readygpu

AMD Radeon RX 7900 XTX vs NVIDIA Tesla P40

For running models locally, priced in the US. At 8k context the first runs 57 models fully in memory, the second 57.

In short

Both have 24 GB of memory for models; the AMD Radeon RX 7900 XTX has 960 GB/s of bandwidth and the NVIDIA Tesla P40 346 GB/s. At 8k context both run the same 57 models we track fully in memory. On the models both run fully in memory, the AMD Radeon RX 7900 XTX is about 2.5× as fast as the NVIDIA Tesla P40 (estimated, median across models). In the US the AMD Radeon RX 7900 XTX costs $1,250 at Amazon and the NVIDIA Tesla P40 $345 at Amazon: the NVIDIA Tesla P40 is $905 cheaper.

Updated . Speeds are estimates; how they are worked out.

AMD Radeon RX 7900 XTX

$1,250

Amazon

Add to build
NVIDIA Tesla P40

$345

Amazon

Add to build

Specifications

SpecAMD Radeon RX 7900 XTXNVIDIA Tesla P40
GPURadeon RX 7900 XTXTesla P40
Memory24 GB GDDR624 GB GDDR5
Memory bus384-bit384-bit
Memory bandwidth960 GB/s346 GB/s
FP16 compute123 TFLOPS11.8 TFLOPS
SoftwareROCmCUDA
Board power355 W250 W
PCIe generationGen 4Gen 3
PCIe lanesx16x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelAMD Radeon RX 7900 XTXNVIDIA Tesla P40
Qwen 3.8 27B23tok/sUD-Q6_K10tok/sUD-Q6_K
Qwen 3.6 27B23tok/sQ6_K9tok/sQ6_K
Gemma 4 31B23tok/sQ5_K_M9tok/sQ5_K_M
Qwen 3.6 35B-A3B152tok/sQ4_K_M62tok/sQ4_K_M
Qwen 3.5 27B23tok/sQ6_K9tok/sQ6_K
Gemma 4 26B-A4B80tok/sUD-Q6_K32tok/sUD-Q6_K
Qwen 3.5 35B-A3B153tok/sQ4_K_M62tok/sQ4_K_M
GLM 4.7 Flash79tok/sQ5_K_M32tok/sQ5_K_M
Devstral Small 2 24B26tok/sQ6_K10tok/sQ6_K
Nemotron 3.5 Lightning 30B-A3B138tok/sMXFP456tok/sMXFP4
gpt-oss 20B120tok/sMXFP449tok/sMXFP4
Qwen3 Coder 30B-A3B108tok/sQ5_K_M44tok/sQ5_K_M
Qwen3 32B24tok/sQ4_K_M10tok/sQ4_K_M
Gemma 4 12B40tok/sQ8_016tok/sQ8_0
Qwen 3.5 9B54tok/sQ8_022tok/sQ8_0
Mistral Small 3.2 24B26tok/sQ6_K10tok/sQ6_K
DeepSeek R1 Distill Qwen 32B24tok/sQ4_K_M10tok/sQ4_K_M
Gemma 3 27B23tok/sQ6_K9tok/sQ6_K
Ministral 3 14B34tok/sQ8_014tok/sQ8_0
Qwen3 14B31tok/sQ8_013tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B53tok/sQ8_022tok/sQ8_0
Qwen3 8B53tok/sQ8_022tok/sQ8_0
Phi-4 14B31tok/sQ8_012tok/sQ8_0
Qwen 3.5 4B110tok/sQ8_045tok/sQ8_0
Ministral 3 8B52tok/sQ8_021tok/sQ8_0
Granite 4.2 8B49tok/sQ8_020tok/sQ8_0
Gemma 4 E4B62tok/sQ8_025tok/sQ8_0
LFM2.5 8B-A1B216tok/sQ8_088tok/sQ8_0
Llama 3.1 8B69tok/sQ6_K28tok/sQ6_K
Qwen2.5 0.5B836tok/sQ8_0340tok/sQ8_0
Qwen3.5 0.8B567tok/sQ8_0230tok/sQ8_0
Qwen3.5 2B340tok/sQ5_K_M138tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B248tok/sQ8_0101tok/sQ8_0
MiniCPM5 2B174tok/sQ8_071tok/sQ8_0
LFM2.5 2.6B156tok/sQ8_064tok/sQ8_0
LFM2.5 VL 3B156tok/sQ8_064tok/sQ8_0
