~/best-gpu-for/mistral-7b
Mistral provider

Best GPU for Mistral 7B v0.3

Ranked recommendations from the 42-GPU catalog for running Mistral 7B v0.3 (7.2B params). Top picks by quality, then by cost.

Top picks.

$ ./rank --by quality
Tightest budget
RTX 3070
8GB · runs at Q5_K_M
5.0 GB weights · 0.60 GB headroom
Best quality
RTX 3090
24GB · runs at FP16/BF16
smallest card supporting the best available quant
Apple Silicon
M2 Max 64
64GB · runs at FP16/BF16
smallest M-series that fits
Datacenter (FP16 ambition)
H100 80GB
80GB · runs at FP16/BF16
best-quality datacenter card that fits

Full ranked list.

$ ./vrambudget --rank mistral-7b
GPUVRAMBest quantWeights at 15% safetyFit
nvidiaRTX 309024GBFP16/BF16
14fitsnvidiaRTX 3090 Ti24GBFP16/BF16
14fitsnvidiaRTX 409024GBFP16/BF16
14fitsamdRX 7900 XTX24GBFP16/BF16
14fitsintelArc Pro B6024GBFP16/BF16
14fitsnvidiaRTX 509032GBFP16/BF16
14fitsnvidiaRTX A600048GBFP16/BF16
14fitsnvidiaRTX 6000 Ada48GBFP16/BF16
14fitsnvidiaL40S48GBFP16/BF16
14fitsamdRadeon Pro W790048GBFP16/BF16
14fitsappleM2 Max 6464GBFP16/BF16
14fitsappleM3 Max 6464GBFP16/BF16
14fitsappleM4 Pro 6464GBFP16/BF16
14fitsappleM5 Pro 6464GBFP16/BF16
14fitsnvidiaH100 80GB80GBFP16/BF16
14fitsappleM3 Max 9696GBFP16/BF16
14fitsnvidiaRTX Pro 600096GBFP16/BF16
14fitsappleM4 Max 128128GBFP16/BF16
14fits

Next steps.

$ ./next
$ checkCan I run Mistral 7B v0.3?see all quants →full quant matrix$ tuneTune the mathopen calculator →/calc?model=mistral-7b

Discussion.

$ gh discussion list

// sign in with github to leave a comment. threads live in the repo's discussions tab.