Dan Nishball
Director of Research, SemiAnalysis
Dan is the Director of Research at SemiAnalysis, where he leads analytical efforts across AI infrastructure and compute. With deep technical expertise and strategic insight, he leads research from transistor-level innovation to hyperscale AI deployment. Dan’s focus areas of research include Networking and Optical, AI Systems and Neocloud Systems and Economics. His work spans the full stack – producing influential research that supports decision makers at the world’s leading technology firms, investment funds, and beyond.
Dan Nishball
Director of Research, SemiAnalysis
Dan is the Director of Research at SemiAnalysis, where he leads analytical efforts across AI infrastructure and compute. With deep technical expertise and strategic insight, he leads research from transistor-level innovation to hyperscale AI deployment. Dan’s focus areas of research include Networking and Optical, AI Systems and Neocloud Systems and Economics. His work spans the full stack – producing influential research that supports decision makers at the world’s leading technology firms, investment funds, and beyond.
Articles by Dan Nishball:
-
Datacenter Anatomy Part 2 – Cooling Systems
L2A, L2L, Immersion, Two-Phase, Google vs Meta vs Microsoft vs Amazon Water Cooling Design, WUE, PUE, Nvidia Rubin Power & Cooling Architecture
-
2025 AI Diffusion Export Controls – Microsoft Regulatory Capture, Oracle Tears, Impacts Quantified, Model Restrictions
Malaysia's Stranded Capacity, Western Hyperscale Strength, AI Accelerator Restrictions, India Brazil Middle East Concerns, Sovereign AI Killed, VEU Restrictions and Implications, Compute Budget, Open-Source Creating Lower Bound, Model-Weights Controlled
-
Nvidia’s Christmas Present: GB300 & B300 – Reasoning Inference, Amazon, Memory, Supply Chain
Blackwell Delays, Microsoft Orders, GB300 BOM, Nvidia Gross Margin, ConnectX-8, VRMs, Micron, Samsung, SK Hynix, Wistron, FII Foxconn, Aspeed, Axiado
-
MI300X vs H100 vs H200 Benchmark Part 1: Training – CUDA Moat Still Alive
Training Performance, User Experience, Usability, Nvidia, AMD, GEMM, Attention, Networking, InfiniBand, Spectrum-X Ethernet, RoCEv2 Ethernet, SHARP, Total Cost of Ownership
-
Scaling Laws – O1 Pro Architecture, Reasoning Training Infrastructure, Orion and Claude 3.5 Opus “Failures”
AI Lab Synthetic Data Infrastructure, Inference Tokenomics of Test Time Compute, The Data Wall, Evaluations are Broken, RLAIF, Inference Time Search, Scale Needed More Than Ever
-
Amazon’s AI Self Sufficiency | Trainium2 Architecture & Networking
Trn2, Trn2-Ultra, Performance, Software, NeuronLinkv3, EFAv3, TCO, 3D Torus, Networking Costs, Supply Chain
-
Datacenter Anatomy Part 1: Electrical Systems
Meta Datacenter Scrapped, Vertiv, Schneider Electric, Eaton, Legrand, Delta, Datacenter Bill Of Materials By Component, Transformers, Switchgear, Redundancy, UPS, OCP Busbar, Generators, Substation
-
AI Neocloud Playbook and Anatomy
H100 Rental Price Cuts, AI Neocloud Giants and Emerging Neoclouds, H100 Cluster Bill of Materials and Cluster Deployment, Day to Day Operations, Cost Optimizations, Cost of Ownership and Returns
-
Multi-Datacenter Training: OpenAI’s Ambitious Plan To Beat Google’s Infrastructure
Gigawatt Clusters, Telecom Networking, Long Haul Fiber, Hierarchical & Asynchronous SGD, Distributed Infrastructure Winners
-
Nvidia’s Blackwell Reworked – Shipment Delays & GB200A Reworked Platforms
MGX GB200A NVL36, B102, B20, CoWoS-L, CoWoS-S, GB200A NVL64, ConnectX-8, Liquid Cooling vs Air Cooling, NVLink Backplane, PCB, CCL, Substrate, BMC, Power Delivery
