Compatibility check
Can I run Llama 3.3 70B Instruct on H100 80GB SXM?
Yes — it runs.best quant:
Decode speed
63 tok/sest
Usable context
131K
Memory
80 GB
Fits in memory63 tok/sest · 131K ctx
Fits in memory53 tok/sest · 131K ctx
Spills to system RAMest
