ATI or AMD R600, or whatever we should call the next generation graphics circuit from ATI, is a bit a of a riddle. It’s still a few months left until the launch and even if there has been some information about the circuit there are a lot of things that are different with the new circuit. At The Inquirer they’ve tried to summarize R600 in a two-part news article where they discuss the architecture, how it’s different from NVIDIA’s G80 despite they both are using unified shaders. But also other interesting details where the power consumption has been a bit exaggerated. The maximal power consumption will be 200-220W, but overall R600 should consume about as much as NVIDIA’s G80 circuit.
The PCB ATI has used with R600 will be the most advanced ATI has ever used up to this day and the board will actually be shorter than NVIDIA’s G80 card, although the specially designed cooler will make the card really big and heavy. The cooler has been designed to handle at least 250W and will thus be returning with future circuits.
We’ve earlier reported about R600’s extreme memory bandwidth and as if a 512-bit memory bus wasn’t enough ATI is planning both a 1GB and 2GB (!) memory buffer for its R600 card. The latter is most likely for workstations and not for us consumers, at first at least. But just 1GB of graphics memory is quite impressive. More interesting information about the R600 architecture and its 64 shader processors can be found in part one, while the information about the PCB and the memory handling can be found in part two.
No active posts found.