> ## Content Index > Fetch the complete content index at: https://adikumar.co/llms.txt > Use this file to discover other available public pages before exploring further. # Two architectures wearing the same name (and how the industry is converging) (2/13) - URL: https://adikumar.co/two-architectures-wearing-the-same-name/ - Published: 2026-05-01T07:08:08.000Z - Updated: 2026-10-07T05:39:14.000Z - Description: The trade press writes about 800 VDC as one architecture. There are two, they behave differently in retrofit, and hyperscalers are choosing between them right now. - Author: Adi Kumar - Tags: #Essay, The DC-DC Transition, #disclosure, AI Infrastructure, Nvidia, 800V DC, OCP Monte Diablo [The DC-DC Transition series](https://adikumar.co/dc-dc-transition/) · Part 2 of 13 **Update.* Since this essay was first published, the picture has moved. OCP's broader LVDC programme now explicitly encompasses 800 VDC as a first-class workstream. Meta, Google, Microsoft and Nvidia are aligning MVAC-to-800 VDC conversion requirements, power quality and system interfaces so suppliers can build interoperable equipment. Both paths stay in scope: the 480 VAC sidecar for existing facilities, and a longer-term MVAC-to-800 VDC hall architecture. The divergence this essay captured is being coordinated inside a shared ecosystem. The deployment-level distinctions between sidecar and native architectures still drive retrofit, capex, protection regime and vendor stack. The essay follows as originally written.* The 800 VDC label covers two distinct deployment architectures under a single voltage. Nvidia's reference design rectifies at row scale and hands a locked 800 VDC bus to a vertically integrated compute rack. The Open Compute Project's Mt. Diablo specification, authored by Meta, Google, and Microsoft, rectifies in a sidecar power rack adjacent to the IT rack, supports a configurable output (bipolar ±400 V or unipolar 800 V), and accepts any accelerator that complies with the spec. Both deliver 800 V of usable potential at the rack inlet. Everything else about the two paths differs. This essay is about why that distinction matters more than it looks, and what it tells us about the structure of the data centre power industry over the next five years. Worth saying up front: at scale, most hyperscalers will operate *both* architectures inside the same estate, but because their workload mix and silicon mix demand it. The choice isn't binary; the *primary* architecture choice is what shapes everything else. Under 800 VDC, electrical capex relocates upstream. The sidecar keeps the conversion boundary beside the rack; the native architecture puts it at the facility perimeter. ## A note on terminology, continued Essay 1 made the case that calling this transition "HVDC" is technically wrong: IEC frameworks put the LVDC boundary at 1,500 V DC. I'll use *800 VDC* for the unipolar architecture, *±400 VDC* or *Mt. Diablo* for the bipolar one, and *the DC transition* when referring to the broader industry shift. All of these sit firmly inside the LVDC regime. The trade press will keep calling it HVDC. We will keep being correct. ## 1\. What are the two 800V DC architectures?, sketched The cleanest way to see the difference is to draw both at the same scale, with the same components, and look at what's structurally different. ![Two-panel architectural diagram: sidecar architecture on the left, native facility-scale on the right](https://adikumar.co/content/images/2026/09/essay-2-fig1-sidecar-vs-native-v2.png) Figure 2.1\. Two architectures under one voltage. Sidecar / hybrid on the left keeps AC/DC conversion beside the rack (Mt. Diablo v0.7.0); native / facility-scale on the right puts it at the perimeter (Nvidia reference and OCP's longer-term MVAC-to-800 VDC path). **NVIDIA 800 VDC reference (row-scale rectification).** Medium-voltage AC arrives from the utility, hits a solid-state transformer at the building perimeter or at the row level, and emerges as a single bus at +800 V to ground. That bus runs along the row to each compute rack, where compact step-down converters deliver the 50 V class to the GPU. The compute and power conversion are designed as one vertically integrated system: the Kyber rack, designed for the Vera Rubin Ultra accelerator, expects 800 VDC at the rack inlet and won't accept anything else. **OCP Mt. Diablo (sidecar rectification, configurable output).