GPU directory · Radeon PRO · RDNA 3 · gfx1100
Cited, not yet measured on our R9700s
AMD Radeon PRO W7800
32 GB gfx1100 RDNA 3 workstation — the W7800, not the 48 GB W7800 SKU and not the 48 GB W7900. AMD product page cites 32 GB GDDR6, 576 GB/s peak bandwidth, and 260 W TBP; the datasheet lists an active double-slot 11 in / 280 mm board. AMD also lists a separate 48 GB W7800 (864 GB/s, 260 W). Official on the cited ROCm 10.1 matrix as gfx1100. llama.cpp first. Fit tables cover 35B, Flash, and the cited /models pages on 1 GPU and 2 GPUs. Speeds are not yet measured on our R9700s.
If the app is not installed, packages are not published yet. There is no public download URL. Read the Stack Lab install notes . If Stack Lab is already running on this machine, open the local app .
How we test — Measured cells cite official amdai-bench/run@1 cards. Estimates stay labeled. Q3 is estimate-only. Q1/Q2 stay off.
Official identity
- Card
- AMD Radeon PRO W7800
- gfx
- gfx1100
- RDNA 3
- Peak bandwidth
- 576 GB/s
- AMD W7800 product page — peak memory bandwidth 576 GB/s (32 GB SKU)
- Board power
- 260 W TBP
- AMD W7800 product page — Total Board Power 260 W; 2×8-pin
- ROCm matrix
- Official
- 32 GB and a separate 48 GB SKU are both listed. Read the sticker.
Form factor: Active cooler, full-height double-slot, 11 in / 280 mm. AMD W7800 datasheet — Active, 11” (280 mm), Double Slot, Full Height. AMD lists a separate W7800 48 GB SKU (48 GB GDDR6, 864 GB/s, 260 W TBP). Read the sticker. This sheet sizes the 32 GB W7800.
Cited support row: ROCm 10.1.0 compatibility matrix and Linux system requirements. We do not invent a support claim past that listing. AMD product page: PRO W7800. Datasheet: PRO W7800 datasheet. 48 GB SKU: W7800 48 GB.
HSA_OVERRIDE: Not needed on current 10.x.
What fits on this card (256K context)
Shopper verdict at 4K, then follow the pair page for 32K and full context. Q4 and up. No printed prices.
| Model | Verdict |
|---|---|
| Qwen3.6-35B on PRO W7800 | Fits fully on GPU |
| Qwen 3.8 Flash on PRO W7800 | Fits with RAM offload |
| gpt-oss 20B on PRO W7800 | Fits fully on GPU |
| Gemma 3 27B on PRO W7800 | Needs 2 GPUs |
| Llama 3.3 70B on PRO W7800 | Fits with RAM offload |
| GLM 5.3 Flash on PRO W7800 | Fits with RAM offload |
| Qwen3.6 27B on PRO W7800 | Needs 2 GPUs |
| Qwen3.8 27B on PRO W7800 | Needs 2 GPUs |
Workstation vs consumer
Built only from cited AMD product-page specs already on this sheet and the consumer /gpus pages. We do not invent a decode number or a street price.
-
32 GB W7800 vs the 48 GB W7800 SKU
This sheet sizes the 32 GB W7800 (576 GB/s, 260 W TBP). AMD’s separate W7800 48 GB page cites 48 GB GDDR6, 864 GB/s, and the same 260 W TBP. Read the sticker. Do not use these 32 GB fit rows for the 48 GB SKU.
-
W7800 vs the measured R9700
Both are 32 GB workstation stickers. The R9700 is gfx1201, 640 GB/s, 300 W TBP, and measured on the local desktop. The W7800 is gfx1100, 576 GB/s, 260 W TBP, official on the cited ROCm 10.1 row — not measured. Same VRAM class is not the same LLVM target.
-
W7800 vs consumer gfx1100
The RX 7900 XT is 20 GB gfx1100 at 315 W typical board power. The W7800 is 32 GB gfx1100 at 260 W TBP. More VRAM on the workstation sticker; different cooler and driver SKU. We do not invent a price or a decode gap.
