NVIDIA Tesla Could Integrate 64-bit ARM Cores
 

Following last week's announcement of the ARM Cortex-A50 series, SoC designers formed a bee-line to license the ARMv8 IP. Among them is NVIDIA, and not just for its Tegra line of high-performance mobile chips, but also for Tesla, its GPU compute accelerator line. In an interview with InfoWorld, NVIDIA CTO for Tesla, Steve Scott, was quoted saying, "Tegra is going to become GPU computing capable in the not-so-distant future. Sometime this decade we are also going to start bringing integrated CPUs and GPUs together in the Tesla line". NVIDIA perhaps is looking at building Tesla GPU compute servers with the x86 CPU completely replaced with efficient ARMv8 64-bit processors.

http://www.techpowerup.com/img/12-11-02/11a_thm.jpg

Source: Tom's Hardware

So where does that leave project Denver/Boulder, if I may ask?

Denver is for tablets and mobile phones. Tesla is for high performance computing.

Tesla + ARM64 core is a good move. Tesla cores are good at parallel floating point tasks but sucks at serial tasks. ARM cores can help tesla here.

May be in the future, a gaming version of this card can play a game without using any CPU resources.

That make sense

This certainly is interesting. Is everyone going to start integrating arm cores now?

Maxwell (The architecture following keplar) will have some thing like this aswell.

Boulder is supposedly a "high-end" (meaning non-mobile, my guess) server/desktop part to replace x86 CPUs in servers and small form factor PC. It didn't look like they would be integrated in the GPU.

Yep. Boulder is the big brother of Denver; with much higher performance at the cost of increased power consumption.

We can expect both projects to move to Cortex-A50 in 2014.

Suddenly the bandwidth of PCIe 3.0 has a market.

It always had -> RAID controllers

now everybody will have arms soon

everyone already had arms, their just putting them to better/some use:p