Based on Navi 32 GPU and packing 48 CUs, 96 AI accelerator, 3072 Stream Prceossors, 64MB of AMD Infinity Cache, and 16GB of 256-bit GDDR6 memory with ECC, the AMD Radeon PRO W7700 provides 28.27 TFLOPs of peak single precision (FP32) compute performance and 567GB/s of memory bandwidth, all at a total board power of 190W. It also features 48 Ray Accelerators, supports AV1 encoding and decoding, and has four DisplayPort 2.1 outputs.
According to AMD's own benchmarks, the $999-priced Radeon PRO W7700 is faster than the $1,250-priced Nvidia RTX 4000 SFF ADA, in SOLIDWORKS, Maya, Photoshop, and DaVinci Resolve benchmarks. Interestingly enough, AMD did not include the rest of its own lineup in these benchmarks, but rather just the rest of the Nvidia RTX lineup, but it should fit well between the Navi 31-based Radeon PRO W7800 and the Navi 33-based Radeon PRO W7600.
"We built and optimized the AMD Radeon PRO W7700 graphics cards to create premium workstation graphics experiences, enabling our customers to boost innovation and productivity," said Dan Wood, corporate vice president, Radeon product management at AMD.
As said, the Radeon PRO W7700 is set to launch this month at $999 SEP.