NVIDIA’s David Kirk Interview on CUDA, CPUs and GPUs

David Kirk, Nvidia’s Chief Scientist, interviewed by the guys at bit-tech.net.

Read the full interview HERE.

Here are some snippets of this 8-page interview:

– page1
“Kirk’s role within Nvidia sounds many times simpler than it actually is: he oversees technology progression and he is responsible for creating the next generation of graphics. He’s not just working on tomorrow’s technology, but he’s also working on what’s coming out the day after that, too.”
“I think that if you look at any kind of computational problem that has a lot of parallelism and a lot of data, the GPU is an architecture that is better suited than that. It’s possible that you could make the CPUs more GPU-like, but then you run the risk of them being less good at what they’re good at now”
“The reason for that is because GPUs and CPUs are very different. If you built a hybrid of the two, it would do both kinds of tasks poorly instead of doing both well,”

– page 2:
“Nvidia has talked about hardware support for double precision in the past—especially when Tesla launched—but there are no shipping GPUs supporting it in hardware yet”
“our next products will support double precision.”
“David talked about expanding CUDA other hardware vendors, and the fact that this is going to require them to implement support for C.”
“current ATI hardware cannot run C code, so the question is: has Nvidia talked with competitors (like ATI) about running C on their hardware?”

– page 3:
“It amazes me that people adopted Cell [the pseudo eight-core processor used in the PS3] because they needed to run things several times faster. GPUs are hundreds of times faster so really if the argument was right then, it’s really right now.”

– page 4:
“Intel’s threaded programming model known as Ct.”
“several people have said that CUDA needs to run on the CPU as well – we can do that and we’re adding it”

– page 5:
“Every CUDA program will run everywhere.”
“The professional OpenGL applications typically can run on GeForce, but just really slowly. This is because OpenGL has good software emulation in the driver for any hardware feature that doesn’t exist – although some applications can detect that and refuse to run.”

– page 6:
“Without being too negative, we see Larrabee as the GPU that a CPU designer would build, not the GPU you’d build if you were a GPU designer.”
“AMD has been declining because it hasn’t built a competitive graphics architecture for almost two years now—ever since the AMD/ATI merger.”
“The third one is the next generation GPU technology—we’re going to invest one billion dollars in here this year and they need to invest on the same level to keep up with us.”

– page 7:
“I think rasterisation will definitely be around in ten years time.”
“I think that the APIs will evolve to include more ray traced components. There’s plenty more for us to do [in graphics] and we’re only scratching the surface at the moment”

– page 8:
“What Intel does well is not just building good CPUs, it’s the fact that they’re a vertically integrated manufacturing company – Intel’s in the business of turning sand into money.”
“but really if you took away the process advantage that Intel had, they would be behind AMD pretty consistently in terms of architecture. They win because they build an efficient and faster process.”
“I think that if we were to compete with Intel, we would need to become a vertically integrated company and have our own fabs.”
“GPUs are much bigger than CPUs, so in the same capacity that they could build 100 million CPUs, they could build 20 million GPUs.”