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

OLMo 2 13B on Intel Gaudi 3 128GB: Does It Fit, How Fast, What It Costs

OLMo 2 13B at FP16 needs 26.0 GB for its weights and 36.7 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
36.7 GB needed / 128 GB available
Decode speed
82.50
tok/s per stream
Aggregate
990.0
tok/s @ 8 concurrent
Cost / 1M tokens
$0.5612
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 GB36.7 GB82.50990.0274.3 ms$0.5612✓
2× Intel256 GB36.7 GB140.33366153.2 ms$0.3301✓
4× Intel512 GB36.7 GB280.51346491.6 ms$0.1650✓
8× Intel1024 GB36.7 GB561.05385660.8 ms$0.0825✓
Concurrency vs context length
On 1× Intel Gaudi 3 128GB at FP16. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K151687.2 MB✓ fits
8K751.34 GB✓ fits
33K185.37 GB✓ fits
66K910.74 GB✓ fits
131K421.47 GB✓ fits
Other GPUs that run OLMo 2 13B
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB2×1828$0.1520
AMD MI300X 192GB1×1590$0.4368
A100 40GB SXM41×465.0$0.5376
AMD MI325X 288GB1×1800$0.5401
A100 80GB SXM41×600.0$0.5556
H100 SXM5 80GB1×1005$0.6910
Frequently asked questions

Can OLMo 2 13B run on Intel Gaudi 3 128GB?

Yes. OLMo 2 13B at FP16 needs 36.7 GB including the KV cache, and Intel Gaudi 3 128GB has 128 GB — so it fits on a single GPU.

How fast is OLMo 2 13B on Intel Gaudi 3 128GB?

82.50 tokens/sec per stream and 990.0 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 274.3 ms, and inter-token latency is 12.1 ms.

What does OLMo 2 13B on Intel Gaudi 3 128GB cost per million tokens?

$0.5612 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 OLMo 2 13B need on Intel Gaudi 3 128GB?

26.0 GB for the weights at FP16, plus 10.7 GB of KV cache for 8 concurrent requests at 8K context — 36.7 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is Intel Gaudi 3 128GB the right choice for OLMo 2 13B?

It works. Compare $0.5612 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 4× RTX 3080 10GB at $0.0717, and the fastest is 1× Cerebras CS-3 (WSE-3) at 6000 tok/s.

Adjust the workload yourself

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

Open in calculator