What is the biggest model the RTX 5090 can run?

Qwen3 32B, which needs about 22 GB, on the 32 GB version at about 49 tok/s.

What is the biggest model the Instinct MI100 can run?

Qwen3 32B, which needs about 22 GB, on the 32 GB version at about 34 tok/s.

RTX 5090 vs Instinct MI100 for local AI

Q: Which is better for local AI, the RTX 5090 or the Instinct MI100?

The RTX 5090, for speed. They fit the same models. Both fit the same models we track, up to Qwen3 32B. The RTX 5090 generates about 46% faster on any model both can run, because it has 1792 GB/s of memory bandwidth to the Instinct MI100's 1229 GB/s.

Both fit the same models we track, up to Qwen3 32B. The RTX 5090 generates about 46% faster on any model both can run, because it has 1792 GB/s of memory bandwidth to the Instinct MI100's 1229 GB/s.

The numbers that matter

Memory and bandwidth

Memory decides which models fit. Bandwidth decides how fast they generate, because every token reads the model's active weights once.

	RTX 5090	Instinct MI100
Memory	32 GB VRAM	32 GB VRAM
Bandwidth	1792 GB/s	1229 GB/s
Biggest model that fits	Qwen3 32B	Qwen3 32B
Best everyday pick	Qwen3 32B, ~49 tok/s	Qwen3 32B, ~34 tok/s

Model by model

Every model, side by side

Estimated generation speed at the largest memory size each was sold in (32 GB and 32 GB).

Model	RTX 5090	Instinct MI100
Qwen3 4B About 4 GB	~390 tok/s Runs great	~270 tok/s Runs great
Qwen3 8B About 7 GB	~190 tok/s Runs great	~130 tok/s Runs great
Gemma 3 12B About 10 GB	~120 tok/s Runs great	~83 tok/s Runs great
Qwen3 14B About 11 GB	~110 tok/s Runs great	~73 tok/s Runs great
gpt-oss 20B About 16 GB	~310 tok/s Runs great	~220 tok/s Runs great
Gemma 3 27B About 19 GB	~58 tok/s Runs great	~40 tok/s Runs great
Qwen3 30B-A3B About 21 GB	~310 tok/s Runs great	~220 tok/s Runs great
Qwen3 32B About 22 GB	~49 tok/s Runs great	~34 tok/s Runs great
Llama 3.3 70B About 45 GB	Needs 45 GB	Needs 45 GB
gpt-oss 120B About 67 GB	Needs 67 GB	Needs 67 GB
Qwen3 235B-A22B About 144 GB	Needs 144 GB	Needs 144 GB
DeepSeek R1 671B About 406 GB	Needs 406 GB	Needs 406 GB

Which should you get?

The RTX 5090, for speed. They fit the same models. Price matters as much as either number: check used prices on the builds page, where every parts list has a measured or estimated speed.

How we estimate

Both columns use the same formula: memory bandwidth divided by the weights read per token, calibrated on our own rack, with Ollama's default 4-bit sizes. Real results are usually within about a third. Measured either one? Send us the number.

Compare

Other matchups

Every comparison