Cognition Becomes First CoreWeave Vera Rubin NVL72 Customer, Sees 4.8X SWE-2 Throughput Boost

altryne · x · 2026-10-01

At Fully Connected '26, Cognition announced it is the first customer on CoreWeave's NVL72 Vera Rubin infrastructure, reporting a 4.8X boost in token throughput on SWE-2. Altryne also notes connecting Devins via Linear to all his assistants is a 'crazy powerup.'

Related event: Cognition First to Adopt CoreWeave's Vera Rubin NVL72(2 posts)→

Original post →

More from coding & agent

coding & agent channel →