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

Llama 3.3 70B on Intel Gaudi 3 128GB: 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 Intel Gaudi 3 128GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
92.1 GB needed / 128 GB available
Decode speed
45.57
tok/s per stream
Aggregate
546.9
tok/s @ 8 concurrent
Cost / 1M tokens
$1.02
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 GB92.1 GB45.57546.91.31 s$1.02✓
2× Intel256 GB92.1 GB77.481859673.3 ms$0.5976✓
4× Intel512 GB92.1 GB155.07438351.6 ms$0.2988✓
8× Intel1024 GB92.1 GB309.929750190.8 ms$0.1494✓
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?
4K421.34 GB✓ fits
8K212.68 GB✓ fits
33K510.74 GB✓ fits
66K221.47 GB✓ fits
131K142.95 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 Intel Gaudi 3 128GB?

Yes. Llama 3.3 70B at FP8 (E4M3) needs 92.1 GB including the KV cache, and Intel Gaudi 3 128GB has 128 GB — so it fits on a single GPU.

How fast is Llama 3.3 70B on Intel Gaudi 3 128GB?

45.57 tokens/sec per stream and 546.9 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 1.31 s, and inter-token latency is 21.9 ms.

What does Llama 3.3 70B on Intel Gaudi 3 128GB cost per million tokens?

$1.02 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 Llama 3.3 70B need on Intel Gaudi 3 128GB?

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 Intel Gaudi 3 128GB the right choice for Llama 3.3 70B?

It works. Compare $1.02 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