Google32 GB VRAM819 GB/s$0.8/hr

GPT-OSS 20B (MoE) on TPU v6e Trillium (32GB): Does It Fit, How Fast, What It Costs

GPT-OSS 20B (MoE) at FP8 (E4M3) needs 20.9 GB for its weights and 22.6 GB in total at 8K context with 8 concurrent users. That means it fits on a single TPU v6e Trillium (32GB). Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
22.6 GB needed / 32 GB available
Decode speed
221.8
tok/s per stream
Aggregate
2662
tok/s @ 8 concurrent
Cost / 1M tokens
$0.0835
1× TPU v6e Trillium (32GB)
What changes as you add TPU v6e Trillium (32GB)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× TPU32 GB22.6 GB221.82662163.0 ms$0.0835✓
2× TPU64 GB22.6 GB377.1905096.9 ms$0.0491✓
4× TPU128 GB22.6 GB754.23620063.5 ms$0.0246✓
8× TPU256 GB22.6 GB150814479946.7 ms$0.0123✓
Concurrency vs context length
On 1× TPU v6e Trillium (32GB) 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?
4K106206.2 MB✓ fits
8K54412.3 MB✓ fits
33K131.61 GB✓ fits
66K63.22 GB✓ fits
131K36.44 GB✓ fits
Other GPUs that run GPT-OSS 20B (MoE)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB1×5824$0.0238
RTX 3090 24GB1×2472$0.0281
RTX 4090 24GB1×3276$0.0339
AMD MI300X 192GB1×17225$0.0403
V100 SXM2 32GB1×2925$0.0475
A100 40GB SXM41×5038$0.0496
Frequently asked questions

Can GPT-OSS 20B (MoE) run on TPU v6e Trillium (32GB)?

Yes. GPT-OSS 20B (MoE) at FP8 (E4M3) needs 22.6 GB including the KV cache, and TPU v6e Trillium (32GB) has 32 GB — so it fits on a single GPU.

How fast is GPT-OSS 20B (MoE) on TPU v6e Trillium (32GB)?

221.8 tokens/sec per stream and 2662 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 163.0 ms, and inter-token latency is 4.5 ms.

What does GPT-OSS 20B (MoE) on TPU v6e Trillium (32GB) cost per million tokens?

$0.0835 per 1M output tokens at 1× TPU v6e Trillium (32GB) ($0.8000/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does GPT-OSS 20B (MoE) need on TPU v6e Trillium (32GB)?

20.9 GB for the weights at FP8 (E4M3), plus 1.6 GB of KV cache for 8 concurrent requests at 8K context — 22.6 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is TPU v6e Trillium (32GB) the right choice for GPT-OSS 20B (MoE)?

It works. Compare $0.0835 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 4× RTX 3080 10GB at $0.006615, and the fastest is 1× Cerebras CS-3 (WSE-3) at 65000 tok/s.

Adjust the workload yourself

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

Open in calculator