NVIDIA192 GB VRAM8,000 GB/s$6/hr

Llama 3.3 70B on B200 SXM 192GB: Does It Fit, How Fast, What It Costs

Llama 3.3 70B at FP8 (E4M3) needs 70.6 GB for its weights and 92.1 GB in total at 8K context with 8 concurrent users. That means it fits on a single B200 SXM 192GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
92.1 GB needed / 192 GB available
Decode speed
110.5
tok/s per stream
Aggregate
1326
tok/s @ 8 concurrent
Cost / 1M tokens
$1.26
1× B200 SXM 192GB
What changes as you add B200 SXM 192GBs
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× B200192 GB92.1 GB110.513261.58 s$1.26✓
2× B200384 GB92.1 GB187.84508806.5 ms$0.7395✓
4× B200768 GB92.1 GB375.618031418.2 ms$0.3697✓
8× B2001536 GB92.1 GB751.372122224.1 ms$0.1849✓
Concurrency vs context length
On 1× B200 SXM 192GB at FP8 (E4M3). KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K901.34 GB✓ fits
8K452.68 GB✓ fits
33K1110.74 GB✓ fits
66K521.47 GB✓ fits
131K242.95 GB✓ fits
Other GPUs that run Llama 3.3 70B
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB4×4039$0.1376
A100 80GB SXM42×1127$0.5916
H100 SXM5 80GB2×1888$0.7358
AMD MI300X 192GB1×878.3$0.7906
L40S 48GB2×486.8$0.7988
AMD MI325X 288GB1×994.3$0.9778
Frequently asked questions

Can Llama 3.3 70B run on B200 SXM 192GB?

Yes. Llama 3.3 70B at FP8 (E4M3) needs 92.1 GB including the KV cache, and B200 SXM 192GB has 192 GB — so it fits on a single GPU.

How fast is Llama 3.3 70B on B200 SXM 192GB?

110.5 tokens/sec per stream and 1326 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 1.58 s, and inter-token latency is 9.1 ms.

What does Llama 3.3 70B on B200 SXM 192GB cost per million tokens?

$1.26 per 1M output tokens at 1× B200 SXM 192GB ($6.00/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Llama 3.3 70B need on B200 SXM 192GB?

70.6 GB for the weights at FP8 (E4M3), plus 21.5 GB of KV cache for 8 concurrent requests at 8K context — 92.1 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is B200 SXM 192GB the right choice for Llama 3.3 70B?

It works. Compare $1.26 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 4080 16GB at $0.1031, and the fastest is 4× Cerebras CS-3 (WSE-3) at 45076 tok/s.

Adjust the workload yourself

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

Open in calculator