AMD and Cerebras Pair Helios With Wafer-Scale Engine for 5x Faster Inference
AMD and Cerebras pair Helios rack-scale GPUs with the Wafer-Scale Engine to deliver the fastest production inference on trillion-parameter models at 5x greater efficiency
- AMD + Cerebras partnership: The two companies announced a disaggregated inference solution pairing AMD Helios rack-scale GPUs with the Cerebras Wafer-Scale Engine.
- 5x efficiency gain: The joint solution claims up to 5x higher tokens per second per watt versus a Cerebras-only setup, on 1T+ parameter models.
- Helios specs: AMD's new Helios rack packs 72 MI455X GPUs, 31 TB HBM4, and 2.9 exaFLOPS FP4 in a single $5-5.5M rack.
- Architecture split: Helios handles compute-heavy prefill (prompt processing); the Cerebras WSE handles ultra-low-latency token decode, each doing what it does best.
- Launch timeline: Available first through Cerebras Inference Cloud in H2 2026; broader availability and Helios mass production expected Q2 2027.
- Competitive context: Directly counters Nvidia's $20B Groq acquisition; Cerebras also has a similar disaggregated deal with AWS Trainium for Amazon Bedrock.
AMD and Cerebras have announced a technical partnership to build a disaggregated inference solution pairing AMD's new Helios rack-scale system with the Cerebras Wafer-Scale Engine. Cerebras CEO Andrew Feldman joined AMD CEO Lisa Su on stage at AMD's Advancing AI 2026 event to make the announcement. The pitch: the fastest production inference tokens in the industry, at 5x greater capacity, on frontier-scale models with over one trillion parameters.
Two chips, one pipeline
LLM inference has two phases with very different hardware needs:
- Prefill: Processing the input prompt and any context. Compute-heavy, parallelizes well across many GPUs.
- Decode: Generating tokens one by one. Memory-bandwidth-bound, benefits enormously from on-chip SRAM rather than off-chip HBM.
Running both phases on the same GPU cluster forces a compromise. GPUs must shuttle the entire model from HBM memory to compute for every token generated. Because the full model cannot fit on the GPU die itself, this process repeats with every token, making GPU inference persistently bandwidth-bound.
Cerebras sidesteps this entirely. The WSE-3 is a wafer-scale chip with 900,000 cores and 44 GB of on-die SRAM connected at 21 PB/s bandwidth. By putting all memory on the wafer itself, it eliminates the HBM memory wall and achieves sub-10ms latency per token at throughput levels where GPU clusters typically stall at 10–12ms.
What Helios brings
AMD's Helios is a single rack containing 72 Instinct MI455X GPUs, 31 terabytes of HBM4 memory, and 2.9 exaflops of FP4 inference compute. It is AMD's first rack-scale AI system and its direct answer to Nvidia's Vera Rubin NVL72. Full specs:
- 1.4 exaFLOPS of FP8 compute, 2.9 exaFLOPS of FP4 compute, and 31 TB of HBM4 memory
- 260 TB/s aggregate intra-rack bandwidth via UALink-over-Ethernet and a co-designed Ethernet switch