What fits in VRAM
llama.cpp, f16 KV, 1 GPU and 2 GPUs (split). Q6 / Q5 / Q4 rows are calculator estimates. UD-Q3_K_XL is a labeled estimate — not a local desktop run. Speed stays not yet measured on our R9700s. Tables cover house 35B, Flash, and every cited /models page. Change quant or context in the calculator.
Qwen3.6-35B-A3B
llama.cpp on PRO W7800 at 32 GB. UD-Q3_K_XL cite: Hugging Face listing for Qwen3.6-35B-A3B-UD-Q3_K_XL.gguf (16.8 GB).
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 26.7 GB | 27.7 GB Fits in VRAM (estimate) | 28.2 GB Fits in VRAM (estimate) | 13.8 GB / GPU Fits in VRAM (estimate) | 14.1 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q5_K_M | Estimate | 24.8 GB | 25.7 GB Fits in VRAM (estimate) | 26.2 GB Fits in VRAM (estimate) | 12.8 GB / GPU Fits in VRAM (estimate) | 13.1 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 19.8 GB | 20.6 GB Fits in VRAM (estimate) | 21.2 GB Fits in VRAM (estimate) | 10.3 GB / GPU Fits in VRAM (estimate) | 10.6 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 15.6 GB | 16.4 GB Fits in VRAM (estimate) | 17.0 GB Fits in VRAM (estimate) | 8.21 GB / GPU Fits in VRAM (estimate) | 8.49 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
Qwen 3.8 Flash
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Hugging Face folder listing for UD-Q3_K_XL (~90 GB across 3 shards; LFS sizes sum 89.99 GB). Converted 90 × 10⁹ bytes → GiB. Not measured VRAM..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 157.4 GB | 160.4 GB Does not fit this GPU | 161.2 GB Does not fit this GPU | 80.2 GB / GPU Does not fit this GPU | 80.6 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| Q5_K_M | Estimate | 147.4 GB | 150.3 GB Does not fit this GPU | 151.1 GB Does not fit this GPU | 75.1 GB / GPU Does not fit this GPU | 75.6 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| Q4_K_M | Estimate | 103.4 GB | 105.6 GB Does not fit this GPU | 106.4 GB Does not fit this GPU | 52.8 GB / GPU Does not fit this GPU | 53.2 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 83.8 GB | 85.7 GB Does not fit this GPU | 86.5 GB Does not fit this GPU | 42.9 GB / GPU Does not fit this GPU | 43.3 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
RAM offload on 1 GPU — Minimum system RAM needed and RAM needed at full context, labeled estimate, rounded up to a real kit.
Minimum system RAM needed
128 GB kit
Estimate · 98.4 GiB raw · 32K context
RAM needed at full context
128 GB kit
Estimate · 105.1 GiB raw · 256K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 128 GB DIMM kit.
Breakdown
- Offloaded weights ·
- 71.4 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 1.05 GiB
- KV in RAM at full context ·
- 7.61 GiB
- Runtime / backend ·
- 1.97 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
RAM offload on 2 GPUs — same two labels, labeled estimate.
Minimum system RAM needed
96 GB kit
Estimate · 66.4 GiB raw · 32K context
RAM needed at full context
96 GB kit
Estimate · 73.1 GiB raw · 256K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 96 GB DIMM kit.
