Cerebras44 GB VRAM20,000 GB/s

Qwen3.8 27B (VLM) on Cerebras CS-3 (WSE-3): Does It Fit, How Fast, What It Costs

Qwen3.8 27B (VLM) at FP8 (E4M3) needs 27.8 GB for its weights and 33.3 GB in total at 8K context with 8 concurrent users. That means it fits on a single Cerebras CS-3 (WSE-3). Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
33.3 GB needed / 44 GB available
Decode speed
701.9
tok/s per stream
Aggregate
8423
tok/s @ 8 concurrent
Cost / 1M tokens
—
1× Cerebras CS-3 (WSE-3) · quote required
What changes as you add Cerebras CS-3 (WSE-3)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× Cerebras44 GB33.3 GB701.98423759.7 ms—✓
2× Cerebras88 GB33.3 GB119328638395.0 ms—✓
4× Cerebras176 GB33.3 GB2387114553212.5 ms—✓
8× Cerebras352 GB33.3 GB4773458212121.2 ms—✓
Concurrency vs context length
On 1× Cerebras CS-3 (WSE-3) 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?
4K381.07 GB✓ fits
8K232.15 GB✓ fits
33K78.59 GB✓ fits
66K317.18 GB✓ fits
131K134.36 GB✓ fits
Other GPUs that run Qwen3.8 27B (VLM)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB2×2566$0.1083
AMD MI300X 192GB1×2232$0.3111
A100 40GB SXM41×652.8$0.3830
AMD MI325X 288GB1×2527$0.3847
A100 80GB SXM41×842.3$0.3957
Intel Gaudi 3 128GB1×1390$0.3997
Frequently asked questions

Can Qwen3.8 27B (VLM) run on Cerebras CS-3 (WSE-3)?

Yes. Qwen3.8 27B (VLM) at FP8 (E4M3) needs 33.3 GB including the KV cache, and Cerebras CS-3 (WSE-3) has 44 GB — so it fits on a single GPU.

How fast is Qwen3.8 27B (VLM) on Cerebras CS-3 (WSE-3)?

701.9 tokens/sec per stream and 8423 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 759.7 ms, and inter-token latency is 1.4 ms.

What does Qwen3.8 27B (VLM) on Cerebras CS-3 (WSE-3) cost per million tokens?

— per 1M output tokens at 1× Cerebras CS-3 (WSE-3). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Qwen3.8 27B (VLM) need on Cerebras CS-3 (WSE-3)?

27.8 GB for the weights at FP8 (E4M3), plus 5.5 GB of KV cache for 8 concurrent requests at 8K context — 33.3 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is Cerebras CS-3 (WSE-3) the right choice for Qwen3.8 27B (VLM)?

It works. Compare — per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 4× RTX 3080 10GB at $0.0511, and the fastest is 1× Cerebras CS-3 (WSE-3) at 8423 tok/s.

Adjust the workload yourself

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

Open in calculator