Nvidia today has announced two big things that will be pivotal to the company?s future both technologically and financially. In a way, today?s announcement is going to reshape the ARM ecosystem as well, since it is more than obvious that two blocks are being established: AMD and Intel are pushing the x86 architecture and now merging CPU silicon with the GPU and NVIDIA, Qualcomm, Samsung and Texas Instruments on the forefront of the ARM ecosystem.
The announcement of the first Tegra 2 based smartphone, the LG Optimus 2X, ushered in a new phase in Nvidia?s Tegra business. There is no secret that even though technologically advanced, Nvidia?s Tegra business has been struggling to effectively place their products with past generations. The problem was the choice of software/OS paired with it from those manufacturers. As a result, Nvidia?s initial Tegra chipset only saw implementations in the failed Microsoft Kin line of phones as well as their marginally successful Zune HDs that failed to crack the iPod market.
The LG Optimus 2X, as well as innovative Asus tablets, and bagging Audi, BMW and Tesla – represents the second coming of Tegra, this time with a vengeance. The fact that nVidia pushed Intel from the BMW Group is an shock to the ego of the blue company from Santa Clara, and one from which Intel will need time to recover. We all knew the hardware performance of Tegra and how the part makes other vendors look obsolete, but the proof is in the pudding, as Brits would say.
Optimus 2X is that ?proof in the pudding?, a device that enables all of the features that Nvidia had for years in the Tegra silicon. With the LG Optimus 2X, it looks like Nvidia has finally delivered on their promise of a ?Super Phone?. The only thing we have to see now is which carriers will pick it up and how well will LG promote this phone as the future of mobile computing. After all, the ability to play cross platform games on a phone that is console quality is definitely something many games await. The golden term ?console quality? is being constantly used by companies like Nvidia and Qualcomm to describe what they see to be one of the most important factors in a mobile platform because they firmly believe that it will be the future of mobile computing.
Following the announcement of the Optimus 2X smartphone from LG, Nvidia followed up with something that we foresaw since the launch of the Tesla C1060 and subsequently, GeForce GTX 280: The inevitable introduction of ARM into the Nvidia ecosystem for every aspect of computing, ranging from handheld all the way to HPC.
Nvidia has given the initiative the name Project Denver as recognition to the team located nearby Denver – who have been linked to false rumors on making an x86 CPU without a license – for years now. Today, nVidia unveiled what the team has been working on – Project Denver is nothing other than a reverse project to what AMD and Intel have been trying to do: instead of adding GPU silicon to the CPU, nVidia is implementing ARM CPU core inside its GPU design, meaning the memory controller and all other benefits of ultra-high speed communication which are trademark of GPU architecture – might become available for CPU technology as well. The reasons for this are both strategic and financial.
Project Denver specifically focuses on implementing ARM in a non-mobile format specifically focusing on HPC to complement their GPU HPC offerings. We will follow this up with a detailed article about this technology at a later time.