Advanced Micro Devices Inc. said that early next year it plans to release a high-end chip intended to speed up application performance. The FireStream 9170 could be used in everything from high-performance computing systems found in scientific research labs to security systems.
The chip, part of AMD's FireStream line of "stream processors," is called a General Purpose General Processing Unit (GP GPU). It traces its origins to video cards, which process data sequentially -- in a straight stream, so to speak, hence the term "stream processing." This processing approach also works well for certain types of mathematical computations. The GP GPU operates as a co-processor to a CPU -- an Opteron, for instance.
AMD released its first processor in this line, FireStream, a year ago. But the company said it didn't push the initial chip in a widespread way, focusing initially on attracting developers. But with the FireStream 9170, AMD says it is eyeing what it sees as a potentially large market -- one that will initially include high-performance computing users and later general business users.
AMD is releasing a software development kit along with the processor, so developers can adapt their applications to work with the accelerator. It has also given the FireStream chip 64-bit double-precision floating point capability. AMD said it is the first chip maker to introduce the greater capability.
Patricia Harrell, director of stream computing at AMD, said that with the right kind of application, the FireStream chip can deliver up to a tenfold improvement in performance without any special application tuning, and the improvement would be greater with tuning.
FireStream technology may be appealing to users of high-performance computing (HPC) systems, such as those involved in financial analysis, seismic processing in oil and gas exploration, or life sciences research. It could also be used in security systems, such as face-recognition or image-matching tools. "This technology has really broad uses across other markets," Harrell said.
Not just throwing CPUs at speed problems
The approach that most HPC users take to improving application performance is to add more microprocessors, said Addison Snell, an analyst at Tabor Research, an HPC market research firm in San Diego. But he said HPC users are beginning to pay more attention to accelerators such as the GP GPU.
Snell noted, however, that there are challenges with effective GP GPU integration. One of those concerns the programmability of applications running with an accelerator: Rather than writing things for a standard processor, developers are trying to get various computational elements to go off the usual chip to a different part of the computer that can do the processing faster, he said. "The other challenge is gaining enough acceleration from doing so that you overcome the latency -- the time lag from going off one chip on to another chip," he said.
The software development kit AMD is releasing along with the chip itself should help developers address those programming issues, said Snell.
AMD said its FireStream chip will be released in the first quarter, priced at US$1,999.