readygpu

Intel Arc B580 vs AMD Radeon RX 7900 XT

For running models locally, priced in the UK. At 8k context the first runs 34 models fully in memory, the second 52.

In short

The AMD Radeon RX 7900 XT has more memory for models: 20 GB against 12 GB. At 8k context the AMD Radeon RX 7900 XT runs 52 of the models we track fully in memory and the Intel Arc B580 34; only the AMD Radeon RX 7900 XT holds Qwen 3.8 27B. On the models both run fully in memory, the AMD Radeon RX 7900 XT is about 1.9× as fast as the Intel Arc B580 (estimated, median across models). In the UK the Intel Arc B580 costs £700 at eBay and the AMD Radeon RX 7900 XT £700 at eBay (used): the Intel Arc B580 is £0 cheaper.

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

Intel Arc B580

£700

eBay

Add to build
AMD Radeon RX 7900 XT

£700

eBay, used (asking)

Add to build

Specifications

SpecIntel Arc B580AMD Radeon RX 7900 XT
GPUArc B580Radeon RX 7900 XT
Memory12 GB GDDR620 GB GDDR6
Memory bus192-bit320-bit
Memory bandwidth456 GB/s800 GB/s
FP16 compute116.5 TFLOPS103 TFLOPS
SoftwareSYCL / oneAPIROCm
Board power190 W315 W
PCIe generationGen 4Gen 4
PCIe lanesx8x16

What each runs

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

ModelIntel Arc B580AMD Radeon RX 7900 XT
Qwen 3.8 27Boffload24tok/sQ4_K_M
Qwen 3.6 27Boffload25tok/sQ4_K_M
Qwen 3.5 27Boffload25tok/sQ4_K_M
Gemma 4 26B-A4Boffload78tok/sQ5_K_M
GLM 4.7 Flashoffload76tok/sQ4_K_M
Devstral Small 2 24Boffload24tok/sQ5_K_M
gpt-oss 20Boffload100tok/sMXFP4
Qwen3 Coder 30B-A3Boffload100tok/sQ4_K_M
Gemma 4 12B22tok/sQ6_K34tok/sQ8_0
Qwen 3.5 9B23tok/sQ8_045tok/sQ8_0
Mistral Small 3.2 24Boffload24tok/sQ5_K_M
Gemma 3 27Boffload25tok/sQ4_K_M
Qwen3 14B22tok/sQ4_K_M26tok/sQ8_0
Ministral 3 14B21tok/sQ5_K_M28tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B23tok/sQ8_044tok/sQ8_0
Qwen3 8B23tok/sQ8_044tok/sQ8_0
Phi-4 14B22tok/sQ4_K_M26tok/sQ8_0
Qwen 3.5 4B48tok/sQ8_092tok/sQ8_0
Ministral 3 8B22tok/sQ8_043tok/sQ8_0
Granite 4.2 8B21tok/sQ8_041tok/sQ8_0
Gemma 4 E4B27tok/sQ8_052tok/sQ8_0
LFM2.5 8B-A1B86tok/sQ8_0180tok/sQ8_0
Llama 3.1 8B30tok/sQ6_K57tok/sQ6_K
Qwen2.5 0.5B361tok/sQ8_0697tok/sQ8_0
Qwen3.5 0.8B245tok/sQ8_0472tok/sQ8_0
Qwen3.5 2B147tok/sQ5_K_M283tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B107tok/sQ8_0207tok/sQ8_0
MiniCPM5 2B75tok/sQ8_0145tok/sQ8_0
LFM2.5 2.6B68tok/sQ8_0130tok/sQ8_0
LFM2.5 VL 3B68tok/sQ8_0130tok/sQ8_0
Spark X2.5 4B67tok/smlx-4bit130tok/smlx-4bit
Kimi VL A3B Thinking 250658tok/sQ4_K_M79tok/sQ8_0
Mellum2.1 12B A2.5B Thinking57tok/sQ5_K_M94tok/sQ8_0
granite vision 4.1 4b53tok/sQ8_0103tok/sQ8_0
Ministral 3 3B Instruct 251250tok/sQ8_097tok/sQ8_0
granite 4.2 3b50tok/sQ8_096tok/sQ8_0
FrogNano 4B 260946tok/sQ8_089tok/sQ8_0
DeepSeek R1 Distill Qwen 7B27tok/sQ8_051tok/sQ8_0
GLM 4 9B 041427tok/smlx-5bit51tok/smlx-5bit
DeepSeek R1 Distill Llama 8B24tok/sQ8_046tok/sQ8_0
Ministral 3 8B Reasoning 251222tok/sQ8_043tok/sQ8_0
GLM 4.6V Flash22tok/sQ8_043tok/sQ8_0
DeepSeek R1 Distill Qwen 14B22tok/sQ4_K_M25tok/sQ8_0
Ministral 3 14B Reasoning 251221tok/sQ5_K_M28tok/sQ8_0
Kimi VL A3Boffload75tok/smlx-8bit
LFM2.5 24B A2Boffload119tok/sQ5_K_M
GLM 4.1V 9B Thinkingoffload36tok/smlx-8bit
Devstral Small 2507offload24tok/sQ5_K_M
Magistral Small 2506offload24tok/sQ5_K_M
Magistral Small 2507offload24tok/sQ5_K_M
Magistral Small 2509offload24tok/sQ5_K_M
Muse Glimmer 30Boffload25tok/sQ4_K_M
Qwen 3.5 122B-A10Bwon't runoffload
Gemma 4 31Boffloadoffload
Mistral Small 4 119Bwon't runoffload
Qwen 3.6 35B-A3Boffloadoffload
GLM 4.5 Airwon't runoffload
Qwen 3.5 35B-A3Boffloadoffload
Nemotron 3.5 Lightning 30B-A3Boffloadoffload
Kimi Linear 48B-A3Boffloadoffload
Qwen3 32Boffloadoffload
DeepSeek R1 Distill Qwen 32Boffloadoffload
Llama 3.3 70Bwon't runoffload
granite 4.2 30boffloadoffload
Qwen3.8 27B OBLITERATEDoffloadoffload
DeepSeek R1 Distill Llama 70Bwon't runoffload
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
Qwen3 Coder Nextwon't runwon't run
gpt-oss 120Bwon't runwon't run
Qwen3 Next 80B-A3Bwon'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 Intel Arc B580 or the AMD Radeon RX 7900 XT?
Choose the AMD Radeon RX 7900 XT to run bigger models. Choose the AMD Radeon RX 7900 XT for speed. Choose the Intel Arc B580 to spend less.
Which can run bigger models, the Intel Arc B580 or the AMD Radeon RX 7900 XT?
At 8k context the AMD Radeon RX 7900 XT runs 52 of the models we track fully in memory and the Intel Arc B580 34; only the AMD Radeon RX 7900 XT holds Qwen 3.8 27B.
Which is faster for local AI, the Intel Arc B580 or the AMD Radeon RX 7900 XT?
On the models both run fully in memory, the AMD Radeon RX 7900 XT is about 1.9× as fast as the Intel Arc B580 (estimated, median across models).
What is the price difference between the Intel Arc B580 and the AMD Radeon RX 7900 XT?
In the UK the Intel Arc B580 costs £700 at eBay and the AMD Radeon RX 7900 XT £700 at eBay (used): the Intel Arc B580 is £0 cheaper.

More comparisons

More hardware for local AI