[edwingao28]

Adds seven News entries to README.md and README_zh.md covering September–November 2026:

• Kimi K3.1 Pro Day 0 benchmarks ([K3.1-Pro] Day-0 support SGLang on B200 #4801, [K3.1-Pro] Day-0 support SGLang on B300 #4803, [K3.1-Pro] Day-0 support vLLM on B300 #4796, [K3.1-Pro] Day-0 support SGLang on GB200 #4804, [K3.1-Pro] Day-0 support SGLang on MI355X #4808)
• PowerX default multi-node power collection ([PowerX] collect multinode power by default / [PowerX] 默认采集多节点功耗 #3670, [PowerX] collect NVL72 CPU and per-host GPU power / [PowerX] 采集 NVL72 CPU 与逐主机 GPU 功耗 #3296, [PowerX] collect AMD x86 CPU package power through RAPL / [PowerX] 通过 RAPL 采集 AMD x86 CPU 封装功耗 #4154)
• GLM-5.3-Flash AgentX on GB200, MI300X, MI325X (feat(bench): add GLM-5.3-Flash FP8 and NVFP4 AgentX on GB200 / feat(bench)：新增 GB200 GLM-5.3-Flash FP8 与 NVFP4 AgentX #4718, Add GLM-5.3-Flash native FP8 MI300X AgentX with HiCache / 新增 GLM-5.3-Flash 原生 FP8 MI300X AgentX（含 HiCache） #4721, Add GLM-5.3-Flash FP8 MI325X AgentX with HiCache / 新增 GLM-5.3-Flash FP8 MI325X AgentX（含 HiCache） #4722)
• MiniMax M3.1 Flash Day 0 AgentX with DSpark ([Klaud Cold] Add day-0 MiniMax-M3.1-Flash B200 vLLM AgentX DSpark recipe / [Klaud Cold] 新增 MiniMax-M3.1-Flash B200 vLLM AgentX DSpark 首日配方 #4426)
• Muse Spark Open AgentX on B300/GB300
• Open-source power model beta (Initial Open Source of Power Model (Beta - Experimental) #3739)
• Vera Rubin NVL72, TPUv7x Ironwood, RTX PRO 6000 Server support ([Klaud Cold] Mark Vera Rubin NVL72, TPUv7x and RTX PRO 6000 Server supported / [Klaud Cold] 将 Vera Rubin NVL72、TPUv7x 与 RTX PRO 6000 Server 标记为已支持 #3510)

Docs-only change.

[claude]

Code review found no issues

No high-confidence issues detected in this change.