Numbers Provenance Register
1,036 canonical active claims across 1,328 chapter occurrences. Each claim appears once by default; expand it to see every place the guide quotes it. The all-rows view lists those occurrences alongside claims not yet placed in a chapter and the figures on the watchlist.
1,036 active claims1,328 chapter occurrences463 with revision history26 flagged contested— review due
Recently revised
How to use this register
Before carrying a figure into a design or model, match its value kind and scope to your decision, check its as-of and review dates, and open the cited source. A contested figure has credible sources that materially disagree, so test the decision against the competing values. Sparklines and the recently revised list show recorded changes; review health follows each figure's review-by date and fast, standard, or slow cadence.
- derived · 25 Derived values are guide calculations; their result is only as fixed as the stated inputs, scope, and method.
- forecast · 92 Forecasts are time-bound outlooks, not specifications; test decisions across plausible scenarios.
- observed · 899 Reported measurements, counts, and specifications keep the precision and scope stated by their source; an exact specification is not treated as a range.
- proposed · 1 Proposed values are not yet final requirements or specifications; track the decision or rulemaking that can change them.
- recommendation · 19 Recommendations are design guidance, not measured facts; validate them against the project boundary conditions.
1,036 matches · showing 1–100
| Canonical claim | Value | Kind / scope | Trend | As of | Review | Updated | Occurrences | Source |
|---|---|---|---|---|---|---|---|---|
| 1.6T switch deployment; ramp faster than 800G, >5M ports within 1-2 years of shipping | first volume yr 2026 | observed | — | 2026 | checking… | — | Dell'Oro Group | |
| 1 GW AI data center: up-front capex ~$37.9B (servers ~$21.2/W with GPUs inside that line, facility ~$11.4/W, network ~$4.9/W, land + utility works ~$0.3/W) and all-in annual TCO (~$8.5M/MW-yr) | ~$38B / ~$8.5B/yr | observed | — | 2026-05 | checking… | — | Epoch AI, AI datacenter cost breakdown | |
| Guide-model 1 GW annual TCO scenarios at 3-/5-/7-year IT useful life; 3-year case is ~71% above 7-year | $12B / $8.5B / $7B | derivedGuide scenario using Epoch AI / AM Compute inputs; the 3-, 5- and 7-year lives are model assumptions, not measured fleet lives. | — | 2025 | checking… | — | Epoch AI / AM Compute synthesis | |
| 20-rack legacy retrofit all-in vs equivalent new build; ~70% of new-build performance at ~20% of cost · contested | $3.3M vs $16M | observed | — | 2025 | checking… | — | STL Partners — liquid-cooling retrofit ~80% less than new build (contested, single-source practitioner estimate) | |
| 2026 HBM3E supply gap, fully allocated; CoWoS advanced-packaging capacity is the true supply gate | ~30% / sold out | observed | — | 2026 | checking… | — | SemiAnalysis / TrendForce | |
| 3M completed its exit from PFAS manufacturing at the end of 2025, ending its Novec 1230 production; FK-5-1-12 remains available from other manufacturers. | Completed at end-2025 | observedcompany: 3M | — | 2026-08 | checking… | — | not placed | 3M, 3M to Exit PFAS Manufacturing by the End of 2025 |
| 3M PFAS settlement with US public water systems (final approval Mar 2024); Novec exit by end-2025 | $12.5B | observed | — | 2024-2025 | checking… | — | DCD; The Cooling Report; 3M filings | |
| 3M's last Novec/PFAS fluorochemical orders; production wind-down through 2025 (drove the two-phase-immersion exit) | Mar 2025 | observed | — | 2025 | checking… | — | ServeTheHome / DCD / C&EN | |
