Despite of the fact that Nvidia Corp. unveiled some details about its next-generation GeForce architecture named Fermi at its GPU Technology Conference this week, introduction of actual products based on the code-named G300/GT300/NV60 graphics processing units (GPUs) are only expected to hit the market in Q1 2010, which is at least three months from now.
Nvidia is officially tight-lipped regarding the time-to-market of its next-generation GeForce, Quadro and Tesla products that are powered by the new Fermi architecture. However, a number of media reports (1, 2) suggest that the first graphics cards from Nvidia that support DirectX 11 and a number of other innovations will only become available in the first quarter of next year.
The lack of Nvidia DirectX 11 graphics processors on the market will leave ATI, graphics business unit of Advanced Micro Devices and the arch-rival of Nvidia, the only supplier of enthusiast-class graphics adapters in the coming months, which will help to boost sales as well as market share. In fact, ATI will quite comfortably sell its ATI Radeon HD 5870, 5850, 5770 and 5750 graphics cards during the holiday season to those, who want DirectX 11 now. However, economic turmoil may not allow AMD to fully take advantage of its current technology success. On the other hand, the first quarter of a year is rarely good for sales of graphics cards, hence, Nvidia will also hardly be able to sell a lot of its high-end graphics boards in Q1 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.
At the GTC, Nvidia demonstrated the A1 revision of the Fermi-G300 graphics chip made in late August, whereas usually the company uses only A2 or even A3 revisions on commercial products. It usually takes months to create a new revision, therefore, it is highly unlikely that the new GeForce GTX 380 arrives in 2009.