OngoingOnline · Tech · advanced · ai, optimization, machine-learning
A global online competition focused on optimizing inference throughput for the MiniCPM5-2B large language model using distributed systems and low-level optimizations. Teams compete to improve latency and memory efficiency without sacrificing accuracy, utilizing frameworks like Fl
OngoingOnline · Tech · advanced · ai, llm, optimization
A global team-based competition to optimize SGLang AI inference kernels for multiple chips using Triton or Triton-TLE. Participants tackle over 200 performance challenges to improve LLM throughput and latency, competing for awards based on optimization breakthroughs and extreme p