OpenAI
OpenAI and Broadcom Built Jalapeño to Make Inference Spicy and NVIDIA Nervous
OpenAI and Broadcom unveiled Jalapeño, a custom LLM inference chip aimed at lower-cost AI. The timing makes the NVIDIA moat story extra awkward.
This Week in Snark
This Week in Snark: From Kiss Cam Collapses to Color-Picking Clouds
A snark-filled roundup of the week's wildest tech news, featuring Astronomer’s scandal, AI paint picks, spider silk deals, Ethernet hype, and more.
Launch
Broadcom Launches Tomahawk Ultra, Promises Ethernet That’s So Fast It Might Time Travel
Broadcom’s Tomahawk Ultra switch promises Ethernet so fast and lossless it might just resurrect your AI model—and your faith in networking.