we_are_coded.by CODE · The world, decoded
БГ

All sections

News

507 stories · page 35 of 57

FrontierOpenAI · 25 August 2026Jalapeño's first numbers: OpenAI claims its chip gives more and faster at onceOpenAI published the first measurements of its own inference chip. They claim higher throughput and shorter waiting at the same time, which until now was treated as a trade. The numbers are theirs and were measured by them.Read
ControlOpenAI · 25 August 2026OpenAI shut down a Russian network building a think tank that did not existThe accounts wrote comments across five social networks to push the website of a self-described expert community. The site itself stood on other people's research papers, copied and credited to the wrong authors.DesignComfyUI · 25 August 2026Wan 3.0 makes 30 seconds of video in a single pass, with no stitchingAlibaba's new video model is live in ComfyUI. It gives up to 30 seconds at once, up to twenty reference assets, and a way to address each of them by name inside the prompt itself.SecurityMeta Newsroom · 25 August 2026WhatsApp replaced the six-digit PIN with a real passwordThe second step at login can now be a long password with letters and symbols instead of six digits. Android also gets more information about calls from unknown numbers, and passkeys are no longer limited to one.InfraMeta Engineering · 24 August 2026Meta unveils MTIA 300, its first chip built to train modelsMTIA 300 is Meta's first accelerator built to train models, not just serve answers from finished ones. It targets recommendation and ranking systems, with the network built straight into the silicon. By Meta's own numbers, total communication time drops 3.9x versus an equivalent GPU cluster.InfraMeta Engineering · 24 August 2026Meta wrote its own AI network protocol and opened it through the Open Compute ProjectMetaRoCE is a transport protocol written from scratch for the conversations between thousands of AI accelerators over plain Ethernet. By the company's own numbers, it keeps about 86 percent of throughput even at one percent packet loss. The spec, the reference code and the compliance test are going public.InfraNVIDIA Blog · 24 August 2026NVIDIA reports 30x more agentic throughput per megawatt, but the number still awaits independent reviewNVIDIA published its own measurements for Vera Rubin NVL72: 30 times more agentic throughput per megawatt and 35 times lower cost per token versus the previous generation. The numbers are theirs, on a workload they ran themselves, and by the company's own words still await review from SemiAnalysis.InfraNVIDIA Blog · 24 August 2026Groq 3 LPX enters production, built together with NVIDIA's platformA rack-level inference accelerator, co-designed with Vera Rubin NVL72. It combines GPUs with dedicated fast-response modules and handles token generation in agentic systems. Already deploying at Nebius, CoreWeave and SpaceXAI.BuildersJetBrains Blog · 24 August 2026JetBrains' Junie now runs entirely on your Mac, no cloud, no tokensJetBrains' code agent now runs locally. No credits, no quota, the code never leaves the machine. The price is the hardware: an M5 chip and at least 64 gigabytes of memory.
507 stories

Sections