** Same medium-voltage AC input, but rectification happens in a dedicated power rack sitting next to one or more IT racks; row-scale rectification sits elsewhere. The IT racks themselves contain no PSUs, every rack unit is available for accelerators and scale-up switching. The Diablo 400 specification permits two output configurations from the sidecar: a three-wire bipolar arrangement at +400 V, 0 V (common), and −400 V, or a two-wire unipolar arrangement at 800 V and return. The IT racks are ORv3 HPR (Open Rack v3 High Power Rack) and accept any accelerator that complies with the spec, NVIDIA's parts work, but so do Trainium, MTIA, Maia, and AMD's MI series. The defining contrast: **where rectification happens** (sidecar vs row), and **whether compute is locked to a single accelerator family**. The voltage-class debate (bipolar vs unipolar) fits inside Mt. Diablo as a configuration choice. Same nominal usable voltage, different rectification topology, different stress to ground in the bipolar case, different supply chain, different commercial intent. ## 2\. The dimensions that matter Past the marketing, there are six dimensions where the architectures diverge. | Dimension | NVIDIA 800 VDC reference | OCP Mt. Diablo | | ------------------------- | -------------------------------------- | ---------------------------------------------- | | Rectification location | Row-scale (perimeter or row-level SST) | Sidecar (rack-adjacent power rack) | | Default rack output | 800 V unipolar | ±400 V bipolar (800 V unipolar permitted) | | Conductor count to a rack | 2 (positive + return) | 3 (+400, common, −400) in default config | | Supply chain leverage | Purpose-built 800 V parts | EV\-derived 400 V parts (default config) | | System scope | Vertically integrated AI factory | Compute and power as separable building blocks | | Accelerator compatibility | NVIDIA Vera Rubin family | Any ORv3 HPR. Compliant accelerator | Figure 2.2\. Architecture comparison across six dimensions. ### Glossary of terms used EV Enterprise Value. Purchase price plus debt assumed minus cash. Total value of the business excluding capital structure. OCP Open Compute Project. Hyperscaler-led standards body developing open reference designs for data centre hardware. VDC Volts Direct Current. Used in 48 VDC and 800 VDC references to rack-level and datacentre power distribution. For the full corpus glossary of acronyms used across all essays, see [adikumar.co/glossary](https://adikumar.co/glossary/). Series expansion · six new supplements The DC-DC Transition series now includes six deep-dive supplements addressing specific questions that the main 13-essay series flagged but did not fully develop. Consulting-grade visuals throughout. Vendor concentration matrices, procurement flow diagrams and capex decomposition views: 1. [Supplement A. The Transition Tax](https://adikumar.co/dc-dc-supplement-a-transition-tax/): capex decomposition across the six domains, covering where the 800V DC premium accumulates in retrofit versus greenfield paths. 2. [Supplement B. Hybrid Forever](https://adikumar.co/dc-dc-supplement-b-hybrid-forever/): why most facilities will run mixed AC/DC through 2035+. Hyperscaler zone strategies compared. 3. [Supplement C. The Voltage Ceiling](https://adikumar.co/dc-dc-supplement-c-voltage-ceiling/): 1500V DC and the MVDC question. Standards + vendor readiness by voltage class. 4. [Supplement D. The Solid-State Transformer](https://adikumar.co/dc-dc-supplement-d-solid-state-transformer/): facility-distribution supplier dynamics. Western vs Chinese vendor field. 5. [Supplement E. The Sidecar Power Rack](https://adikumar.co/dc-dc-supplement-e-sidecar-power-rack/): the rack-shelf architecture. OCP Mt. Diablo v0.7.0 reference. 80+ ecosystem partners. 6. [Supplement F. The Connector Wars](https://adikumar.co/dc-dc-supplement-f-connector-wars/): rack and on-package interface engineering. High-current connector vendor concentration + failure modes. Full series index: [/tag/dc-transition/](https://adikumar.co/tag/dc-transition/)All supplements now live. Further reading + sources **Related essays** - [The Architecture Map](https://adikumar.co/the-architecture-map/). the framework this essay builds on - [Supplement B. Hybrid Forever](https://adikumar.co/dc-dc-supplement-b-hybrid-forever/). why most facilities will run mixed AC/DC through 2035+ - [Supplement E. The Sidecar Power Rack](https://adikumar.co/dc-dc-supplement-e-sidecar-power-rack/). the specific rack-shelf architecture that drives both branches **Primary sources** - [OCP Rack + Power](https://www.opencompute.org/projects/rack-power?ref=adikumar.co). the specification defining the sidecar form at Layer 4 - [Uptime Institute Global Data Center Survey](https://uptimeinstitute.com/resources/asset/global-data-center-survey?ref=adikumar.co). operator-side deployment pattern data In brief 800V DC covers two deployment paths. OCP Mt. Diablo rectifies in a rack-adjacent sidecar with a default ±400V bipolar output; an 800V unipolar option is permitted in the same spec. Nvidia's reference rectifies at row scale on a unipolar 800V bus. The distinctions drive retrofit, capex, protection regime and vendor stack. Both paths are converging inside the OCP 800VDC subcommittee. Several of those dimensions deserve real attention before we move on, because the marketing on each side oversells them. **On output voltage class.