by Anton Shilov
01/06/2011 | 05:33 PM
The chief scientist of Nvidia has released certain plans concerning Nvidia's intentions to developer client- and server-class solutions with Nvidia stream cores and ARM "latency" cores onside.
"Nvidia's project Denver will usher in a new era for computing by extending the performance range of the ARM instruction-set architecture, enabling the ARM architecture to cover a larger portion of the computing space. Coupled with an Nvidia GPU, it will provide the heterogeneous computing platform of the future by combining a standard architecture with awesome performance and energy efficiency," said William Dally, the chief scientist officer of Nvidia.
ARM is already the standard architecture for mobile devices. Project Denver is supposed to extend the range of ARM-based systems upward to PCs, data center servers, and supercomputers.
Announcement of Windows 8 essentially frees PCs, workstations and servers from the hegemony of Windows and x86 architecture, which will naturally enable a lot of motivation to develop numerous apps for Windows x86 based on ARM.
For several years, makers of high-performance computing platforms (HPC) have had no choice about instruction-set architecture, but x86. Denver provides a choice, but requires to change the architecture of software.
Nvidia's initiative code-named Denver describes an Nvidia CPU running the ARM instruction set, which will be fully integrated on the same chip as the Nvidia GPU. This new processor stems from a strategic partnership, also announced today, in which Nvidia has obtained rights to develop its own high performance CPU cores based on ARM's future processor architecture. In addition, Nvidia licensed ARM's current Cortex-A15 processor for its future-generation Tegra mobile processors.
Nvidia's intention to integrate ARM-based microprocessor cores had been known for over a year now. Back in November, 2010, the company even unveiled plans to develop a so-called Echelon design chip that incorporates a large number (~1024) of stream cores and a smaller (~8) number of latency-optimized CPU-like cores on a single chip. As a result, the current announcement is just a formal confirmation of Nvidia's plans to enter the market of accelerated processing units (APUs, the chips that feature both CPU and GPU cores on the same die) eventually.
Bill Dally is a specialist on multi-parallel chips.