readygpu

Intel Arc Pro B60 vs NVIDIA Tesla P40

For running models locally, priced in Canada. At 32k context the first runs 26 models fully in memory, the second 26.

In short

Both have 24 GB of memory for models; the Intel Arc Pro B60 has 456 GB/s of bandwidth and the NVIDIA Tesla P40 346 GB/s. At 32k context both run the same 26 models we track fully in memory. On the models both run fully in memory, the Intel Arc Pro B60 is about 1.1× as fast as the NVIDIA Tesla P40 (estimated, median across models).

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

Intel Arc Pro B60

$1,060

Amazon

Add to build
NVIDIA Tesla P40

Not priced here

Add to build

Specifications

SpecIntel Arc Pro B60NVIDIA Tesla P40
GPUArc Pro B60Tesla P40
Memory24 GB GDDR624 GB GDDR5
Memory bus192-bit384-bit
Memory bandwidth456 GB/s346 GB/s
FP16 compute98.5 TFLOPS11.8 TFLOPS
SoftwareSYCL / oneAPICUDA
Board power200 W250 W
PCIe generationGen 5Gen 3
PCIe lanesx8x16

What each runs

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

ModelIntel Arc Pro B60NVIDIA Tesla P40
Qwen 3.8 27B10tok/sUD-Q5_K_M10tok/sUD-Q5_K_M
Qwen 3.6 27B11tok/sQ5_K_M10tok/sQ5_K_M
Gemma 4 31B10tok/sQ4_K_M10tok/sQ4_K_M
Qwen 3.6 35B-A3B49tok/sUD-Q4_K_M50tok/sUD-Q4_K_M
Qwen 3.5 27B10tok/sQ5_K_M10tok/sQ5_K_M
Gemma 4 26B-A4B30tok/sUD-Q5_K_M29tok/sUD-Q6_K
Qwen 3.5 35B-A3B49tok/sQ4_K_M50tok/sQ4_K_M
GLM 4.7 Flash24tok/sQ5_K_M24tok/sQ5_K_M
Devstral Small 2 24B10tok/sQ5_K_M10tok/sQ5_K_M
gpt-oss 20B39tok/sMXFP440tok/sMXFP4
Qwen3 Coder 30B-A3B25tok/sQ4_K_M26tok/sQ4_K_M
Gemma 4 12B17tok/sQ8_016tok/sQ8_0
Qwen 3.5 9B21tok/sQ8_020tok/sQ8_0
Mistral Small 3.2 24B10tok/sQ5_K_M10tok/sQ5_K_M
Gemma 3 27B10tok/sQ5_K_M10tok/sQ5_K_M
Ministral 3 14B12tok/sQ8_011tok/sQ8_0
Qwen3 14B11tok/sQ8_010tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B17tok/sQ8_016tok/sQ8_0
Qwen3 8B17tok/sQ8_016tok/sQ8_0
Phi-4 14B10tok/sQ8_010tok/sQ8_0
Qwen 3.5 4B41tok/sQ8_038tok/sQ8_0
Ministral 3 8B17tok/sQ8_016tok/sQ8_0
Granite 4.2 8B16tok/sQ8_015tok/sQ8_0
Gemma 4 E4B25tok/sQ8_023tok/sQ8_0
LFM2.5 8B-A1B47tok/sQ8_048tok/sQ8_0
Llama 3.1 8B21tok/sQ6_K20tok/sQ6_K
Qwen 3.5 122B-A10Boffloadoffload
Qwen3 Coder Nextoffloadoffload
Mistral Small 4 119Boffloadoffload
GLM 4.5 Airoffloadoffload
Qwen3 Next 80B-A3Boffloadoffload
Nemotron 3.5 Lightning 30B-A3Boffloadoffload
Kimi Linear 48B-A3Boffloadoffload
Qwen3 32Boffloadoffload
DeepSeek R1 Distill Qwen 32Boffloadoffload
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
Llama 3.3 70Bwon't runwon't run

Questions

Which is better for local AI, the Intel Arc Pro B60 or the NVIDIA Tesla P40?
Choose the Intel Arc Pro B60 for speed.
Which can run bigger models, the Intel Arc Pro B60 or the NVIDIA Tesla P40?
At 32k context both run the same 26 models we track fully in memory.
Which is faster for local AI, the Intel Arc Pro B60 or the NVIDIA Tesla P40?
On the models both run fully in memory, the Intel Arc Pro B60 is about 1.1× as fast as the NVIDIA Tesla P40 (estimated, median across models).

More comparisons