SemiAnalysis

SemiAnalysis argues AMD's gap to NVIDIA is software, not silicon: theoretical perf-per-dollar favors AMD (more so with the rebates going to Meta and OpenAI), but engineers lack internal R&D GPU capacity and stable QA clusters. AgentX is the benchmark they are optimizing against.

SemiAnalysis tracks ICLR submissions rising from 4,938 in 2023 to 19,525 in 2026 and NeurIPS acceptances up 49% in a year. It argues cheap AI-assisted drafting is outpacing reviewer capacity and eroding what acceptance signals.

SemiAnalysis walks through recent vLLM work aimed at agentic serving: one KV-cache save per shared prefix instead of a full per-agent re-save each turn, and complete state transfer between prefill and decode for hybrid models. Derived from replayed agent traffic and their AgentX multi-turn benchmark

SemiAnalysis's institutional financial model analyzes the ownership economics of AI clouds ('NeoClouds') that buy GPUs and resell compute, covering GPU rental price forecasts, cost-per-token training/inference economics, and a full three-statement financial model with IRR, NPV, and residual value analysis. It's aimed at AI cloud operators, investors, and large compute buyers evaluating long-term GPU procurement or cluster investment decisions.

SemiAnalysis walks through OpenAI's Jalapeño accelerator on its podcast: performance per megawatt against Blackwell, where the benchmark comparisons break down, the programming model and CUDA lock-in, Samsung HBM4 supply, and how far AI-assisted silicon design carried the project.

Co-packaged optics moves fiber attach onto the photonic IC, where sub-micron alignment decides coupling loss. SemiAnalysis walks through active alignment's serial cost, kinematic coupling's deterministic six-point constraint, and elastic averaging's cheaper compliant approach, with the yield and str

SemiAnalysis frames subscriptions as monthly credit grants consumed at model-specific rates that diverge from API prices. It shows how the API-equivalent value of one $200 Claude plan changes with model and workload.

SemiAnalysis open-sources AgentX 1.0, an Apache-2.0 inference benchmark built on replayed multi-turn Claude Code traces at 1M context. It compares NVIDIA and AMD systems across DeepSeek, Kimi, MiniMax and Qwen models, and details the vLLM, SGLang and TensorRT-LLM fixes it drove.

SemiAnalysis examines OpenAI's Jalapeño inference ASIC after lab access: HBM4 and 700W TDP specs, a unified design with no prefill-decode split, rack topology, Gluon kernels written with Codex, and InferenceX results against Blackwell and Rubin. Data is vendor-supplied and limited to 8k/1k workloads

SemiAnalysis discussion on subscription economics: a $200 Claude plan can yield about $12,000 of Opus 5.5 tokens at API pricing, versus roughly $5,000 to $6,000 on comparable ChatGPT plans, depending on workload and model.

Claude Code 2.1.299 adds idle compaction. Sessions with 200k or more context compact after about 54 minutes idle, refreshing the one-hour cache TTL so resumed sessions avoid uncached input costs. Reported by SemiAnalysis, not confirmed against the changelog.

On DeepSeek v4.1 Flash's day-zero release, NVIDIA's vLLM worked cleanly on H100 through GB300 while AMD's announced ROCm image for the model was still unavailable a day later, per SemiAnalysis.

SemiAnalysis published ClusterMAX 3.0, its latest deep-dive rating of GPU cloud providers on reliability, performance, support, pricing, and security, intended as the reference for choosing compute vendors.