Archive
All analysis
Every article on GPU infrastructure, cloud pricing and AI hardware, newest first.
31 articles
GPU VPS vs Cloud GPU: The Contract You Rent Decides the Cost, Not the Hourly Rate
GPU VPS vs cloud GPU: four contracts behind the label, three providers bill powered-off GPUs, and one 24-month flat plan beats hourly above 10.49 hours.
Read the analysis →Nemotron Labs Dory 1.0 4B on an 8 GB GPU: What Fits and What Does Not
Nemotron Labs Dory 1.0 4B fits an 8 GB GPU in Q5_K_M or FP4 formats, but BF16 does not. Weight sizes, loop trade-offs and what is still unverified.
Read the analysis →Local AI Hardware in the DRAM Crunch: Buy Now or Wait? (DGX Spark vs Strix Halo vs Mac Studio vs a GPU Rig)
Local AI hardware: DGX Spark is $4,699, 128GB Strix Halo boxes sit near $3,500, and the new Mac Studio starts at $2,499. Who should buy, wait, or rent.
Read the analysis →GPU Electricity Cost: What a 700-Watt H200 Adds at July 2026 U.S. Prices
GPU electricity cost at 700 watts is $0.0684 an hour at the July 2026 U.S. industrial price and $0.1017 at the commercial price, before the building.
Read the analysis →NVIDIA Data Center Revenue in Q2 FY2027: The Split Buyers Should Use
NVIDIA data center revenue was $89.0 billion in Q2 FY2027. Hyperscale grew 13.15% and the other line 25.21%. The mix, recomputed from the 8-K.
Read the analysis →Server Useful Life: What Five AI-Infrastructure 10-Ks Actually Depreciate
Server useful life in five 2025–2026 10-Ks runs from 2 to 6 years. Meta lengthened, Amazon shortened a subset, and the dollar gap is recomputed.
Read the analysis →KV Cache Memory: How Many Llama 3 Tokens Fit on One H200
KV cache memory on one H200: Llama 3 70B at 16-bit weights leaves 1.0 GB, about 3,052 tokens. Byte formula, unit trap, and a worked example.
Read the analysis →MLPerf® Inference v6.1 Per-GPU Results: The 5.7× Headline, Recomputed for Hardware You Can Buy
Per-GPU recomputation of MLPerf® Inference v6.1: the 5.7× headline is 3.04× for Available hardware, and software moved AMD 37.9%.
Read the analysis →Edge AI vs Cloud GPU Inference 2026: The Data Sovereignty and Latency-Cost Framework
A construction-site monitoring system that ships raw video to the cloud for object detection is not primarily an engineering decision in Europe in 2026 — it is a…
Read the analysis →Custom AI Silicon vs GPU Rental 2026: When TPU, Trainium, and Maia Beat NVIDIA and AMD
Google will tell you its seventh-generation Ironwood TPU delivers roughly 44% lower total cost of ownership than a comparable GB200 configuration. What Google will not tell you, three…
Read the analysis →vLLM vs SGLang vs TensorRT-LLM 2026: The Enterprise Inference Engine Decision
A platform engineer migrating a production chatbot from vLLM to TensorRT-LLM in early 2026 discovered the throughput gain everyone promised — roughly 13% more tokens per second —…
Read the analysis →The Enterprise GPU Procurement Playbook 2026: Contracts, Capacity Guarantees, and Provider Financial Risk
A mid-market AI company signed what its infrastructure lead described as a "capacity guarantee" for 200 GPUs with a growing neocloud in early 2026. Six months later, during…
Read the analysis →