** The Diablo 400 spec (v0.7.0, March 2026) explicitly permits both bipolar and unipolar output from the sidecar. So the "NVIDIA = 800 V, Mt. Diablo = ±400 V" framing in trade press is incomplete. The default and most-promoted Mt. Diablo configuration is bipolar ±400 V. That's the configuration that captures the EV supply chain argument. But operators that want unipolar 800 V output can get it within the same spec. The real contrast is sidecar vs row. **On fault behaviour.** It's tempting to frame Mt. Diablo's bipolar option as a reliability win, since half the bus stays alive when one rail faults to ground. In theory, yes. In practice, most protection schemes will trip the entire bus on a single pole-to-ground fault for safety reasons, regardless of whether the surviving rail could handle partial load. The bipolar advantage holds in fault energy (each conductor is at 400 V to ground (half the 800)), but the "ride-through on one rail" framing is more aspirational than operational at this point. **On conductor count.** Bipolar ±400 V needs three conductors plus careful neutral handling per rack; unipolar 800 V needs two. At 600 kW+ per rack, where the industry is heading, this registers as a non-trivial difference for busway sizing, distribution complexity, and per-rack copper capex. More on this below. **On neutral balance.** Bipolar ±400 V achieves minimum copper only when the two rails hold symmetric loads. AI workloads step-load in the millisecond band as thousands of accelerators synchronise, so asymmetric draw between rails pushes current down the neutral conductor. Active balancing in the sidecar mitigates this. The bipolar default imposes it on the vendor stack. **On DC fault clearing.** AC breakers rely on the current zero-crossing every 8.3 milliseconds to extinguish the arc. DC has no zero-crossing. Interrupting a short at 800 V with capacitive or battery reserve behind it requires solid-state circuit breakers with series SiC switches, hybrid mechanical-plus-semiconductor breakers, or pyrotechnic disconnects. The industrial supply chain for those parts is thin. This is part of why UL 98B is on the signals list below. The dimension that counts most for the next three years, though, is the supply chain. ## 3\. Why ±400 V is really an EV play in disguise The under-discussed feature of Mt. Diablo's default bipolar configuration is that 400 V came from automotive supply pricing, not technical elegance. They picked it because there is already a planet-scale supply chain for components rated at 400 V, built by the electric vehicle industry over the past decade. The Diablo 400 specification itself names this rationale, *"selecting 400 VDC as the nominal voltage leverages the supply chain established by electric vehicles, for greater economies of scale, proven quality, and more efficient manufacturing."* A precision note before the argument: most BEV traction systems still use \~400 V battery packs (Tesla, BYD volume models, most Chinese EVs). The components inside an EV carry mixed ratings relative to 400 V to ground. Pack architecture, motor inverter topology, and isolation strategy all vary. But the discrete 400 V-class parts (IGBTs, SiC MOSFETs, capacitors, contactors, fuses, connectors) are produced at scale specifically because the EV industry needs them in this voltage class. Annual production of EV-grade 400 V power modules is in the millions of units across the major suppliers (Infineon, ST, Wolfspeed, ON Semi, Mitsubishi). The 400 V class as a *standardized component design point* really consolidated between 2015 and 2018 as second-generation EV platforms moved off bespoke voltages onto common ones. That curve has been compounding for a decade. The 800 V class, by contrast, is much earlier. Yes, EV fast charging stations and 800 V powertrains (Porsche Taycan, Hyundai E-GMP, Lucid Air, Kia EV6) operate at this class but volumes are roughly an order of magnitude below the 400 V base, the components are different (rectifiers and