Intel128 GB VRAM3,300 GB/s$2/hr

Mixtral 8x7B (MoE) on Intel Gaudi 3 128GB: 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 fits on a single Intel Gaudi 3 128GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
55.3 GB needed / 128 GB available
Decode speed
249.4
tok/s per stream
Aggregate
2993
tok/s @ 8 concurrent
Cost / 1M tokens
$0.1856
1× Intel Gaudi 3 128GB
What changes as you add Intel Gaudi 3 128GBs
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× Intel128 GB55.3 GB249.42993264.4 ms$0.1856✓
2× Intel256 GB55.3 GB424.010176147.5 ms$0.1092✓
4× Intel512 GB55.3 GB848.04070588.8 ms$0.0546✓
8× Intel1024 GB55.3 GB169616282059.4 ms$0.0273✓
Concurrency vs context length
On 1× Intel Gaudi 3 128GB 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?
4K151549.8 MB✓ fits
8K751.07 GB✓ fits
33K184.29 GB✓ fits
66K98.59 GB✓ fits
131K417.18 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
L40S 48GB2×2664$0.1460
AMD MI325X 288GB1×5442$0.1787
A100 80GB SXM41×1814$0.1838
H100 SXM5 80GB1×3038$0.2286
Frequently asked questions

Can Mixtral 8x7B (MoE) run on Intel Gaudi 3 128GB?

Yes. Mixtral 8x7B (MoE) at FP8 (E4M3) needs 55.3 GB including the KV cache, and Intel Gaudi 3 128GB has 128 GB — so it fits on a single GPU.

How fast is Mixtral 8x7B (MoE) on Intel Gaudi 3 128GB?

249.4 tokens/sec per stream and 2993 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 264.4 ms, and inter-token latency is 4.0 ms.

What does Mixtral 8x7B (MoE) on Intel Gaudi 3 128GB cost per million tokens?

$0.1856 per 1M output tokens at 1× Intel Gaudi 3 128GB ($2.00/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 Intel Gaudi 3 128GB?

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 Intel Gaudi 3 128GB the right choice for Mixtral 8x7B (MoE)?

It works. Compare $0.1856 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