AMD288 GB VRAM6,000 GB/s$3.5/hr

Qwen2 7B on AMD MI325X 288GB: Does It Fit, How Fast, What It Costs

Qwen2 7B at FP16 needs 15.2 GB for its weights and 19.0 GB in total at 8K context with 8 concurrent users. That means it fits on a single AMD MI325X 288GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
19.0 GB needed / 288 GB available
Decode speed
255.9
tok/s per stream
Aggregate
3071
tok/s @ 8 concurrent
Cost / 1M tokens
$0.3166
1× AMD MI325X 288GB
What changes as you add AMD MI325X 288GBs
Tensor-parallel degree vs VRAM, throughput, and cost. Cost scales linearly with GPU count; decode throughput scales near-linearly too, so cost per token stays roughly flat.
GPUsTotal VRAMVRAM neededDecode tok/sAggregate tok/sTTFTCost / 1M tokFits?
1× AMD288 GB19.0 GB255.93071224.9 ms$0.3166✓
2× AMD576 GB19.0 GB435.010441127.8 ms$0.1862✓
4× AMD1152 GB19.0 GB870.14176478.9 ms$0.0931✓
8× AMD2304 GB19.0 GB174016705554.5 ms$0.0466✓
Concurrency vs context length
On 1× AMD MI325X 288GB at FP16. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K1161240.5 MB✓ fits
8K580481.0 MB✓ fits
33K1451.88 GB✓ fits
66K723.76 GB✓ fits
131K367.52 GB✓ fits
Other GPUs that run Qwen2 7B
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 3090 24GB1×479.1$0.1450
RTX 5090 32GB1×917.2$0.1514
RTX 4090 24GB1×515.9$0.2154
AMD MI300X 192GB1×2713$0.2560
V100 SXM2 32GB1×460.6$0.3015
A100 40GB SXM41×793.3$0.3151
Frequently asked questions

Can Qwen2 7B run on AMD MI325X 288GB?

Yes. Qwen2 7B at FP16 needs 19.0 GB including the KV cache, and AMD MI325X 288GB has 288 GB — so it fits on a single GPU.

How fast is Qwen2 7B on AMD MI325X 288GB?

255.9 tokens/sec per stream and 3071 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 224.9 ms, and inter-token latency is 3.9 ms.

What does Qwen2 7B on AMD MI325X 288GB cost per million tokens?

$0.3166 per 1M output tokens at 1× AMD MI325X 288GB ($3.50/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Qwen2 7B need on AMD MI325X 288GB?

15.2 GB for the weights at FP16, plus 3.8 GB of KV cache for 8 concurrent requests at 8K context — 19.0 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is AMD MI325X 288GB the right choice for Qwen2 7B?

It works. Compare $0.3166 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 2× RTX 3080 10GB at $0.0840, and the fastest is 1× Cerebras CS-3 (WSE-3) at 10236 tok/s.

Adjust the workload yourself

Same formulas, your context length, concurrency, and quantisation.

Open in calculator