chargers), and the certification frameworks for stationary 800 V data centre equipment are still maturing. The most cited recent UL milestone, ABB earning UL 98B certification for a 2 kV switch-disconnector in July 2025, is illustrative precisely because of how narrow it is. That certification was for **utility-scale solar** (data centres out of scope), and as the standards press noted at the time, UL 98B itself only tests up to 1500 V; ABB's 2 kV cert was a stretch case ahead of UL writing the broader voltage range into the standard. The standards extension to genuinely cover 800 V stationary data centre topologies, with the arc, protection, and isolation requirements that come with that environment, is an active area of work. ![Bar chart showing 400 V class component maturity vs 800 V class](https://adikumar.co/content/images/2026/05/essay-2-fig3-component-supply-chain.png) Figure 2.3\. Component supply chain maturity. The 400 V class rides the EV ecosystem; the 800 V class is purpose-built. So when Meta, Google, and Microsoft jointly authored Mt. Diablo with bipolar ±400 V as the default configuration, they were doing something specific: they were borrowing the EV industry's component cost curve. By 2027 they'll be deploying a power architecture whose discrete components are riding a learning curve that started consolidating a decade ago, while NVIDIA's vendors will still be on the steeper early portion of an entirely new 800 V curve. ## 4\. The unbundling pattern Mt. Diablo is the visible piece of a larger pattern. Hyperscalers are systematically unbundling Nvidia's vertical stack wherever they can: silicon, form factor, power, network, software. ![NVIDIA vertical stack vs hyperscaler parallel options](https://adikumar.co/content/images/2026/05/essay-2-fig4-unbundling-map.png) Figure 2.4\. The unbundling map. Hyperscaler alternatives across Nvidia's vertical stack, tier-coded by maturity. But the framing "alternatives exist at every layer" is doing work it shouldn't. Some of those alternatives are genuinely contested with real production deployment behind them. Others are aspirational specs with limited silicon to back them. Worth being precise about which is which: **Genuinely contested (real production volume, real deployment):** \- *Silicon:* Google TPU (a decade of internal deployment, now v7 Ironwood in volume), AWS Trainium 3, Meta MTIA v3, Microsoft Maia 200\. NVIDIA's data centre share is projected to drop from \~86% in FY2025 toward \~75% by end of 2026, with custom ASIC volume growing at roughly 45% CAGR through 2028\. - *Form factor:* OCP ORv3 HPR is shipping at scale across Meta, Google, Microsoft fleets while Nvidia reference systems still dominate neoclouds. - *Power:* NVIDIA 800 VDC vs. Mt. Diablo. The fight this essay is about. **Partially contested (real specs, limited silicon, dependent on future tape-outs):** \- *Networking:* Google ICI is a fact and at scale internally, but Ultra Ethernet Consortium specs have shipped while UEC silicon volume remains modest. NVLink and InfiniBand still dominate in non-Google deployments. - *Software at training:* JAX and PyTorch-XLA are credible, but the developer base remains small relative to CUDA. OpenXLA matters strategically but commands a fraction of CUDA's ecosystem mindshare. **Aspirational (specs exist, production is the open question):** \- *Software at inference:* vLLM and SGLang have material production deployment, but the dependence on CUDA underneath persists for any workload that doesn't have a dedicated compiler team. The honest summary: the unbundling is complete for silicon, form factor and power. Networking and training software are partial. Application software remains aspirational because CUDA's two-decade lead, four-million-developer ecosystem, and framework-first integration create switching costs measured in years. ## 5\. The hyperscaler silicon investment The unbundling in silicon specifically deserves its own look, because it funds all the others. ![Mekko chart showing hyperscaler NVIDIA vs custom silicon mix](https://adikumar.co/content/images/2026/05/essay-2-fig5-hyperscaler-silicon.png) Figure 2.5\. Hyperscaler AI compute by silicon source. NVIDIA's data centre revenue ran $115B in FY2025 and is tracking toward $194B in FY2026\. That number is going up. But the *share* of new AI compute that goes to NVIDIA is declining, and the rate of decline is the part that counts. Consensus from Bloomberg Intelligence, TrendForce, SemiAnalysis, and Counterpoint converges on three claims that should drive power architecture decisions: - NVIDIA's AI accelerator