Computex 2024 | AMD Radeon Pro W7900 Dual Slot gets $500 price cut, up to 52% better perf-per-dollar compared to RTX 6000 Ada
Last year, AMD launched the Radeon Pro W7900 as a 3-slot workstation GPU based on full Navi 31 (RDNA 3). Today, at Computex 2024, AMD is once again re-introducing the Radeon Pro W7900 albeit in a more compact 2-slot form factor as the Radeon Pro W7900 Dual Slot (DS) and at a lower MSRP.
Specifications of the W7900 DS remain identical to the previous iteration. The W7900 DS features 96 compute units (CUs) or 6,144 stream processors (SPs) with up to 2.5 GHz boost and 192 AI accelerators. The card supports 48 GB of 384-bit GDDR6 ECC memory that yields an 864 GB/s bandwidth. There are a total of four DisplayPort 2.1-outs including one mini-DisplayPort.
Power requirements continue to remain the same at 295 W total board power (TBP). The Radeon Pro W7900 DS is slated to offer 61 peak single-precision FP32 TFLOPs, which is significantly lower than the 91 TFLOPs of the Nvidia RTX 6000 Ada.
On the other hand, Nvidia does not officially specify the FP16 peak half precision performance for the RTX 6000 Ada, but AMD claims 123 TFLOPs with the Radeon Pro W7900 DS. AMD also showed a relative ~2x performance-per-dollar benefit for the W7900 DS over the RTX 6000 Ada in enterprise workloads. The Navi 31 workstation flagship is also said to offer 38% better performance-per-dollar in Llama3 70 billion Q4 LLM inference compared to the RTX 6000 Ada.
The W7900 DS also gets a price cut this time and will retail at US$3,499 MSRP from June 19 instead of US$3,999 of the original, which itself is currently retailing at US$3,663 on Amazon.
ROCm 6.1 for Radeon Pro
ROCm, traditionally made for the AMD Instinct datacenter cards, is also officially supported on the Radeon VII (deprecated) and RX 7900 series GPUs such as the RX 7900 XTX on the consumer side along with professional cards like the Radeon Pro V620, W6800, W7800, and the W7900.
AMD is updating the ROCm software stack for Radeon Pro to version 6.1, which now offers Ubuntu 22.04.3 (Jammy Jellyfish) support both natively and via WSL 2 on Windows. ROCm 6.1 also adds rocDecode support that allows ROCm HIP post-processing of decoded compressed video streams within the GPU's VRAM itself instead of having to constantly ferry the same across the PCIe bus.
Are you a techie who knows how to write? Then join our Team! Wanted:
- News translator (DE-EN)
- Review translation proofreader (DE-EN)
Details here
Source(s)
AMD Press Release