Model Releases
Our GB300 cluster went down yesterday, just as Deepseek released 😱 We were 😥 but @CoreWeave came through to contribute to the Open Source.…
Our GB300 cluster went down yesterday, just as Deepseek released 😱 We were 😥 but @CoreWeave came through to contribute to the Open Source. They scrambled in the compute crisis, finding 2 spare dev rac
Our GB300 cluster went down yesterday, just as Deepseek released 😱 We were 😥 but @CoreWeave came through to contribute to the Open Source. They scrambled in the compute crisis, finding 2 spare dev racks of GB300 Our team is running Deepseekv4 now for InferenceX benchmarks! SAN FRANCISCO, April 24, 2026 — InferenceX has added Day 0 DeepSeek v4 support for @sgl_project’s amazing Blackwell B300 performance, 5x faster than Hopper even on Day 0. Great work to @radixark, @NVIDIAAIDev, and @ying11231 for the first-principles engineering. Huge shoutout to …
Source: Dylan Patel (X) | 2026-04-25