market share has declined from a peak of \~95% to an estimated 80-86% in 2025-2026, with most analysts forecasting further decline toward 75% by end of 2026 as custom ASIC volume scales. - Custom ASIC shipments are projected to grow at roughly 45% CAGR through 2028, with shipments crossing GPU shipments in 2028 per Counterpoint. - The displacement is most pronounced in \*\*inference\*\*, where custom ASICs are approaching parity in deployed FLOPs by 2027-2028\. Frontier \*\*training\*\* remains NVIDIA-dominated and probably will through 2028, anchored by CUDA's ecosystem advantage and Vera Rubin's raw performance. NVIDIA's revenue keeps growing in absolute terms because the AI compute pie is growing faster than the share shift. The displacement is a fact but asymmetric, with share down and dollars up. Three production realities matter most for power architecture decisions: **Google TPU v7 (Ironwood)** entered volume production in late 2025, with industry sources estimating roughly 4.3M units shipping in 2026 and analyst forecasts pointing to cumulative deployment in the tens of millions across 2026-2028\. Anthropic specifically has committed to deploying over one million Ironwood chips, representing more than a gigawatt of compute. Ironwood is designed and deployed in ORv3 HPR-class racks with sidecar power. **AWS Trainium 3** reached general availability in December 2025 at re:Invent, built on TSMC 3nm. The Trn3 UltraServer scales to 144 chips across 36 servers in a multi-rack configuration; the volume per-rack SKU (NL32x2 Switched) is 32 chips per rack. Project Rainier, AWS's earlier deployment of 500,000+ chips with Anthropic, was Trainium 2-based; Trainium 3 is the next generation and Anthropic is among the named launch customers. **Meta MTIA v3** is in full deployment for inference workloads (ranking, recommendations, ad serving). MTIA v4 is in fabrication. **Microsoft Maia 200** runs Copilot and Azure OpenAI Services internally. Microsoft also remains one of NVIDIA's largest Vera Rubin customers. The dual-track strategy is explicit, and the implication for power architecture is that Microsoft will operate both Mt. Diablo halls (for Maia) and 800 VDC halls (for Vera Rubin) inside the same data centre estate. When the operator of the workload also controls the silicon, the bargaining position with the rest of the stack changes. ## 6\. So which architecture wins? Wrong question. The right question: *which architecture wins for which kind of operator?* There's also a real engineering case for the row-scale unipolar approach that the supply-chain argument tends to bury. At 600 kW+ per rack, where Vera Rubin Ultra and similar accelerators are heading, the unipolar 800 VDC bus uses fewer conductors and less copper than the bipolar default. Mt. Diablo in its default three-wire config needs three conductors plus careful neutral handling; unipolar needs two. At extreme rack densities this matters for busway sizing, distribution complexity, and per-rack capex on conductors alone. NVIDIA's 800 VDC pitch isn't pure supply-chain naïveté. There is a real density argument underneath. The supply chain economics may dominate for the next five years; the density economics may dominate beyond that. ![Decision tree showing six operator archetypes](https://adikumar.co/content/images/2026/05/essay-2-fig6-operator-decision-tree.png) Figure 2.6\. Operator decision tree across six archetypes. With that in the picture, six operator archetypes, and the architectural answer is different for each. **The hyperscaler running its own silicon (Meta, Google, Microsoft, Amazon).** Mt. Diablo, with very high probability, as the primary architecture. They authored the spec, they want the EV supply chain on the bipolar default, and their compute platform is accelerator-agnostic by design. NVIDIA's row-scale 800 VDC architecture is the secondary architecture for this archetype. They will operate it where they host NVIDIA-tenant or Vera Rubin workloads, but their default new-build power architecture is sidecar. **The AI-specialist neocloud (CoreWeave, Lambda, Crusoe, Nebius).** Almost certainly NVIDIA 800 VDC. Their entire value proposition is "fastest path to NVIDIA capacity." They don't have the volume to influence Mt. Diablo's roadmap, they don't run alternative silicon, and time-to-revenue dominates everything else. The density advantage of unipolar 800 V matters here too as Vera Rubin Ultra deployment scales. **A note on stranded capex risk for any single-architecture bet.