| ±400 VDC HVDC cables feeding an IT rack from the sidecar (Diablo later rev) | 16 × 50 kW | observed | — | 2026 | checking… | — | OCP Diablo 400 v0.7.0 | |
| 512+ GPU job failure rate after lemon-node ejection — a sensitivity result the model must reproduce | 14% → 4% | observed | — | 2024 | checking… | — | Meta, Revisiting Reliability (arXiv 2410.21680) | |
| 800G optical transceiver lead time — volume orders ~36–50 wk (floor raised after NVIDIA's 2026-03 $4B EML pre-buy), non-priority buyers quoted past 2027; EML lasers are the constraint, 800G supply 40–60% below demand through 2027 | ~36–50 wk | observed | — | 2026-05 | checking… | — | not placed | McKinsey via TechTimes (2026-05); Supply Chain Connect (2026); TrendForce (2026-04) |
| 800G optics power: DSP pluggable vs LPO; CPO under ~6 W | 14-17 W → 7-8.5 W | observed | — | 2025 | checking… | — | Vitex Direct | |
| abandoned factory shell to 100k-GPU training cluster (xAI Colossus, Memphis); doubled to 200k in a further 92 days | 122 days | observed | — | 2024–2025 | checking… | — | NVIDIA Newsroom (Spectrum-X / xAI Colossus) | |
| ABB synchronous-condenser orders announced for VoltaGrid: 27 booked in Q1-Q3 2025 plus 35 additional orders announced 26 March 2026; ordered, not installed | 62 sync condensers | observed | — | 2026-03-26 | checking… | — | ABB | |
| accelerated GPU economic life — contested: bear case 2–3 yr on obsolescence; Goldman Sachs estimates 4–6 yr useful life vs 5–6 yr book · contested | 2–3 yr (bear case) vs 4–6 yr (GS) | forecastGoldman Sachs characterizes four to six years as the estimated useful life of AI accelerators; it does not substantiate a general two-to-three-year economic life or a universal 20–40% three-year residual value. | — | 2026 | checking… | — | Goldman Sachs | |
| accelerator-utilization threshold for a passing MLPerf Storage result (70% for Cosmoflow); holding this line is the loader's core objective | ≥90% | observed | Aug 2025 | checking… | — | MLCommons, MLPerf Storage v2.0 | ||
| accelerators served per storage system in MLPerf Storage v2.0 vs v1.0 — storage scaling tracking compute scaling | ~2x accelerators | observed | — | 2025 | checking… | — | MLCommons (MLPerf Storage v2.0) | |
| acres per MW of IT on GW-scale AI campuses (full campus incl. substation/cooling yard) | 0.7–0.9 | observed | — | 2025–2026 | checking… | — | Reuters / operator filings (Hyperion, Stargate) | |
| active generation + storage in US interconnection queues (end-2025; >1.5x US installed capacity; down from 2,290 GW end-2024) | >2,060 GW | observed | — | end-2025 | checking… | — | LBNL, US interconnection queue data through 2025 (May 2026) | |
| active in-row cooler capacity per rack (contained); fan-wall close-coupled similar | ~30–50 kW | observed | — | 2025 | checking… | — | Vertiv | |
| active local opposition groups across 49 states (more than doubled YoY) | 833 | observed | — | Q1 2026 | checking… | — | Data Center Watch / 10a Labs | |
| active RDHx capacity per rack — mainstream ~50–100 kW; SuperMicro's July-2026 DCBBS line rates ten door models at 10–120 kW at the door (rack-level claim 240 kW — entering-water temperature unpublished, likely dual-door or door+D2C) | ~50–120 kW | observed | — | 2026-07 | checking… | — | Motivair | |
| additional US pipeline capacity INGAA projects is needed by 2052 to serve load growth | +39% | forecast | — | 2025 | checking… | — | Natural Gas Intelligence / INGAA | |
| advanced Nvidia accelerators authorized for the Gulf (G42 + HUMAIN, ~35k each) under the RTE compact | ~70,000 chips | observed | — | Nov 2025 | checking… | — | US Commerce statement; CNBC; DCD | |