Spark X2.5 4B155tok/smlx-4bit63tok/smlx-4bit
LFM2.5 24B A2B130tok/sQ6_K53tok/sQ6_K
granite vision 4.1 4b123tok/sQ8_050tok/sQ8_0
granite 4.2 3b116tok/sQ8_047tok/sQ8_0
Mellum2.1 12B A2.5B Thinking112tok/sQ8_046tok/sQ8_0
FrogNano 4B 2609107tok/sQ8_043tok/sQ8_0
Kimi VL A3B Thinking 250694tok/sQ8_038tok/sQ8_0
Kimi VL A3B90tok/smlx-8bit37tok/smlx-8bit
DeepSeek R1 Distill Qwen 7B62tok/sQ8_025tok/sQ8_0
GLM 4 9B 041462tok/smlx-5bit25tok/smlx-5bit
DeepSeek R1 Distill Llama 8B55tok/sQ8_022tok/sQ8_0
Ministral 3 8B Reasoning 251252tok/sQ8_021tok/sQ8_0
GLM 4.6V Flash51tok/sQ8_021tok/sQ8_0
GLM 4.1V 9B Thinking44tok/smlx-8bit18tok/smlx-8bit
DeepSeek R1 Distill Qwen 14B31tok/sQ8_012tok/sQ8_0
Devstral Small 250729tok/sQ5_K_M12tok/sQ5_K_M
Magistral Small 250729tok/sQ5_K_M12tok/sQ5_K_M
Magistral Small 250929tok/sQ5_K_M12tok/sQ5_K_M
Magistral Small 250626tok/sQ6_K10tok/sQ6_K
Qwen3.8 27B OBLITERATED25tok/smlx-4bit10tok/smlx-4bit
granite 4.2 30b23tok/sQ5_K_M9tok/sQ5_K_M
Qwen 3.5 122B-A10Boffloadoffload
Qwen3 Coder Nextoffloadoffload
Mistral Small 4 119Boffloadoffload
GLM 4.5 Airoffloadoffload
Qwen3 Next 80B-A3Boffloadoffload
Kimi Linear 48B-A3Boffloadoffload
Llama 3.3 70Boffloadoffload
DeepSeek R1 Distill Llama 70Boffloadoffload
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
Qwen 3.8 Flash Nextwon't runwon't run
MiniMax M2.7won't runwon't run
Nemotron 3 Super 120B-A12Bwon't runwon't run
gpt-oss 120Bwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Zerowon't runwon't run
DeepSeek V3won't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.1 Terminuswon't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.5won't runwon't run
GLM 4.5Vwon't runwon't run
GLM 4.6won't runwon't run
GLM 4.6Vwon't runwon't run
GLM 4.7won't runwon't run
GLM 5won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi Dev 72Bwon't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the AMD Radeon RX 7900 XTX or the NVIDIA Tesla P40?
Choose the AMD Radeon RX 7900 XTX for speed. Choose the NVIDIA Tesla P40 to spend less.
Which can run bigger models, the AMD Radeon RX 7900 XTX or the NVIDIA Tesla P40?
At 8k context both run the same 57 models we track fully in memory.
Which is faster for local AI, the AMD Radeon RX 7900 XTX or the NVIDIA Tesla P40?
On the models both run fully in memory, the AMD Radeon RX 7900 XTX is about 2.5× as fast as the NVIDIA Tesla P40 (estimated, median across models).
What is the price difference between the AMD Radeon RX 7900 XTX and the NVIDIA Tesla P40?
In the US the AMD Radeon RX 7900 XTX costs $1,250 at Amazon and the NVIDIA Tesla P40 $345 at Amazon: the NVIDIA Tesla P40 is $905 cheaper.

More comparisons