AI 导读
5.6-sol 重写了那个解决了自身问题的生产内核,最终有效带来: 🟠 GPU kernel 改进使服务成本降低 20% 🟠 改进的投机解码使 token 生成效率提升 15% 🟠 节省的成本传导至 Luna/Terra 降价 (2/2)
正文
5.6-sol rewrote the production kernel that solved itself which effectively resulted in:
🟠 20% lower serving costs from GPU kernel improvements
🟠 15% better token-generation efficiency from improved speculative decoding
🟠 Savings passed into Luna/Terra price cuts (2/2)
来源:@SemiAnalysis_ · x.com