by Anton Shilov
12/26/2009 | 01:05 PM
Nvidia Corp. first demonstrated its code-named GF100 (NV60, G300, GT300, etc) graphics processing units (GPU) based on Fermi architecture over three months ago. Since the product is still not released, it is not a surprise that there are loads of speculations about performance of the next-generation GeForce. According to a representative at Nvidia, the company is happy about performance of the novelty.
“We expect [Fermi] to be the fastest GPU in every single segment. The performance numbers that we have [obtained] internally just [confirms] that. So, we are happy about this and are just finalizing everything on the driver side,” said Luciano Alibrandi, the head of Nvidia public relations department in EMEAI region, in an interview with DonanimHaber web-site.
It was untraditional for Nvidia to demonstrate capabilities of its forthcoming graphics card months ahead of actual release. It is generally believed that Nvidia showed off its Fermi product in late September in an attempt to slowdown sales of its arch-rival ATI’s DirectX 11 graphics processors. The company yet has to show performance numbers of its next-gen graphics products. However, according to Mr. Alibrandi, the company is not only working hard on the new chip itself, but is also developing “perfect” drivers in an attempt to provide ultimate experience for the end-user.
“We just want to make sure it is as perfect as we want it to be in both graphics and computing [performance and quality],” said Mr. Alibrandi.
To date, Nvidia has demonstrated its next-generation GeForce GF100 graphics card rendering a DirectX 11 benchmark and working in multi-GPU SLI mode.
The chief of Nvidia’s public relations department in EMEAI did not reveal when exactly Nvidia plans to launch its GF100 GPU for consumers. The only official comment on the matter was made by the company’s chief executive Jen-Hsun Huang, who said that production of the GF100 would be ramped only in Q1 FY2011. Nvidia’s first quarter of fiscal year 2011 begins on the 26th of January and ends on the 26th of April, 2010.
The flagship Fermi graphics processor will feature 512 stream processing engines (which are organized as 16 streaming multi-processors with 32 cores in each) that support a type of multi-threading technology to maximize utilization of cores. Each stream processor has a fully pipelined integer arithmetic logic unit (ALU) and floating point unit (FPU). The top-of-the-range chip contains 3 billion of transistors, features 384-bit memory GDDR5 memory controller with ECC and features rather unprecedented 768KB unified level-two cache as well as rather complex cache hierarchy in general. Naturally, the Fermi family is compatible with DirectX 11, OpenGL 3.x and OpenCL 1.x application programming interfaces (APIs). The new chips will be made using 40nm process technology at TSMC.