~/best-gpu-for/yi-34b
01.AI

Best GPU for Yi 34B

Ranked recommendations from the 42-GPU catalog for running Yi 34B (34B params). Top picks by quality, then by cost.

Top picks.

$ ./rank --by quality
Tightest budget
RTX 3090
24GB · runs at Q3_K_M
15 GB weights · 4.1 GB headroom
Best quality
M3 Max 96
96GB · runs at FP16/BF16
smallest card supporting the best available quant
Apple Silicon
M2 Max 64
64GB · runs at Q8_0
smallest M-series that fits
Datacenter (FP16 ambition)
H200
141GB · runs at FP16/BF16
best-quality datacenter card that fits

Full ranked list.

$ ./vrambudget --rank yi-34b
GPUVRAMBest quantWeights at 15% safetyFit
appleM3 Max 9696GBFP16/BF16
68fitsnvidiaRTX Pro 600096GBFP16/BF16
68fitsappleM4 Max 128128GBFP16/BF16
68fitsappleM5 Max 128128GBFP16/BF16
68fitsnvidiaDGX Spark128GBFP16/BF16
68fitsintelGaudi 3128GBFP16/BF16
68fitsnvidiaH200141GBFP16/BF16
68fitsnvidia2× H100 NVL188GBFP16/BF16
68fitsappleM2 Ultra 192192GBFP16/BF16
68fitsnvidiaB200192GBFP16/BF16
68fitsamdMI300X192GBFP16/BF16
68fitsappleM3 Ultra 512512GBFP16/BF16
68fitsnvidiaRTX A600048GBQ8_0
36fitsnvidiaRTX 6000 Ada48GBQ8_0
36fitsnvidiaL40S48GBQ8_0
36fitsamdRadeon Pro W790048GBQ8_0
36fitsappleM2 Max 6464GBQ8_0
36fitsappleM3 Max 6464GBQ8_0
36fits

Next steps.

$ ./next
$ checkCan I run Yi 34B?see all quants →full quant matrix$ tuneTune the mathopen calculator →/calc?model=yi-34b

Discussion.

$ gh discussion list

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