Tag: Inference

1. Epoch measures why long context feels slower on GPT than Claude (epoch.ai)
   AI · 2 days ago · September 11, 2026
2. Running a 2.8-trillion-parameter model on a laptop by streaming from SSDs (github.com)
   AI · 4 days ago · September 9, 2026
3. Mercury 2.5 pushes diffusion language models toward production use (inceptionlabs.ai)
   AI · 4 days ago · September 9, 2026
4. A 4-Bit Model That Beats Its Full-Precision Original (huggingface.co)
   AI · 5 days ago · September 8, 2026
5. Cheap, fast models make a whole class of AI products viable (calv.info)
   AI · 2 weeks ago · August 28, 2026
6. OpenAI and Cerebras push GPT-5.6 Sol to 750 tokens per second (cerebras.ai)
   AI · 1 month ago · August 14, 2026
7. AMD buys Taalas, which etches models straight into silicon (theregister.com)
   AI · 1 month ago · August 7, 2026
8. How Cloudflare serves Kimi and GLM at scale with FP8 and INT4 (blog.cloudflare.com)
   AI · 1 month ago · August 4, 2026
9. OpenAI cuts GPT-5.6 Luna 80% after a model rewrote its own inference code (openai.com)
   AI · 1 month ago · August 1, 2026
10. Why model routing is an optimization problem, not a classifier (huggingface.co)
    AI · 1 month ago · July 17, 2026
11. A proxy that turns your prompt into a PNG to cut token costs (github.com)
    Engineering · 2 months ago · July 4, 2026
12. A startup bets oscillator chips can cut AI inference power 1,000x (techcrunch.com)
    AI · 2 months ago · June 28, 2026
13. OpenAI shows its first custom chip, Jalapeño, built for inference (techcrunch.com)
    AI · 2 months ago · June 25, 2026
14. Moebius: a 0.2B inpainting model that keeps up with 12B ones (hustvl.github.io)
    AI · 2 months ago · June 23, 2026
15. DiffusionGemma trades quality for speed in open text generation (blog.google)
    AI · 3 months ago · June 11, 2026
16. NVIDIA's open diffusion LLMs run six times faster than autoregressive baselines (huggingface.co)
    AI · 3 months ago · May 23, 2026
17. Modal cuts serverless GPU cold starts by 40x (modal.com)
    Engineering · 3 months ago · May 19, 2026
18. Hugging Face cut 22 percent off generation time by overlapping CPU and GPU (huggingface.co)
    Engineering · 4 months ago · May 14, 2026
19. Fix the inference engine before you patch the RL objective (huggingface.co)
    Engineering · 4 months ago · May 6, 2026

← all tags