe in heterogeneous computing. The CPU and GPU should work together and the CPU is x86, so I’m x86 as well ..."

This seems to be change in tone from few months ago when NV's PR used to talk about CPU is dead and [GP]GPU would rule computing world. Looks like, NV is working on x86 CPU ( which they should have done years ago).
More Inq hate-on of Nvidia... You know it is possible to just give us the facts of the article without all the spin like, "he puffed on his chest." Seriously guys. Weak, weak, weak.
Huang's opinion on the technical stuff are actually quite insightful. Cut the guy some slack!

There're some dark clouds gathering on his visual computing *ecosystem* though. It ain't what the inquirer reporters think it is...

While the idea of larrabee excites me as does cuda and gpgpu I then come back down to earth and think....how many applications actually make use of the 4 cores of my quad core cpu....not many.....programmers seem to have difficulty it doing this....perhaps many applications only can be single threaded and as such will not benefit from the multi core parallel hardware explosion so whether its cuda or x86 larrabee using god knows how many cores many many and perhaps the majority of applications and tasks simply cannot be executed using a parallel approach.
While a certain amount of parallelism is necessary and of use in reality the focus needs to return to more efficient architecture and higher clock speeds.