NVIDIA80 GB VRAM3,350 GB/s$2.5/hr

DeepSeek R1 671B (MoE) on H100 SXM5 80GB: Does It Fit, How Fast, What It Costs

DeepSeek R1 671B (MoE) at AWQ 4-bit needs 369.1 GB for its weights and 631.0 GB in total at 8K context with 8 concurrent users. That means it needs 8× H100 SXM5 80GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 8×
631.0 GB needed / 640 GB available
Decode speed
640.3
tok/s per stream
Aggregate
61469
tok/s @ 8 concurrent
Cost / 1M tokens
$0.0904
8× H100 SXM5 80GB
What changes as you add H100 SXM5 80GBs
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× H10080 GB631.0 GB0.0000.000——✗
2× H100160 GB631.0 GB0.0000.000——✗
4× H100320 GB631.0 GB0.0000.000——✗
8× H100640 GB631.0 GB640.361469184.6 ms$0.0904✓
Concurrency vs context length
On 8× H100 SXM5 80GB at AWQ 4-bit. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K1616.37 GB✓ fits
8K832.75 GB✓ fits
33K2131.00 GB✓ fits
66K1261.99 GB✓ fits
131K0523.99 GB✗ does not fit
Other GPUs that run DeepSeek R1 671B (MoE)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
A100 80GB SXM48×36698$0.0727
H200 SXM5 141GB8×88075$0.1009
B200 SXM 192GB4×36698$0.1817
Frequently asked questions

Can DeepSeek R1 671B (MoE) run on H100 SXM5 80GB?

Yes, but it needs 8× H100 SXM5 80GB. At AWQ 4-bit the total footprint is 631.0 GB (369.1 GB of weights plus 262.0 GB of KV cache at 8K × 8 concurrent users), against 640 GB of available VRAM across 8 GPUs.

How fast is DeepSeek R1 671B (MoE) on H100 SXM5 80GB?

640.3 tokens/sec per stream and 61469 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 184.6 ms, and inter-token latency is 1.6 ms.

What does DeepSeek R1 671B (MoE) on H100 SXM5 80GB cost per million tokens?

$0.0904 per 1M output tokens at 8× H100 SXM5 80GB ($2.50/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does DeepSeek R1 671B (MoE) need on H100 SXM5 80GB?

369.1 GB for the weights at AWQ 4-bit, plus 262.0 GB of KV cache for 8 concurrent requests at 8K context — 631.0 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is H100 SXM5 80GB the right choice for DeepSeek R1 671B (MoE)?

It works. Compare $0.0904 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× A100 80GB SXM4 at $0.0727, and the fastest is 8× H200 SXM5 141GB at 88075 tok/s.

Adjust the workload yourself

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

Open in calculator