Google95 GB VRAM819 GB/s$4.2/hr

Llama 4 Scout 17B-16E (MoE, VLM) on TPU v5p (per-chip, 95GB): Does It Fit, How Fast, What It Costs

Llama 4 Scout 17B-16E (MoE, VLM) at AWQ 4-bit needs 60.0 GB for its weights and 72.8 GB in total at 8K context with 8 concurrent users. That means it fits on a single TPU v5p (per-chip, 95GB). Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
72.8 GB needed / 95 GB available
Decode speed
50.10
tok/s per stream
Aggregate
601.2
tok/s @ 8 concurrent
Cost / 1M tokens
$1.94
1× TPU v5p (per-chip, 95GB)
What changes as you add TPU v5p (per-chip, 95GB)s
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× TPU95 GB72.8 GB50.10601.21.26 s$1.94✓
2× TPU190 GB72.8 GB85.182044648.6 ms$1.14✓
4× TPU380 GB72.8 GB170.48177339.3 ms$0.5707✓
8× TPU760 GB72.8 GB340.732708184.6 ms$0.2854✓
Concurrency vs context length
On 1× TPU v5p (per-chip, 95GB) at AWQ 4-bit. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K43824.6 MB✓ fits
8K211.61 GB✓ fits
33K56.44 GB✓ fits
66K212.88 GB✓ fits
131K125.77 GB✓ fits
Other GPUs that run Llama 4 Scout 17B-16E (MoE, VLM)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB4×17891$0.0311
AMD MI300X 192GB1×3891$0.1785
L40S 48GB2×2157$0.1803
AMD MI325X 288GB1×4405$0.2207
A100 80GB SXM41×1468$0.2270
Intel Gaudi 3 128GB1×2423$0.2293
Frequently asked questions

Can Llama 4 Scout 17B-16E (MoE, VLM) run on TPU v5p (per-chip, 95GB)?

Yes. Llama 4 Scout 17B-16E (MoE, VLM) at AWQ 4-bit needs 72.8 GB including the KV cache, and TPU v5p (per-chip, 95GB) has 95 GB — so it fits on a single GPU.

How fast is Llama 4 Scout 17B-16E (MoE, VLM) on TPU v5p (per-chip, 95GB)?

50.10 tokens/sec per stream and 601.2 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 1.26 s, and inter-token latency is 20.0 ms.

What does Llama 4 Scout 17B-16E (MoE, VLM) on TPU v5p (per-chip, 95GB) cost per million tokens?

$1.94 per 1M output tokens at 1× TPU v5p (per-chip, 95GB) ($4.20/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Llama 4 Scout 17B-16E (MoE, VLM) need on TPU v5p (per-chip, 95GB)?

60.0 GB for the weights at AWQ 4-bit, plus 12.9 GB of KV cache for 8 concurrent requests at 8K context — 72.8 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is TPU v5p (per-chip, 95GB) the right choice for Llama 4 Scout 17B-16E (MoE, VLM)?

It works. Compare $1.94 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 3080 10GB at $0.0146, and the fastest is 2× Cerebras CS-3 (WSE-3) at 49920 tok/s.

Adjust the workload yourself

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

Open in calculator