AMD Radeon Pro W6800, AI & Machine Learning Benchmarks & Specs

32GB · AI Score 3.0/100 · anchored estimate vs 51 measured cards

3 AI Score Includes estimates

We have not run AMD Radeon Pro W6800 on our bench. These figures are anchored estimates, interpolated per workload against the 51 GPUs we did measure (confidence: low). On Llama 3.1 8B (Q4_K_M) AMD Radeon Pro W6800 should deliver about 80.3 tokens/sec. Stepping up to Qwen3 32B it should hold roughly 18 tok/s. Llama 3.3 70B does not fit. It needs roughly 42GB and this card has 32GB. 1 of the 12 workloads won't fit on 32GB at the tested precision, Llama 3.3 70B. We publish those as hard gates rather than quietly dropping to a smaller quant. Image, video and 3D models aren't listed for this card. They're built for NVIDIA's CUDA, and on AMD they only run through ROCm workarounds whose speed depends more on the driver and OS than on the card, so we don't estimate them, and they count as not supported in the AI score. For text generation (llama.cpp runs natively on ROCm and Vulkan), the numbers above hold up.

AI & Machine Learning benchmark results

Text Generation tok/s 5

Qwen3 4B129.1

Llama 3.1 8B80.3

Qwen2.5-Coder 14B43.6

Qwen3 32B18.1

Workload	Result	Telemetry	Data
Qwen3 4B	129.1 tok/s	estimated	Est.
Llama 3.1 8B	80.3 tok/s	estimated	Est.
Qwen2.5-Coder 14B	43.6 tok/s	estimated	Est.
Qwen3 32B	18.1 tok/s	estimated	Est.
Llama 3.3 70B	✕ Won't fit	VRAM-gated at this precision	Est.

How this estimate is derived. This card hasn’t been through our bench yet, so its numbers are anchored estimates, interpolated from the 51 first-party measured cards (Bandwidth Theil-Sen ladder off NVIDIA anchors × vendor factor (LLM ×0.85, diffusion ×0.5); realistic VRAM floors). The VRAM “won’t fit” gates are exact, since they’re pure capacity limits. Confidence: low. Estimates are replaced with measured data as more silicon goes through the bench. Full methodology →

AMD Radeon Pro W6800 specifications

Architecture	RDNA 2
Stream processors	3,840
VRAM	32GB GDDR6
Memory bus	256-bit
Memory bandwidth	512 GB/s
Boost clock	2,320 MHz
TDP	250 W
Process	7nm
Interface	PCIe 4.0 x16
Release date	2021-06-08
Launch MSRP	$2,249

Verdict, capable, but 32GB sets the ceiling

AMD Radeon Pro W6800 scores 3.0/100, #59 of 102. It ran 4 of 12; 1 exceeded its 32GB. Figures are anchored estimates, not measurements, we flag that on every row.

Relative performance: where the AMD Radeon Pro W6800 lands

100% = this card, AI & Machine Learning headline metric (AI Score). #17 of 20 workstation cards in this vertical.

GPU	%	AI Score
AMD Radeon Pro W7900	150%	4.5
NVIDIA RTX A4000	137%	4.1
NVIDIA Quadro RTX 5000	117%	3.5
NVIDIA RTX 2000 Ada Generation	103%	3.1
AMD Radeon Pro W6800	100%	3
AMD Radeon Pro W7800	100%	3
Intel Arc Pro A60	60%	1.8
NVIDIA RTX A2000	43%	1.3

One tier up: NVIDIA RTX 2000 Ada Generation (+3%)
One tier down: AMD Radeon Pro W7800 (0%)
Best value alternative ($625): NVIDIA RTX 2000 Ada Generation (+3%)

← All AI & Machine Learning GPU rankings

The silicon

Transistors	26,800 million
Die size	520 mm²
Process node	7 nm
Fabricated by	TSMC
Transistor density	51.5 million per mm²

Denser than 85% of the 746 cards we have silicon data for. Density is the clearest measure of what a process node bought: a card that gained it without growing the die got its speed from the fab rather than the architecture.

Silicon figures from Wikipedia (CC BY-SA 4.0). Benchmarks on this page are our own. Compare every chip.

What this card can build

Whole-job timings, composed from our measured per-model results on this card.

Workflow	Time	Energy	Basis
Full codebase review	22.9 min	n/a	estimate, 0 of 1 stage measured

Can't run: Long-form article batch (needs Llama 3.3 70B).