** NVIDIA's pattern across the Hopper-to-Blackwell-to-Vera Rubin transitions tells you something about the speed at which rack-level specs change. In three generations the rack jumped from \~50 kW air-cooled, to 140 kW liquid-cooled, to 600 kW liquid-cooled with 800 VDC, with each transition stranding prior facility-side investment. CBRE puts the liquid-cooling retrofit cost alone at $2-5M per rack row. If you build an 800 V Kyber row in 2026 and NVIDIA's 2029 generation shifts the voltage class again, or the mechanical interface, or the cooling spec, the building-side capex doesn't transfer cleanly. The chips amortize over 2-3 years; the infrastructure is supposed to amortize over 15-25\. That mismatch is a strategic problem for any operator whose architecture choice is locked to a single vendor's roadmap. It's a particular problem for AI-specialist neoclouds because their balance sheet cannot take on the asymmetry. It's a smaller problem for hyperscalers because they have multi-tenant flexibility and silicon optionality. It's why the enterprise "wait" answer below is pattern recognition, not conservatism. **The major colocation operator with mixed AI tenants (Equinix, Digital Realty, Vantage, NTT).** Hybrid as a product. Different halls for different tenants. Some operators will offer Mt. Diablo halls for hyperscaler tenants, some 800 VDC halls for neocloud tenants, and a lot of legacy AC for the long tail. The commercial sophistication comes from pricing each architecture distinctly. **The OEM / ODM (Supermicro, Wiwynn, Foxconn, Quanta, Inventec)** build both. Watch which lines they're tooling up: the ratio of capex going to ORv3 HPR-compatible chassis versus NVIDIA reference (MGX, Kyber) lines is one of the cleanest leading indicators of which architecture is winning by volume. Today the ratio splits roughly evenly across the major ODMs; the trajectory matters more than the snapshot. The ODMs have no architectural loyalty. They go where the tonnage is. **The enterprise running on-prem AI for regulated workloads (banks, defense primes, healthcare systems).** Probably neither, for the next three years. Stay AC, deploy a small DC island for the AI sub-zone, take the inefficiency hit. The skills, the certifications, and the supplier relationships aren't there yet, and the workloads (mostly inference, mostly small training runs) don't justify the architectural commitment. **The sovereign AI operator (national champions in EU, UAE, India, etc.).** Depends entirely on supply chain politics. If the priority is geopolitical independence from NVIDIA's ecosystem, Mt. Diablo provides a credible path. If the priority is parity with the US AI frontier, NVIDIA 800 VDC. Most operators in this segment hedge between the two. ## 7\. What to watch over the next eighteen months Three signals will tell you which architecture is gaining ground faster. **Component pricing parity.** Track the bill of materials for an 800 V SST versus a ±400 V bipolar rectifier of equivalent capacity. Today the 800 V part is roughly 15-25% more expensive at the component level. The gap will close, that's the direction the learning curve points, but the timing depends on annual unit volumes hitting the threshold where 800 V parts move down the experience curve at EV-class rates. Watch component supplier price lists. **ODM tooling ratio.** Same watch as component pricing. What fraction of new chassis tooling at Supermicro, Wiwynn, Foxconn, Quanta, Inventec is going to ORv3 HPR versus Nvidia reference designs. Quarterly capex disclosures and supplier earnings calls are the source. The ODMs have no loyalty; their tooling decisions are the ground truth. **Standards extensions.** UL 98B coverage extending cleanly to 800 V data centre topologies. IEEE 1709 next revision. IEC 61140 DC supplements. If standards bodies write language that explicitly accommodates both unipolar and bipolar, hybrid wins. If they prefer one, that one wins. ## 8\. The operator's question If you're the executive responsible for an AI infrastructure budget in 2026, the question isn't "which architecture should I bet on?" The question is: **given my workload mix, my supplier relationships, and my time-to-revenue requirement, which architecture costs me less to be wrong about?** For most operators, the answer is the architecture your biggest tenant or partner is already committed to. Inertia is rational here. The sophistication is in recognizing that you'll probably be operating both architectures by 2030\. And writing your supplier contracts now to keep that option open. --- *Essay 2 of 13 in The DC Transition series. Subscribe to get each essay as it publishes.