Yet Another Chip Company Claiming Fantastical AI Results - Just Taped Out

[KevinK]

Tensordyne Tapes Out LNS-Based AI Chip, Claims Huge Power Advantages​

AI chip startup Tensordyne has taped out its data center inference chip, which the company said will offer an order-of-magnitude improvement in power efficiency compared with leading GPU alternatives. The company said systems based on its chip can achieve 17× the tokens per second per Watt versus an Nvidia GB300-based system for the same workload, or 13× the tokens per second per rack (see editor’s note at the end of this article).

Faster and lower power for AI Inference in data-centers than anything in the known universe, all based on simulated benchmark results. I'm going to wait on InferenceX and Artificial Intelligence (AA-AgentPerf) benchmark results on real rack-level hardware before believing.

https://www.eetimes.com/tensordyne-tapes-out-lns-based-ai-chip-claims-huge-power-advantages/