TechInsights' own product page states a different 2023 NVIDIA GPU count and market share than trade journalism attributes to its study

ai-energygpunvidiatechinsightsmarket-sharequantitative-verificationcitation-accuracy

TechInsights' own product page for "Data-Center AI Accelerator Market – Q1 2024 Update" (product code ARI-2404-801, release-dated 2024-04-12 — inside the same reporting window HPCwire and DataCenterDynamics cite) states: "NVIDIA tripled its revenue and shipped 3.5 million GPU accelerator cards in 2023. NVIDIA controlled 95% of the market, with no major competition in sight in the AI accelerator market." This is a different unit count (3.5M vs. the 3.76M reported by HPCwire and DCD) and a different market-share figure (95% vs. 98%) than what trade journalism attributes to "a study conducted by semiconductor analyst firm TechInsights" for the same calendar year — see claim-nvidia-2023-gpu-shipment-figure-traces-only-to-paywalled-techinsights-study.

The discrepancy plausibly reflects two different TechInsights market definitions — "data-center GPU" shipments broadly (3.76M, including non-AI-accelerator data-center GPUs) versus the narrower "AI accelerator" segment (3.5M) — rather than a contradiction; neither the product page nor the journalism discloses enough methodology to confirm that reading. Either way, TechInsights' own public self-description does not reproduce the exact figure Luccioni, Strubell & Crawford's citation chain uses, from any angle: not the year (see claim-luccioni-2025-nvidia-gpu-figure-misdated-2023-as-2024), and not, cleanly, the number.

A market-concentration figure diverging between a firm's own account and journalism's retelling of it is a familiar shape in this vault — compare claim-ibm-market-share-fell-during-antitrust-suit-price-strategy, where a market-share number also reached the vault only through a secondary retelling of an uncited figure.