nVidia’s Co-founder and CEO recently hosted a dinner for selected press where he discussed the company past, present and future, revealing interesting bits about the future architectures as well. The dinner was covered on several publications, but we took note of one sentence published on VentureBeat:

"To take the ARM processor, partner with them to develop a next-generation 64-bit processor to extend it so that all of computing can have the benefits of that instruction set architecture. It is backward-compatible with today?s ARM processors."

Jen-Hsun Huang, co-founder and CEO of nVidia holds smallest motherboard featuring Tegra 2 System-on-Chip This was a clear confirmation what our sources were saying about Project Denver: that nVidia is bringing a 64-bit instruction set to the ARM CPU architecture. According to our sources, in order to fully develop a new high-performing CPU architecture, you need to dedicate roughly between 1.1-1.8 billion dollars. The figure comes from companies that actually develop CPU architectures, thus they have previous experience. The cost of Project Denver could be much higher, as nVidia doesn’t have previous experience in building a CPU architecture. The company however, has a lot of experience in building a complex GPU architecture, which are now (arguably) more complex than any other computer chip on the planet. GPUs in fact, cover first (nVidia GF100/GF110), second (AMD 6900 "Cayman") and third place (AMD 5800 "Cypress"). Intel’s ultra high-end CPU, Itanium "Tukwila" CPU comes in fourth place, with only 50 million transistors more than fifth-placed nVidia’s GF110/114, used in mainstream cards such as GeForce GTX 460 and 560 Titanium.

According to Jen-Hsun, nVidia is "working on a CPU internally for about three and half years or so. It takes about five years to build any full custom CPU. And Project Denver has a few hundred engineers working on it for this period of time and our strategy with Project Denver was to extend the reach of ARM beyond the mobile, the handheld computing space."

Thus, we don’t expect Project Denver to appear before late 2012 or early 2013 – in line with Maxwell GPU architecture, which is expected to integrate Project Denver architecture and become the first shipping GPU which could boot an operating system. It would not be the first GPU to boot an operating system, though. According to several PR representatives, the company already managed to boot an special build of Linux using Fermi GPU, but resources for that were abandoned as it proved too much of a hassle. Kepler of Maxwell for a public demonstration? Our money is on Maxwell?

Two major reasons why nVidia wanted to develop the processor – with or without Microsoft’s dedication to base Windows 8 on ARM architecture – were Tesla and console business. Tesla is a growing business unit still gives a lot of revenue to either Intel or AMD, whomever ends up being selected for CPUs that feed the Tesla GPGPU array (1 CPU feeds 4 GPGPUs). Pairing Tegra 3/4 with Kepler-based Tesla is a start, and Maxwell/Project Denver should be a viable business solution, abandoning the need for an x86 CPU in order to feed the parallel GPGPU arrays.

As far as console business is concerned, Jen-Hsun confirmed that the company is working on an next-generation console. Chances are, the only chips inside the unnamed console should come from Santa Clara. Sony or Microsoft? We’ll leave that guessing game to others.