Storia in 2 fonti

A startup claims it broke through a bottleneck that’s holding back LLMs

Subquadratic has now shared more details about its new model. But some are still skeptical.

Raccontata da

technologyreview.com

thenextweb.com

Confronto fonti

2 prospettive sulla stessa storia

AI · summaries

technologyreview.comStai leggendo1 mese fa

A startup claims it broke through a bottleneck that’s holding back LLMs

Subquadratic released SubQ using sparse attention to process 12x more text at lower cost, validated by third-party tests. For tech leaders: could end the transformer era and cut inference costs 30-50%, reshaping enterprise LLM decisions.

originale

thenextweb.com1 mese fa

A startup says it cracked the bottleneck holding back AI

Subquadratic's sparse attention cuts long-context costs 99% ($8 vs $2,600) and runs 56x faster, per Appen validation. Efficient models enable enterprise AI agent deployments with drastically lower infrastructure costs and energy consumption.

Leggi questa versione →

Timeline cronologica

venerdì 19 giugno 2026·technologyreview.com
A startup claims it broke through a bottleneck that’s holding back LLMs
Subquadratic has now shared more details about its new model. But some are still skeptical.
venerdì 19 giugno 2026·thenextweb.com
A startup says it cracked the bottleneck holding back AI
Miami startup Subquadratic claims its SubQ model breaks the 'quadratic attention' bottleneck. Independent tests back much of it, but doubts remain.