Storia in 4 fonti

KI-Inferencing: US-deutsches Start-up will Nvidia ausstechen

Dank logarithmischer Mathematik soll der 3-Nanometer-Chip Tensordyne Napier KI wesentlich schneller und effizienter berechnen als Nvidia-Beschleuniger.

Raccontata da

Confronto fonti

4 prospettive sulla stessa storia

AI · summaries

heise.deStai leggendo2 g fa

KI-Inferencing: US-deutsches Start-up will Nvidia ausstechen

Tensordyne meldet Tape-Out des KI-Chips Napier mit 13x besseren Durchsätzen und 17x besserer Effizienz als Nvidia GB300; 200 Millionen Dollar Vorbestellungen. CTOs müssen ihre Inference-Architektur überdenken: die Kosten pro Token könnten sich fundamental verändern.

originale

forbes.com3 g fa

Tensordyne Revives Logarithmic Math In A Bid To Cut AI Power Use

Tensordyne says logarithmic computing could reduce AI inference costs and power demands, offering an alternative to conventional chip designs.

Leggi questa versione → originale

cryptobriefing.com13 h fa

Tensordyne claims 13x throughput improvement over Nvidia's NVL72 GB300 with new TDN AI Processor

Tensordyne's Napier TDN processor claims 13x throughput and 17x efficiency over Nvidia's GB300 NVL72 rack using logarithmic math on TSMC's 3nm process.

Leggi questa versione → originale

spectrum.ieee.org2 g fa

Tensordyne's Wild Log Math Aims to Leave Nvidia’s AI Chips In the Dust

The startup uses logarithmic math to speed up inference

Leggi questa versione → originale

Timeline cronologica

lunedì 15 giugno 2026·forbes.com
Tensordyne Revives Logarithmic Math In A Bid To Cut AI Power Use
Tensordyne says logarithmic computing could reduce AI inference costs and power demands, offering an alternative to conventional chip designs.
lunedì 15 giugno 2026·heise.de
KI-Inferencing: US-deutsches Start-up will Nvidia ausstechen
Dank logarithmischer Mathematik soll der 3-Nanometer-Chip Tensordyne Napier KI wesentlich schneller und effizienter berechnen als Nvidia-Beschleuniger.
lunedì 15 giugno 2026·spectrum.ieee.org
Tensordyne's Wild Log Math Aims to Leave Nvidia’s AI Chips In the Dust
The startup uses logarithmic math to speed up inference
mercoledì 17 giugno 2026·cryptobriefing.com
Tensordyne claims 13x throughput improvement over Nvidia's NVL72 GB300 with new TDN AI Processor
Tensordyne's Napier TDN processor claims 13x throughput and 17x efficiency over Nvidia's GB300 NVL72 rack using logarithmic math on TSMC's 3nm process.