💻 API Update
🔹 V3.2: Same usage pattern as V3.2-Exp.
🔹 V3.2-Speciale: Served via a temporary endpoint: base_url="https://t.co/VfiTuW4Lbm". Same pricing as V3.2, no tool calls, available until Dec 15th, 2025, 15:59 (UTC Time).
💡 V3.2 now supports Thinking in Tool-Use — details: https://t.co/MUPiwkDI8T
4/n
@HardwareUnboxed@maroon420@TeflonGr The dont bother to allow game run on older hardwa re is good enough indication for not pre-optimization intend or not
@HardwareUnboxed Nvidia and has been doing this same thing for years. They don't give day one updates to GTX 1000 or 900 even if supported, same with the rtx 2000 and 3000. Just look at bf6 using 3/4 of the tdp on pascal. They don't optimize games day one for older archs and never did.
@__tinygrad__@mikestaub Just curious. How scalable with this kind of solution regarding training? how many GPU is usable regarding training llm and diffuser
Fun fact: If you see 9070 XT's sold out shortly after release, it will mean retailers will have sold more 9070 XT's than all GeForce 50 series GPUs combined.
(this includes RTX 5070 stock)
Wisdom is learning the lessons we thought we already knew. DeepSeek reminds us of three important learnings from computing history:
1) Computing obeys the gas law. Making it dramatically cheaper will expand the market for it. The markets are getting it wrong, this will make AI much more broadly deployed.
2) Engineering is about constraints. The Chinese engineers had limited resources, and they had to find creative solutions.
3) Open Wins. DeepSeek will help reset the increasingly closed world of foundational AI model work. Thank you DeepSeek team.