| aeroderivative turbine lead time (refurb under 12 mo; some cores quoted to ~243 weeks) — new-build order slots quoted for 2028–2030 as of 2026 | 18–36 mo+ | observed | — | 2025 | checking… | — | 2025 synthesis; Data Center Frontier / power-eng aeroderivative coverage (2026) | |
| vendor-projected shift toward more host CPU capacity for some agentic workloads; ratios require a named accelerator, CPU-package/socket and host denominator | directional; profile host throughput per workload | observedAn eight-accelerator dual-socket host is 8 accelerators per host but 4 accelerators per CPU package. Do not compare ratios with different denominators. | — | 2026 | checking… | — | TrendForce; Intel (via TrendForce) | |
| aggregate NVLink bandwidth in one GB200 NVL72 domain (72 GPUs, ~13.4 TB unified memory) | ~130 TB/s | observed | — | 2025 | checking… | — | NVIDIA NVLink / OCP NVL72 contribution | |
| aggregate storage bandwidth per 1,024 training GPUs (~2 PB initial); place on front-end, not back-end | ~250–400 GB/s | observed | — | 2025 | checking… | — | SemiAnalysis storage sizing | |
| AI-app gross margin vs 70-90% for mature SaaS | ~41% to ~52% | observed | — | 2026 | checking… | — | ICONIQ State of AI 2026; Bessemer | |
| AI-driven incremental US gas demand projected by 2030; US gas demand to ~150 Bcf/d by 2031 | ~6 Bcf/d | forecast | — | 2030-2031 | checking… | — | SemiAnalysis / RAND; INGAA | |
| AI hardware refresh interval compressing from a 5-7 yr norm — the engine of recurring embodied carbon | 18-36 mo | observedhardware refresh interval | — | 2026 | checking… | — | Goldman/CNBC | |
| AI/high-performance storage market 2025, growing ~24–25% CAGR; vendor consolidation underway | ~$36B / ~24% CAGR | observed | — | 2025 | checking… | — | Market research (DataM, SNS Insider et al.) | |
| AI/HPC scheduler share: Slurm / Kubernetes / in-house (rule of thumb) | ~70% / ~20% / ~10% | observed | — | 2026 | checking… | — | HPCwire, ‘Slurm vs Kubernetes in the Age of AI’; ClusterMAX | |
| AI inference capacity to 2030 (~35% CAGR) vs training 23.1 → 62.2 GW (~22%) | 20.9 → 93.3 GW | observed | — | 2026 | checking… | — | McKinsey, 'The next big shifts in AI workloads' | |
| AI-optimized facility build cost, facility-only (shell + fit-out, excl. IT/servers) (JLL 2026 Outlook; Goldman models ~$15M/MW; ~$11.3M/MW all-build global avg) | ~$15-25M/MW | observed | — | 2026 | checking… | — | JLL 2026 Outlook; Goldman Sachs; iRecruit / Construct Elements market data | |
| dense-hall slab capacity has no universal screening allowance; engineer the declared dead, live, concentrated, rolling, rigging, seismic and anchorage load states under the adopted code | Project-specific structural design | recommendationFinal capacity follows site soil, slab, reinforcement, joints, punching, bearing, deflection, vibration, and the declared equipment and movement load states. | — | 2026-08-24 | checking… | — | ASCE/SEI 7-22; ACI CODE-318-25 | |
| AI-spec slab-on-grade thickness (post-tensioned, epoxy rebar) vs 6-8 in legacy; +$50-75/sq ft | 12-18 in | observed | — | 2025 | checking… | — | gbc-engineers (slab/live-load design) | |
| AI storage market (2025), ~24% CAGR; QLC + data reduction displacing HDD across hot and capacity tiers | ~$36B | observed | — | 2025 | checking… | — | Precedence Research; MarketsandMarkets (AI-Powered Storage) | |
| AI storage market size and CAGR — the consolidation pressure behind certified-appliance lock-in | ~$36B / ~24% | observed | — | 2025 | checking… | — | Precedence Research, AI-Powered Storage Market (2025) | |
