Pinned post

As inference splits into prefill and decode, Nvidia's Groq deal could enable a "Rubin SRAM" variant optimized for ultra-low latency agentic reasoning workloads (Gavin Baker/@gavinsbaker)

Gavin Baker / @gavinsbaker : As inference splits into prefill and decode, Nvidia's Groq deal could enable a “Rubin SRAM” variant optim...

27 January 2025

How DeepSeek outpaced OpenAI at 3% of the cost: open-source approach, pure reinforcement learning, not supervised fine-tuning, and building on DeepSeek-R1-Zero (Matt Marshall/VentureBeat)

Matt Marshall / VentureBeat:
How DeepSeek outpaced OpenAI at 3% of the cost: open-source approach, pure reinforcement learning, not supervised fine-tuning, and building on DeepSeek-R1-Zero  —  DeepSeek R1's Monday release has sent shockwaves through the AI community, disrupting assumptions about what's required to achieve cutting-edge AI performance.

Posted from: this blog via Microsoft Power Automate.

Daily Deals

By on