750 Tokens Per Second: Why Cerebras Wins the Inference Speed Race
Cerebras and OpenAI launched Ultrafast, running GPT-5.6 Sol at 750 tokens per second — 14x the standard rate. The secret is on-chip memory, not raw compute.
Read ArticleCategory: AI Engineering
Articles about AI Engineering
IBM's $240M Open Inference Bet Is the Inflection Point Enterprise AI Needed
IBM and Together AI just committed $240M to proving open-source inference can undercut the hyperscalers. Here's why this changes the math for every team running AI in production.
Read Article
One GPU, Thirty Billion Parameters: What Meta's Muse Glimmer Actually Changes
Meta shipped a 30B agent model that runs on a single consumer GPU under Apache 2.0. The benchmarks are competitive, the hardware bar is lower than you think, and the build-vs-buy math just shifted.
Read Article
Terafab: Tesla and SpaceX's $16.8B Bet That the Chip Supply Chain Is Broken
Tesla and SpaceX just committed $16.8 billion to build a chip factory in Texas. The real story isn't the money—it's what vertical integration at this scale reveals about who actually controls AI compute.
Read Article
The Model IS the Chip: AMD's Taalas Acquisition and the End of the Memory Wall
AMD acquired Taalas, which etches AI model weights directly into silicon. At 16,960 tokens/sec on Llama 3.1 8B with no HBM--this isn't optimization, it's a paradigm shift for inference.
Read Article
The 10-Hour CUDA Clone: How AI Coding Agents Are Eating Nvidia's Software Moat
A one-year-old startup built CUDA-equivalent chip software in 10 hours using AI coding agents. The implications for Nvidia's moat—and for every team running AI infrastructure—are significant.
Read Article
Feeling Faster While Going Slower: What 2026's AI Coding Data Actually Shows
Three rigorous 2026 studies reveal a striking paradox: engineers using AI feel 20% faster while measuring 19% slower, and aggregate PR throughput is up just 10% despite 93% adoption.
Read Article
The Engineers Are Asking for a Kill Switch — And I Think They're Right
On July 28, 1,178 engineers at OpenAI, Anthropic, Google, and Meta asked the US government to build an AI slowdown mechanism. When the builders raise the alarm, that's the signal practitioners can't ignore.
Read Article
Qualcomm Closes Its $3.9B Modular Bet: CUDA Lock-In Now Has a Real Challenger
Qualcomm closed its $3.9B Modular acquisition on July 29. For the first time, there's a serious, well-funded alternative to CUDA's software moat—here's the operational read.
Read Article