| air-permit timing for combustion bridge power is jurisdiction- and case-specific; EPA Region 8 examples span roughly 60 days to a year or more depending on modeling, comments and permit path | ~60 days-1 year+; case-specific | observed | — | 2026-08-24 | checking… | — | U.S. Environmental Protection Agency, Region 8 | |
| all-in cost per 8-GPU H100 server (excl. storage); ~$31k/GPU/yr enterprise all-in | $283–318k | observed | — | 2025 | checking… | — | SemiAnalysis AI Neocloud Playbook | |
| all three suppliers' entire 2026 HBM output pre-booked by end-2025; ~30% demand-supply gap | Sold out 2026 | observed | — | 2026 | checking… | — | All Weather Capital | |
| allocation-constrained silicon per GB200 NVL72 rack — theft/destruction economics differ from generic cloud hardware | ~$3M+ | observed | — | 2025 | checking… | — | SemiAnalysis (GB200 NVL72 TCO, 2025); ~$3.1M server-only / ~$3.9M all-in | |
| AMD EPYC 9965 (Turin) host CPU, DDR5-6000/6400; Intel Xeon 6 up to 128 P-cores w/ MRDIMM-8800 | 192 cores / 500 W | observed | — | 2025 | checking… | — | AMD; Intel; Phoronix | |
| AMD Helios (72× MI455X) rack-scale FP4 (1.4 EF FP8), 31 TB HBM4, ORW double-wide | ~2.9 EF FP4 | observed | — | 2026 | checking… | — | AMD / NextPlatform / DCD | |
| AMD Helios / MI400-series rack availability — MI455X UALoE72 low volume 2H 2026, mass ramp Q2 2027; AMD Q1 2026 call: on track, with anchor deals (OpenAI 6 GW) absorbing early supply | 2H 2026 | observed | — | 2026-05 | checking… | — | not placed | The Next Platform |
| AMD MI355X availability — ramping through 1H26; Oracle first hyperscaler with public instances, building a 27,000+ node MI355X cluster | ramping through 1H26 | observed | — | 2025-11 | checking… | — | not placed | Investing.com earnings-call transcript |
| AMD MI355X FP4 (no sparsity) and HBM3E capacity; 1,400 W peak board power | ~10.1 PFLOPS / 288 GB | observed | — | 2025 | checking… | — | AMD Instinct MI355X GPU datasheet (LE-92703-00, 10/25) | |
| AMD MI355X FP4 (sparse), CDNA4, with native FP6 support | ~20 PFLOPS | observed | — | 2025 | checking… | — | AMD MI355X product page | |
| AMD MI355X/MI325X 8-GPU system lead time — editorial proxy, no OEM-published SKU clock as of mid-2026. MI325X ~6–12 wk standard; MI355X ~8–16 wk; 12–24 wk for large or custom liquid-cooled deployments | ~8–20 wk | observed8-GPU systems, accepted orders; proxy range | — | 2026-07 | checking… | — | not placed | Editorial proxy triangulated from SLYD OEM lead-time benchmarks (2026-01) and Crusoe MI355X program ~3–6 mo announcement-to-service (2025-06-12); no public OEM SKU-level quote found as of 2026-07 |
| AMD–OpenAI Instinct supply agreement over 5 yr; first 1 GW (MI450) H2 2026; up to 160M AMD shares to OpenAI | 6 GW | observed | — | 2025 | checking… | — | AMD / TechCrunch / DCD | |
| AMD price discount vs comparable NVIDIA SXM; inference effective-TCO gap → ~0/favorable | ~15–30% | observed | — | 2025-2026 | checking… | — | PromptCost.org | |
| AMD stack with RCCL NCCL-API parity; MI350X/MI355X support (7.0 Sep 2025); production line 7.14 on TheRock (2026-07-15; 7.9–7.13 were tech previews); six-week feature cadence committed Jul 2026 | ROCm 7.14 (production) | observed | — | 2026-07 | checking… | — | AMD ROCm release history; Phoronix (2026-07-16 production release; 2026-07-24 six-week cadence, Advancing AI 2026) | |
| Americans opposing a local data center / worried it raises their energy bill | ~70% / 78% | observed | — | 2026 | checking… | — | Consumer Reports; survey synthesis | |
| annual AI revenue shortfall by 2030 even after AI-driven savings (the bubble-case gap) | ~$800B | forecast | — | 2025 | checking… | — | Bain & Company, Global Technology Report | |