Breakdown
- Offloaded weights ·
- 39.4 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 1.05 GiB
- KV in RAM at full context ·
- 7.61 GiB
- Runtime / backend ·
- 1.97 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
gpt-oss 20B
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Estimate: total_params × 3.91 bpw (Unsloth UD Q3 bucket; uses Q3_K_M bpw unless a publisher file size is listed) / 8. Active params do not shrink the file..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 11.2 GB | 11.9 GB Fits in VRAM (estimate) | 12.6 GB Fits in VRAM (estimate) | 5.95 GB / GPU Fits in VRAM (estimate) | 6.28 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q5_K_M | Estimate | 10.9 GB | 11.6 GB Fits in VRAM (estimate) | 12.3 GB Fits in VRAM (estimate) | 5.80 GB / GPU Fits in VRAM (estimate) | 6.13 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 10.8 GB | 11.5 GB Fits in VRAM (estimate) | 12.2 GB Fits in VRAM (estimate) | 5.75 GB / GPU Fits in VRAM (estimate) | 6.08 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 9.79 GB | 10.4 GB Fits in VRAM (estimate) | 11.1 GB Fits in VRAM (estimate) | 5.22 GB / GPU Fits in VRAM (estimate) | 5.56 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
Gemma 3 27B
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Estimate: total_params × 3.91 bpw (Unsloth UD Q3 bucket; uses Q3_K_M bpw unless a publisher file size is listed) / 8. Active params do not shrink the file..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 20.6 GB | 22.3 GB Fits in VRAM (estimate) | 24.5 GB Fits in VRAM (estimate) | 11.1 GB / GPU Fits in VRAM (estimate) | 12.3 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q5_K_M | Estimate | 17.9 GB | 19.6 GB Fits in VRAM (estimate) | 21.8 GB Fits in VRAM (estimate) | 9.78 GB / GPU Fits in VRAM (estimate) | 10.9 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 15.4 GB | 17.0 GB Fits in VRAM (estimate) | 19.2 GB Fits in VRAM (estimate) | 8.49 GB / GPU Fits in VRAM (estimate) | 9.60 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 12.5 GB | 14.0 GB Fits in VRAM (estimate) | 16.2 GB Fits in VRAM (estimate) | 7.00 GB / GPU Fits in VRAM (estimate) | 8.11 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
Llama 3.3 70B
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Estimate: total_params × 3.91 bpw (Unsloth UD Q3 bucket; uses Q3_K_M bpw unless a publisher file size is listed) / 8. Active params do not shrink the file..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 53.9 GB | 56.4 GB Does not fit this GPU | 65.3 GB Does not fit this GPU | 28.2 GB / GPU Fits in VRAM (estimate) | 32.6 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| Q5_K_M | Estimate | 46.5 GB | 48.9 GB Does not fit this GPU | 57.8 GB Does not fit this GPU | 24.4 GB / GPU Fits in VRAM (estimate) | 28.9 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 39.6 GB | 41.9 GB Does not fit this GPU | 50.7 GB Does not fit this GPU | 20.9 GB / GPU Fits in VRAM (estimate) | 25.4 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 32.1 GB | 34.3 GB Does not fit this GPU | 43.2 GB Does not fit this GPU | 17.1 GB / GPU Fits in VRAM (estimate) | 21.6 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
RAM offload on 1 GPU — Minimum system RAM needed and RAM needed at full context, labeled estimate, rounded up to a real kit.
Minimum system RAM needed
48 GB kit
Estimate · 42.7 GiB raw · 32K context
RAM needed at full context
48 GB kit
Estimate · 35.1 GiB raw · 8K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 48 GB DIMM kit.
Breakdown
- Offloaded weights ·
- 7.60 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 10.0 GiB
- KV in RAM at full context ·
- 2.50 GiB
- Runtime / backend ·
- 1.14 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
RAM offload on 2 GPUs — same two labels, labeled estimate.
Minimum system RAM needed
32 GB kit
Estimate · 25.1 GiB raw · 32K context
RAM needed at full context
32 GB kit
Estimate · 25.0 GiB raw · 8K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 32 GB DIMM kit.
Breakdown
- Offloaded weights ·
- 0 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 0 GiB
- KV in RAM at full context ·
- 0 GiB
- Runtime / backend ·
- 1.14 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
GLM 5.3 Flash
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Estimate: total_params × 3.91 bpw (Unsloth UD Q3 bucket; uses Q3_K_M bpw unless a publisher file size is listed) / 8. Active params do not shrink the file..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 271.8 GB | 276.5 GB Does not fit this GPU | 277.0 GB Does not fit this GPU | 138.2 GB / GPU Does not fit this GPU | 138.5 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| Q5_K_M | Estimate | 223.8 GB | 227.8 GB Does not fit this GPU | 228.3 GB Does not fit this GPU | 113.9 GB / GPU Does not fit this GPU | 114.2 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| Q4_K_M | Estimate | 186.0 GB | 189.4 GB Does not fit this GPU | 189.9 GB Does not fit this GPU | 94.7 GB / GPU Does not fit this GPU | 95.0 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 145.7 GB | 148.5 GB Does not fit this GPU | 149.0 GB Does not fit this GPU | 74.2 GB / GPU Does not fit this GPU | 74.5 GB / GPU Does not fit this GPU | not yet measured on our R9700s |
RAM offload on 1 GPU — Minimum system RAM needed and RAM needed at full context, labeled estimate, rounded up to a real kit.
