ArticleJulie BortNews
Inference computation forms the largest ongoing operational cost for deploying advanced AI models. Tangible data centre adoption of alternative architectures by institutional players highlights viable competition against established incumbents like Nvidia.
Etched targets the dual bottleneck of inference by separating prompt context processing from token output generation. Its low-voltage chip avoids typical thermal limits during compute-heavy prefill, while the proprietary interconnect and cluster-scale memory reduce latency during decode operations. Validated by Jane Street's purchase and deployment, Etched demonstrates that purpose-built inference hardware can offer flexibility across frontier models without being locked to a single architecture.
Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.
Similar in content, found by the AI search.
Link:Etched
AI hardware startup Etched has introduced dedicated inference clusters designed specifically for frontier models. Following the delivery of A0 silicon from TSMC N4P, the company announced $800M in raised funding and $1B in customer demand.
Link:ricursive.com
Frontier AI lab Ricursive Intelligence is developing self-improving systems starting with semiconductor design, closing the feedback loop between AI models and hardware.
Article:Bryan Shan, Myron Xie, Jordan Nanos
At the Hot Chips conference, OpenAI revealed details of its custom LLM inference chip codenamed Jalapeño. Developed in partnership with Broadcom, the ASIC outperforms Nvidia Blackwell and Vera Rubin in token throughput per megawatt according to initial lab benchmarks by SemiAnalysis.
Article:Kyt Dotson
AI startup Nous Research has raised $50 million in a Series A round led by Paradigm. The company leverages the Solana blockchain to coordinate distributed AI model training across contributed consumer and enterprise GPUs.
Link:Modular
Modular presents its AI inference platform designed to execute workloads portably across hardware architectures including NVIDIA, AMD, TPUs, Trainium, and Apple silicon, combining the MAX serving framework with the Mojo language.
Link:prismml.com
PrismML has introduced the Bonsai model family, offering ultra-dense AI models based on research from Caltech. The flagship Bonsai 27B model runs locally on smartphones in a 1-bit variant requiring only 3.9 GB of memory.
Parts of this content were created with the assistance of AI.