| annual cooling-energy cut from water-side economizing vs mechanical chilling, temperate climates | ~40–65% | observed | — | 2023–2025 | checking… | — | Journal of Korean Institute of Architectural Sustainable Environment and Building Systems | |
| annual EED reporting deadline (prior calendar year); FY2023 first cycle was 15 Sep 2024 | 15 May | observedAnnual operator submission to the European database for the preceding calendar year | 2026-08-24 | checking… | — | Commission Delegated Regulation (EU) 2024/1364 | ||
| annual free-cooling hours in a Nordic climate; PUE as low as ~1.09 | ~7,000-8,000+ hr | observed | — | 2025 | checking… | — | w.media | |
| ANSI/BICSI 002-2024 — the most comprehensive lifecycle design+implementation standard; 2024 ed. expanded liquid/immersion | ~575 pp | observed | — | 2024 | checking… | — | BICSI | |
| ANSI/NETA Acceptance Testing Specifications — the current as-installed bar for switchgear, breakers, relays and primary injection | ATS-2025 | observed | — | 2025 | checking… | — | ANSI/NETA ATS-2025; NETA World Journal | |
| ANSI/TIA-942-C revision date; renames "Tier" to "Rated-1..4" and adds AI-growth & sustainability accommodations | May 2024 | observed | — | 2024 | checking… | — | TIA TR-42.1 | |
| approximate silicon value concentrated in a single GB200 NVL72 rack (1.36 t) — the asset-value density driving target priority | $3–4M | forecast | — | 2025 | checking… | — | NVIDIA / SemiAnalysis (derived) | |
| architecture cadence — Blackwell 2024, Blackwell Ultra 2025, Rubin 2026, Rubin Ultra 2027, Feynman 2028 | annual | observed | — | 2026 | checking… | — | The Next Platform | |
| ASHRAE commissioning-process / Basis-of-Design / data-center-specific Cx guidelines; Std 202 formalizes the Cx-Process | Gd 0 / 1.1 / 1.6 | observed | — | 2025 | checking… | — | ASHRAE; ACHR News | |
| ASHRAE liquid classes by max supply water temp (17/27/32/40/45 °C, plus W+); all share a 2 °C lower limit | W17–W45+ | observed | — | 2025 | checking… | — | ASHRAE TC 9.9 Thermal Guidelines 5th ed.; Upsite | |
| ASHRAE recommended inlet band; allowable A1 15–32, A2 10–35, A3 5–40, A4 5–45 °C | 18–27 °C | observed | — | 2024 | checking… | — | ASHRAE TC 9.9 Thermal Guidelines (5th ed.) | |
| ASHRAE TC 9.9 air classes and liquid W-classes (5th ed. + 2024 liquid-cooling resiliency addendum) | A1–A4 / W17–W45 | observed | — | 2024 | checking… | — | ASHRAE TC 9.9 Thermal Guidelines | |
| ASHRAE TC 9.9 Thermal Guidelines edition introducing the liquid-cooling chapter and W-classes | 5th ed. | observed | — | 2024–2025 | checking… | — | ASHRAE TC 9.9; Upsite | |
| assessed posture of frontier labs vs OC4–OC5 adversaries that want the weights — the central gap | ~SL2–SL3 | observed | — | 2026 | checking… | — | RAND RRA2849-1; IFP / IST SL5 Task Force | |
| attack vectors infeasible for OC1–OC3 but feasible for OC4–OC5 — why nation-state defense is categorically harder | 8 of 38 | observed | — | 2024 | checking… | — | RAND RRA2849-1 | |
| attenuation per doubling of distance from a point source — the cheapest, most reliable decibel; setback is the first lever | ~6 dB | observed | — | 2026 | checking… | — | Acoustic geometric-spreading (inverse-square) principle | |
| total Sweden-Finland transfer capability in each direction after the 380 km Aurora Line was commissioned on 13 Nov 2025 (+~900 MW FI→SE / +~800 MW SE→FI) — which also raised permitted withdrawal around Luleå/Boden from 345 to 745 MW | ~1,900 MW | observedNext planned link, Fenno-Skan 3 (~800 MW), is not expected until ~2038. | — | Nov 2025 | checking… | — | not placed | Fingrid; Svenska kraftnät |