Minimum system RAM needed
192 GB kit
Estimate · 181.9 GiB raw · 32K context
RAM needed at full context
192 GB kit
Estimate · 181.5 GiB raw · 8K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 192 GB DIMM kit. The local desktop ships 256 GB DDR5; measured host-RAM usage is not recorded until Local AI numbers arrive.
Breakdown
- Offloaded weights ·
- 154.0 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 0.74 GiB
- KV in RAM at full context ·
- 0.29 GiB
- Runtime / backend ·
- 3.20 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
RAM offload on 2 GPUs — same two labels, labeled estimate.
Minimum system RAM needed
192 GB kit
Estimate · 149.9 GiB raw · 32K context
RAM needed at full context
192 GB kit
Estimate · 149.5 GiB raw · 8K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 192 GB DIMM kit. The local desktop ships 256 GB DDR5; measured host-RAM usage is not recorded until Local AI numbers arrive.
Breakdown
- Offloaded weights ·
- 122.0 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 0.74 GiB
- KV in RAM at full context ·
- 0.29 GiB
- Runtime / backend ·
- 3.20 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
Qwen3.6 27B
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Hugging Face LFS size for unsloth/Qwen3.6-27B-GGUF Qwen3.6-27B-UD-Q3_K_XL.gguf (14.47 GB). Q3 is a labeled estimate — not a local desktop run. Converted 14.47 × 10⁹ bytes → GiB. Not measured VRAM..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 21.0 GB | 22.1 GB Fits in VRAM (estimate) | 23.9 GB Fits in VRAM (estimate) | 11.0 GB / GPU Fits in VRAM (estimate) | 11.9 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q5_K_M | Estimate | 18.2 GB | 19.2 GB Fits in VRAM (estimate) | 21.0 GB Fits in VRAM (estimate) | 9.62 GB / GPU Fits in VRAM (estimate) | 10.5 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 15.7 GB | 16.7 GB Fits in VRAM (estimate) | 18.5 GB Fits in VRAM (estimate) | 8.35 GB / GPU Fits in VRAM (estimate) | 9.24 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 13.5 GB | 14.5 GB Fits in VRAM (estimate) | 16.3 GB Fits in VRAM (estimate) | 7.24 GB / GPU Fits in VRAM (estimate) | 8.13 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
Qwen3.8 27B
llama.cpp on PRO W7800 at 32 GB. Q3 weights note: Hugging Face LFS size for unsloth/Qwen3.8-27B-GGUF Qwen3.8-27B-UD-Q3_K_XL.gguf (13.15 GB). Q3 is a labeled estimate — not a local desktop run. Converted 13.15 × 10⁹ bytes → GiB. Not measured VRAM..
| Quant | Kind | Weights | 1 GPU 4K | 1 GPU 32K | 2 GPUs 4K | 2 GPUs 32K | Speed |
|---|---|---|---|---|---|---|---|
| Q6_K | Estimate | 20.5 GB | 21.6 GB Fits in VRAM (estimate) | 23.4 GB Fits in VRAM (estimate) | 10.8 GB / GPU Fits in VRAM (estimate) | 11.7 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q5_K_M | Estimate | 18.4 GB | 19.5 GB Fits in VRAM (estimate) | 21.3 GB Fits in VRAM (estimate) | 9.75 GB / GPU Fits in VRAM (estimate) | 10.6 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| Q4_K_M | Estimate | 15.3 GB | 16.4 GB Fits in VRAM (estimate) | 18.1 GB Fits in VRAM (estimate) | 8.18 GB / GPU Fits in VRAM (estimate) | 9.07 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
| UD-Q3_K_XL | Q3 estimate | 12.2 GB | 13.2 GB Fits in VRAM (estimate) | 15.0 GB Fits in VRAM (estimate) | 6.62 GB / GPU Fits in VRAM (estimate) | 7.50 GB / GPU Fits in VRAM (estimate) | not yet measured on our R9700s |
House 35B Q3 cite: Hugging Face listing for Qwen3.6-35B-A3B-UD-Q3_K_XL.gguf (16.8 GB). Other Q3 rows use the calculator publisher-file note for that model.
