Pengzhan Zhao (@borontion) profile picture
Joined July 2026
685 output tokens/s per user on @Alibaba_Qwen Qwen3.8-Flash-Next. Running TokenSpeed on 4× NVIDIA GB300 with NVFP4. A look at the kernel and serving optimizations behind the results: 👉 lightseek.org/blog/tokenspee… ⭐ github.com/lightseekorg/toke…
14
7
11
131
206,145
Kimi K3 on AMD Instinct MI355X, powered by TokenSpeed ⚡ 216 output tok/s per user on 8× MI355X: 1.4× the decode speed of ATOM at C1, while running higher precision. Measured on a real agentic workload: multi-turn SWE-smith, 50K+ context, 10–15 turns. Technical deep dive with @AIatAMD 🧵
2
4
1
26
8,814
Pengzhan Zhao retweeted
MI450X launched last week. We’re kicking off an engineering blog series on its hardware features and how to use them from Gluon. First one is from @borontion and team on decode attention: using explicit WMMA layouts, async TDM→LDS loads, four-stage triple-buffered pipelining, and fused Split-K to deliver 17.1 TB/s, 85% of peak HBM. rocm.blogs.amd.com/software-…
4
23
1,595