| automated node replacement on a best-in-class fleet — the day-2 lifecycle target | ~90 sec | forecast | — | 2026 | checking… | — | SemiAnalysis AI Neocloud Playbook / ClusterMAX | |
| average annualized insider-risk cost per organization in the cited DTEX/Ponemon report (349 organizations) | $17.4M | observedUse the report's named sample and definition. The source record does not establish a cross-study 'largest ever' comparison. | 2026 report (2025 data) | checking… | — | Ponemon Institute / DTEX 2026 Cost of Insider Risks Global Report (2025 data; 349 organizations) | ||
| average breach cost in IBM's 2025 Cost of a Data Breach report: $4.44M global and $10.22M US | $4.44M global / $10.22M US | observedOne IBM/Ponemon report vintage. Do not combine 2026 values or AI-breach findings with the 2025 value, as-of date and locator. | — | 2025 | checking… | — | IBM/Ponemon Cost of a Data Breach 2025 | |
| average commercial rate in the top-10 US data center states (≈ identical to other states) | ~14.4-14.5¢/kWh | observed | — | 2025-26 | checking… | — | Electric Choice / RealClearEnergy synthesis | |
| average cost of data-center downtime (≈$5,000/min) | $300k/hr | observed | — | 2025 | checking… | — | Uptime Institute / industry downtime analyses | |
| average eCrime breakout time (initial access to first lateral movement) in 2025, down from 48 min in 2024; fastest 27 s | 29 min | observed | — | 2025 | checking… | — | CrowdStrike 2026 Global Threat Report | |
| average energy savings from retro-/re-commissioning of an existing facility (often more) | 10–20% | observed | — | 2023 | checking… | — | U.S. DOE FEMP / LBNL retro-commissioning studies | |
| average time to detect and contain an insider incident (down from 86 in 2023); far longer than a checkpoint copy takes | 81 days | observed | — | 2025 | checking… | — | Ponemon 2025 Cost of Insider Risks | |
| AWS facilities directly hit by drones (UAE) + 1 blast-damaged (Bahrain), Mar 2026 — first confirmed combat strike on US-run hyperscale DC | 2 + blast | observed | — | Mar 2026 | checking… | — | DefenseScoop / DCK / MWI (West Point) | |
| AWS Trainium3 availability — shipping since start of 2026 and nearly fully subscribed; Trainium2 largely sold out; Trainium4 planned for 2027; AWS in-house chips (Trainium+Graviton+Nitro, blended) ~$25B annualized with ~$225B multi-year commitments | nearly fully subscribed | observed | — | 2026-08 | checking… | — | not placed | Amazon Q1 2026 earnings, Andy Jassy (2026-04-30); Q2 2026 earnings + Bezos remarks (Aug 2026): Trainium4 2027, ~$25B annualized in-house chip revenue, $225B commitments |
| back-calculated NVL72 footprint loading (3,000-3,300 lb over 0.64 m²; ~21-23 kN/m²); NVIDIA publishes no official floor spec | ~435-480 psf | observed | — | 2026 | checking… | — | HPE | |
| back-end fabric cost avoided with 2:1 oversubscription (virtual-modular-switch design) vs the non-oversubscribed reference — where the workload allows (contested — single-source) · contested | ~31% | observedSemiAnalysis models the 31.6% saving for 2:1 oversubscription and recommends 2:1 for small clusters; the cited passage does not model ~3:1 (Meta ran 7:1 between pods). | 2024 | checking… | — | SemiAnalysis, AI Neocloud Playbook and Anatomy | ||
| back-end NICs per training node = 3.2 Tb/s/node; 1:1 non-blocking 8-rail fat-tree | 8× 400 Gb/s | observedOne named 8-rail training reference design, not a workload-name mandate. Validate the port rate, rail count, and blocking ratio against the target traffic matrix, topology, placement, failure headroom, and step-time objective. | — | 2025 | checking… | — | SemiAnalysis AI Neocloud Playbook | |