* --- The DC-DC Transition · Navigate the series - **← PREVIOUS** · [Essay 1 · The real reason data centres are going DC](https://adikumar.co/the-real-reason-data-centres-are-going-dc-and-its-not-capacity-alone/) - **→ NEXT** · [Essay 3 · The architecture map](https://adikumar.co/the-architecture-map/) - **FRAMEWORK MAP** · [The AI Power Chain framework](https://adikumar.co/framework-ai-power-chain/) - **DEEP DIVE** · [Supplement B · Hybrid forever (quantitative scenario)](https://adikumar.co/dc-dc-supplement-b-hybrid-forever/) ## Sources - NVIDIA FY2025 data centre revenue ($115.2B): NVIDIA Q4 FY2025 earnings. - NVIDIA FY2026 data centre revenue (\~$194B): Silicon Analysts (April 2026 update) referencing FY2026 trajectory. - AI accelerator market share trajectory (peak \~95% → 80-86% in 2025-26 → \~75% by end-2026): consensus across Silicon Analysts, TrendForce (October 2025), SemiAnalysis Q4 2025 model, Bloomberg Intelligence (January 2026). - Custom ASIC CAGR (44.6%): Bloomberg Intelligence "AI Accelerator Chips 2026 Outlook Deep Dive," January 2026. - ASIC shipments crossing GPU shipments by 2028: Counterpoint Research (cumulative 40M ASIC chips deployed across top 10 hyperscalers, 2024-2028). - TPU v7 Ironwood 2026 shipment scale (\~4.3M units, cumulative tens of millions across 2026-2028): industry market sources covered in The Next Web (April 2026), Jon Peddie Research (December 2025). - Anthropic Ironwood commitment (>1M chips, >1 GW): Google Cloud / Anthropic joint announcement, late 2025. - Trainium 3 architecture (144 chips per UltraServer, 32 per rack in NL32x2 SKU, 3nm, GA December 2025): AWS re:Invent 2025; AWS Neuron documentation; SemiAnalysis "AWS Trainium3 Deep Dive," December 2025. - Project Rainier (Trainium 2, 500K+ chips with Anthropic): AWS, 2025. - ABB UL 98B 2 kV certification (July 2025, utility-scale solar application, UL 98B native test scope is to 1500 V): ABB press release, July 16 2025; pv-magazine USA, July 17 2025. - OCP Diablo 400 specification: Diablo 400 Project: Rack and Power v0.7.0 Base Specification, March 2026 (prior versions: v0.5.0 May 2025, v0.5.2 May 2025). - Sidecar vs row-scale architectural framing: Glenn K. Lockwood, "Mt. Diablo" reference note (May 2025); Diablo 400 v0.7.0 spec text on permitted output configurations. - Stranded capex risk framing: Hopper-to-Blackwell-to-Vera Rubin rack power trajectory (50 kW → 140 kW → 600 kW) per NVIDIA technical disclosures and IntuitionLabs deep-dive (March 2026); Goldman Sachs AI Infrastructure report (2025) on Blackwell facility-upgrade implications; CBRE Group infrastructure report (2025) on liquid-cooling retrofit costs ($2-5M per rack row). ## Frequently asked What 800V DC power rail solutions are hyperscale operators deploying? Two deployment paths are marketed as "800V DC". OCP's Mt. Diablo (Microsoft, Meta, Google, spec v0.7.0 March 2026) uses a rack-adjacent sidecar with a default ±400V bipolar output and a permitted 800V unipolar option in the same spec. Nvidia's reference architecture uses row-scale rectification with a unipolar 800V bus. Both paths are being coordinated inside the OCP 800VDC subcommittee. The deployment-level distinctions still drive retrofit, capex, protection regime and vendor stack. What are the differences between 800V DC and Nvidia 800V vs OCP Mount Diablo? The label "800V DC" covers two architectural approaches. OCP Mt. Diablo rectifies in a sidecar power rack adjacent to the IT rack, with a default ±400V bipolar output and a permitted 800V unipolar option. Nvidia's reference rectifies at row scale and hands a single 800V unipolar bus to the compute rack. The main differences: where rectification happens, whether compute is locked to a specific accelerator family, and the grounding and fault-clearing regime each entails. Can 800V DC be retrofit into existing data centres? The sidecar architecture is designed for retrofit. It converts existing 480 VAC to ±400V or 800V DC in a rack-adjacent power rack, preserving the upstream AC infrastructure. Native 800 VDC (MVAC to 800 VDC) is typically greenfield because it changes facility-level distribution. Retrofit economics depend on rack density and available AC capacity. Which vendors are named in the 800V DC ecosystem? Vendor participation in the 800V DC ecosystem is broad. Vertiv, Delta, ABB, Eaton and Schneider are named as ecosystem partners in the OCP 800VDC standardisation work. Wolfspeed and Infineon supply wide-bandgap semiconductors. Vicor supplies on-package delivery. Specific production-site deployments are generally not publicly disclosed. See the AI Power Chain vendor screen for named design wins.