Bros, breaking news. Su Zifeng brought a host computer the size of a lunch box to the stage and ran a…
Bros, breaking news. Su Zifeng brought a host computer the size of a lunch box to the stage and ran a large model with 235 billion parameters on the spot. It doesn’t rely on the cloud or rents a GPU, it relies entirely on AMD Ryzen AI Max+ 395, the world’s first x86 processor with 128GB of unified memory where the CPU and GPU share unified memory. Let’s compare: the RTX 5090 only has 32GB of video memory, the RTX…
Ok so the RLM is kinda done almost.
Ngl the RLM nuances are sill lowkey bonkers. Had to read and revise some 2-3 times to get the mental model right and then some more. I hit the async/sync on single thread issue with Terminal UI event loop and the RLM engine which itself is
Top-5 Best Value #LLM Models of the Day at UTC-04
| Model | #AAII | Price | | GLM 5.3 Flash | 58 | $0.06 | | DeepSeek V4 Flash 0731 | 52 | $0.05 | | GPT-5.6 Luna | 52 | $0.18 | | Gemini 3.7 Flash | 56 | $0.57 | | GLM 5.2 | 53 | $0.55 |
POV: Monday + deadline + slow API + rising invoices 🙃
3 diseases are one medicine: ✦ GPT-5.6 LUNA — IDR 180/1M tokens ✦ GLM 5.3 Flash — Rp. 1,350/1M ✦ DeepSeek V4 Flash — Rp. 1,080/1M 1M context, vision support, pay as you go. My Monday is bright,
You have reached the end of the archive
All of deepseek