NVIDIA48 GB VRAM864 GB/s$0.7/hr

Mixtral 8x7B (MoE) on L40S 48GB: Does It Fit, How Fast, What It Costs

Mixtral 8x7B (MoE) at FP8 (E4M3) needs 46.7 GB for its weights and 55.3 GB in total at 8K context with 8 concurrent users. That means it needs 2× L40S 48GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 2×
55.3 GB needed / 96 GB available
Decode speed
111.0
tok/s per stream
Aggregate
2664
tok/s @ 8 concurrent
Cost / 1M tokens
$0.1460
2× L40S 48GB

L40S 48GB has no NVLink, so 2-way tensor parallelism runs over PCIe. tokcalc applies a flat 0.85× multi-GPU efficiency factor for this, which is realistic for NVLink-equipped datacenter cards but optimistic for PCIe scale-out — treat the throughput and cost figures for this configuration as an upper bound, and measure before committing.

What changes as you add L40S 48GBs
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× L40S48 GB55.3 GB0.0000.000——✗
2× L40S96 GB55.3 GB111.026642.35 s$0.1460✓
4× L40S192 GB55.3 GB222.0106571.19 s$0.0730✓
8× L40S384 GB55.3 GB444.142629609.1 ms$0.0365✓
Concurrency vs context length
On 2× L40S 48GB 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?
4K91549.8 MB✓ fits
8K451.07 GB✓ fits
33K114.29 GB✓ fits
66K58.59 GB✓ fits
131K217.18 GB✓ fits
Other GPUs that run Mixtral 8x7B (MoE)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB2×5526$0.0503
AMD MI300X 192GB1×4807$0.1445
AMD MI325X 288GB1×5442$0.1787
A100 80GB SXM41×1814$0.1838
Intel Gaudi 3 128GB1×2993$0.1856
H100 SXM5 80GB1×3038$0.2286
Frequently asked questions

Can Mixtral 8x7B (MoE) run on L40S 48GB?

Yes, but it needs 2× L40S 48GB. At FP8 (E4M3) the total footprint is 55.3 GB (46.7 GB of weights plus 8.6 GB of KV cache at 8K × 8 concurrent users), against 96 GB of available VRAM across 2 GPUs.

How fast is Mixtral 8x7B (MoE) on L40S 48GB?

111.0 tokens/sec per stream and 2664 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 2.35 s, and inter-token latency is 9.0 ms.

What does Mixtral 8x7B (MoE) on L40S 48GB cost per million tokens?

$0.1460 per 1M output tokens at 2× L40S 48GB ($0.7000/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Mixtral 8x7B (MoE) need on L40S 48GB?

46.7 GB for the weights at FP8 (E4M3), plus 8.6 GB of KV cache for 8 concurrent requests at 8K context — 55.3 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is L40S 48GB the right choice for Mixtral 8x7B (MoE)?

It works. Compare $0.1460 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 3080 10GB at $0.0119, and the fastest is 2× Cerebras CS-3 (WSE-3) at 61674 tok/s.

Adjust the workload yourself

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

Open in calculator