System RAM
Minimum system RAM needed and RAM needed at full context for Qwen3.6-35B-A3B on llama.cpp at Q4_K_M, 32K selected and the model’s full context. Estimate until a first-party amdai-bench/run@1 card exists for this SKU. Rounded up to a real DIMM kit. Offload rows on the other /models tables repeat both labels.
Minimum system RAM needed
32 GB kit
Estimate · 24.7 GiB raw · 32K context
RAM needed at full context
32 GB kit
Estimate · 24.8 GiB raw · 256K full context
Estimate. Offloaded weights + KV in RAM + runtime + OS/desktop + headroom, rounded up to a 32 GB DIMM kit.
Breakdown
- Offloaded weights ·
- 0 GiB
- Separable embed in RAM ·
- 0 GiB
- KV in RAM at selected ·
- 0 GiB
- KV in RAM at full context ·
- 0 GiB
- Runtime / backend ·
- 0.71 GiB
- OS / desktop baseline ·
- 8.00 GiB
- Headroom ·
- 16.0 GiB
llama.cpp-first stack
-
Step 1
llama.cpp first
Vulkan and ROCm/HIP are both in scope on this official gfx1100 workstation row. HIP when rocminfo lists gfx1100. Ollama and vLLM are secondary — not the house path for this card.
-
Step 2
gfx1100 family package
gfx1100 needs the gfx110x ROCm family package, same pin as the W7900. Do not mix a gfx1201 (R9700 / 9070 XT) or gfx1101 (W7700 / 7800 XT) install. Re-check the cited matrix before you install.
-
Step 3
32 GB sheet, 48 GB SKU exists
This sheet sizes the 32 GB W7800. AMD product page: 32 GB GDDR6, 576 GB/s, 260 W TBP. The 48 GB W7800 is a different listing (864 GB/s, 260 W). Read the sticker and open the calculator before you download.
-
Step 4
Setup, then Stack Lab doctor
First-run setup is Ubuntu. Open Stack Lab doctor for the gfx110x install family, then record an amdai-bench/run@1 card if you measure this SKU. House daily and Flash profiles are R9700 / local desktop — they do not apply here. Packages are not published yet.
Compare with…
Head-to-head pages use the same VRAM, bandwidth, and calculator fit already on the sheets. Q3 is a labeled estimate. No invented prices. PRO W7800 against the other published sheets:
Also: PRO W7700 · RX 6700 XT · RX 6800 XT · RX 6950 XT · RX 7600 · RX 7600 XT · RX 7700 XT · RX 7800 XT · RX 7900 GRE · RX 7900 XT · RX 7900 XTX · RX 9060 XT · RX 9070 · RX 9070 GRE · RX 9070 XT · Strix Halo
Check price · no live or invented prices
AMD Radeon PRO W7800
32 GB workstation listing. AMD also lists a separate 48 GB W7800 SKU. No live or invented price.
- Amazon: check price
- Newegg: check price
- B&H Photo: check price
- Micro Center: check price
- AMD Store: check price
Return terms vary by seller and item. Newegg's Hassle-Free page covers only Sold-and-Shipped-by-Newegg listings; marketplace items use the seller's policy. See the listing's Warranty & Returns tab.
As an Amazon Associate I earn from qualifying purchases. Some Newegg links use Rakuten Advertising and some B&H links use B&H’s affiliate program when those IDs are configured. Micro Center and AMD Store links are plain “also available at” links — no tracking. Buttons say “Check price” and never invent or live-quote a price. Full disclosure.
Configurations we have
No first-party amdai-bench/run@1 card exists for this SKU. llama.cpp first. Empty cells stay empty. Fit numbers on this page are calculator estimates. Shopper RAM lines stay Minimum system RAM needed and RAM needed at full context — labeled estimate until measured.
No tested recipe yet — contribute. This card stays listed so the gap is visible.