| back-end SuperNIC provisioning (8×400G/800G per 8-GPU server = 3.2–6.4 Tb/s/node) + a separate front-end/storage NIC | 1 per GPU | observed | 2025 | checking… | — | NVIDIA RA | ||
| Bartz v. Anthropic settlement — largest US copyright payout; ~$3,000 per work across ~500,000 works; pirated copies ordered destroyed | $1.5B | observed | 2026-07-20 (final judgment) | checking… | — | Bartz v. Anthropic (N.D. Cal.); Authors Guild; Fortune | ||
| ISO-Base platform displacement (standard ~6 in; XD 10.25 in) and supported equipment weight (~$15-20k/unit, practitioner estimate — vendor does not publish pricing) | 6–10 in / up to 10,000 lb | observed | — | 2026 | checking… | — | WorkSafe Technologies (ISO-Base / ISO-Base XD) | |
| battery-capacity cut and double-conversion-loss cut from distributed BBU + supercaps vs central 2N | ~50% / 2–3% | observed | — | 2026 | checking… | — | SemiAnalysis; Eaton 800 VDC reference architecture | |
| behind-the-meter gas announced cumulatively by mid-2026 (~60 GW with equipment orders, ~two-thirds of projects; only ~1 GW under construction) | ~90 GW | forecast | 2026-08 | checking… | — | Cleanview, 'Bypassing the Grid' (mid-2026) | ||
| behind-the-meter (mostly gas) announced since Jan 2025; the speed-vs-carbon tradeoff in one number | ~82 GW | forecast | 2026 | checking… | — | Cleanview | ||
| reported MTBF for one 512-H100 cluster at a top-tier operator | ~7 days / one 512-H100 cluster | observedSemiAnalysis reported this for one 512-H100 cluster at a top-tier operator in October 2024. It is not a per-GPU rate, a scaling law, or a portable fleet baseline; do not extrapolate it by accelerator count. | — | 2024-10 | checking… | — | SemiAnalysis, AI Neocloud Playbook and Anatomy | |
| BF16 MFU achieved pre-training Llama 3 on 16k H100s (frontier-scale reference point) | ~41% | observed | — | 2024 | checking… | — | Meta, The Llama 3 Herd of Models | |
| BF16 MFU gain on H100 (GPT-3 175B) from software/kernel maturation over ~12 months, Jan–Dec 2024 (≈57% throughput from software alone) | 34% → 54% | observed | — | 2024 | checking… | — | SemiAnalysis (H100 vs GB200 NVL72 Training Benchmarks, Oct 2025) | |
| BIS moves named advanced-compute exports to China/Macau from presumption-of-denial to case-by-case review | Jan 15 2026 | observed | — | 2026 | checking… | — | US BIS final rule; Baker McKenzie sanctions analysis | |
| Blackwell (B200/GB200) allocation backlog for new orders — the last cleanly reported order-to-delivery lead time; NVIDIA has since communicated allocation status, not weeks | ~12 mo | observed | — | 2024-10 | checking… | — | Morgan Stanley via Data Center Dynamics (2024-10-11) | |
| Blackwell CC overhead on large matrix ops (encrypted HBM + TEE-I/O over NVLink); Hopper far heavier on small/PCIe transfers | under ~3% | observed | — | 2025 | checking… | — | NVIDIA; independent Hopper CC benchmark (arXiv 2409.03992) | |
| Blackwell vs Rubin 2026 shipment mix — TrendForce raised Blackwell to ~71% and cut Rubin to 22% on HBM4 validation, the ConnectX-8→9 migration, power and cooling steps | Blackwell ~71% of 2026 | observed | — | 2026-04 | checking… | — | not placed | TrendForce via The Register (2026-04-08) |
| Bloom–Oracle SOFC framework (1.2 GW initially contracted); near-zero criteria pollutants; ~$3,000–4,000/kW | up to 2.8 GW | observed | — | 2026 | checking… | — | Bloom Energy IR / press | |
| BlueField-4 Arm Neoverse V2 complex (64B transistors, ~6× BF-3 compute); 128 GB LPDDR5; PCIe Gen6 | 64-core | observed | — | 2026 (announced) | checking… | — | NVIDIA / WWT / HPCwire |