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

GTE Large (1024-dim) on H100 SXM5 80GB: Does It Fit, How Fast, What It Costs

GTE Large (1024-dim) at FP16 needs 0.9 GB for its weights and 7.3 GB in total at 8K context with 8 concurrent users. That means it fits on a single H100 SXM5 80GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
7.3 GB needed / 80 GB available
Decode speed
2474
tok/s per stream
Aggregate
29693
tok/s @ 8 concurrent
Cost / 1M tokens
$0.0234
1× 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 GB7.3 GB24742969345.0 ms$0.0234✓
2× H100160 GB7.3 GB420710095737.5 ms$0.0138✓
4× H100320 GB7.3 GB841340382733.8 ms$0.006879✓
8× H100640 GB7.3 GB16826161530931.9 ms$0.003439✓
Concurrency vs context length
On 1× H100 SXM5 80GB at FP16. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K196412.3 MB✓ fits
8K98824.6 MB✓ fits
33K243.22 GB✓ fits
66K126.44 GB✓ fits
131K612.88 GB✓ fits
Other GPUs that run GTE Large (1024-dim)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 3080 10GB1×6736$0.008247
RTX 3090 24GB1×8296$0.008370
RTX 5090 32GB1×15884$0.008744
RTX 4090 24GB1×8935$0.0124
RTX 4080 16GB1×6355$0.0131
AMD MI300X 192GB1×46977$0.0148
Frequently asked questions

Can GTE Large (1024-dim) run on H100 SXM5 80GB?

Yes. GTE Large (1024-dim) at FP16 needs 7.3 GB including the KV cache, and H100 SXM5 80GB has 80 GB — so it fits on a single GPU.

How fast is GTE Large (1024-dim) on H100 SXM5 80GB?

2474 tokens/sec per stream and 29693 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 45.0 ms, and inter-token latency is 404.1 µs.

What does GTE Large (1024-dim) on H100 SXM5 80GB cost per million tokens?

$0.0234 per 1M output tokens at 1× 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 GTE Large (1024-dim) need on H100 SXM5 80GB?

0.9 GB for the weights at FP16, plus 6.4 GB of KV cache for 8 concurrent requests at 8K context — 7.3 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is H100 SXM5 80GB the right choice for GTE Large (1024-dim)?

It works. Compare $0.0234 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 1× RTX 3080 10GB at $0.008247, and the fastest is 1× Cerebras CS-3 (WSE-3) at 177273 tok/s.

Adjust the workload yourself

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

Open in calculator