LIVING INDEX · NO. 2 IN THE SERIES
AI rack power by GPU generation
Living index of rack-scale power requirements across NVIDIA + AMD + adjacent AI accelerator generations. Categorical cooling requirement, 800V DC ecosystem readiness, integrator + power partners, per-row primary source. Quarterly cadence.
11Accelerator generations
7 kW → 500+ kWRack power range
QuarterlyUpdate cadence
v1.0Aug 2026
01Rack power comparison table
Note on undisclosed values. Where specific TDP or rack power is not publicly disclosed by the vendor, cell reads "Not publicly disclosed" rather than a fabricated estimate. Where an estimate is presented, it is attributed as such. NVIDIA Vera Rubin specifics have not been fully released as of August 2026; entries reflect what has been publicly announced.
| Accelerator generation | TDP per GPU | Rack config | Rack power | Cooling requirement | 800V DC readiness | Integrator + power partners | Primary source |
|---|---|---|---|---|---|---|---|
| NVIDIA H100 (Hopper) Shipped 2022 | 700W SXM5 | 8-GPU HGX / 4-GPU/8-GPU DGX | 30-60 kW typical (HGX H100 8-GPU) / 10 kW (DGX H100 single node) | Air (up to ~40 kW/rack); DLC required above | Ships in AC-powered racks; 800V DC racks introduced by ecosystem partners 2025+ | Foxconn, Wiwynn, Supermicro, Dell, HPE, Lenovo (multiple integrators) | NVIDIA H100 product page |
| NVIDIA H200 (Hopper refresh) Shipped 2024 | 700W SXM5 (same as H100) | 8-GPU HGX | 30-60 kW typical (HGX H200 8-GPU) | Air-capable; DLC preferred at 8-GPU HGX density | Ships in AC-powered racks; 800V DC via ecosystem partners | Same integrator ecosystem as H100 | NVIDIA H200 product page |
| NVIDIA B100 / B200 (Blackwell) Shipped 2024-2025 | 700W (B100) / 1000W (B200 dense) | 8-GPU HGX / MGX modular | 60-120 kW (HGX B200 8-GPU) | Liquid cooling standard (DLC or immersion); air-cooled variants limited | Ecosystem transitioning to 800V DC rack designs for Blackwell density | Foxconn, Wiwynn, Quanta, Supermicro, Dell, Delta Electronics (integrators + power) | NVIDIA HGX platform |
| NVIDIA GB200 NVL72 (Grace Blackwell) Shipped 2025 | 2700W per GB200 superchip (2×B200 + Grace CPU) | 72-GPU NVL72 (full rack), 36-GPU NVL36 (subset) | 120 kW per NVL72 rack (public NVIDIA guidance) | Liquid cooling required (DLC); air-cooled not a supported configuration | NVIDIA reference design assumes 800V DC facility feed; OCP Mt Diablo rack fits | Foxconn, Wiwynn, Quanta, Delta (800V DC power shelf), Flex (800V DC rack), CoolIT (thermal) | NVIDIA GB200 NVL72 + 800V DC ecosystem blog |
| NVIDIA Vera Rubin (announced) Shipped 2026-2027 target | Not publicly disclosed | Not publicly disclosed (successor to GB200 architecture) | Estimated 500-600 kW per rack based on public commentary; not officially disclosed | Liquid cooling required; NVIDIA guidance emphasises 800V DC + advanced cooling | 800V DC facility feed assumed; 1500V DC discussion emerging | Same partner ecosystem as GB200; specific product roadmaps not publicly disclosed | NVIDIA 800V HVDC architecture blog |
| AMD MI300X Shipped 2023 | 750W per GPU | 8-GPU OAM Universal Baseboard | 30-60 kW typical (8-GPU UBB) | Air-capable; DLC preferred at 8-GPU density | AC-powered racks primary; 800V DC via ecosystem partners | Supermicro, Dell, HPE, Lenovo (multiple integrators) | AMD MI300X product page |
| AMD MI325X Shipped 2024 | 1000W per GPU | 8-GPU OAM UBB | 60-90 kW typical (8-GPU UBB) | DLC standard at 1000W density | AC-powered primary; 800V DC via ecosystem partners | Same integrator ecosystem as MI300X | AMD MI325X product page |
| AMD MI350 series (announced) Shipped 2025 | Not publicly disclosed | Not publicly disclosed | Not publicly disclosed | DLC + advanced cooling expected | Not publicly disclosed | Not publicly disclosed | AMD Instinct product line |
| Google TPU v5 Shipped 2023 | Not publicly disclosed (Google internal) | Google internal pod architecture | Not publicly disclosed (Google internal) | Liquid cooling (Google DC standard) | Google internal DC architecture; not publicly detailed for TPU | Google internal + ODM partners (Quanta et al) | Google Cloud TPU |
| Cerebras WSE-3 Shipped 2024 | ~23 kW per wafer-scale engine | CS-3 system (1 WSE-3 per system chassis) | ~23 kW per CS-3 system | Liquid cooling required (integrated Cerebras thermal) | Standard AC input; not part of 800V DC ecosystem | Cerebras direct; system-level integration | Cerebras CS-3 system page |
| Groq LPU Shipped 2024 | Not publicly disclosed | GroqRack (server-scale) | Not publicly disclosed | Air-cooled variant available; liquid options | Standard AC input; not part of 800V DC ecosystem | Groq direct + system integrators | Groq product page |
02What this tracks
What this map ISN'T
Not a benchmark. Not a performance comparison. Not investment advice.
Not exhaustive of all AI accelerators (Cerebras + Groq included as representative non-NVIDIA/non-AMD points; other silicon vendors deferred to next revision). Not a substitute for direct vendor engagement or published product datasheets. Cooling + 800V DC readiness columns reflect ECOSYSTEM CONFIGURATION rather than accelerator-inherent limits.
03Method + sourcing
How this index is built
- Coverage. Includes accelerators with public product-page presence + verified generation. Each row cites a canonical vendor source.
- TDP. Per-GPU thermal design power as publicly disclosed by the vendor. Blank or "not publicly disclosed" where vendor has not released.
- Rack power. Reflects typical rack configurations (e.g. HGX 8-GPU, NVL72 full-rack). Range spans lower-density variants + higher-density configs.
- Cooling requirement. Categorical: air-capable, DLC preferred, DLC required, liquid required. Reflects ecosystem-standard configurations, not necessarily accelerator-inherent thermal limits.
- 800V DC readiness. Whether the accelerator generation ships in racks that assume 800V DC facility feed, OR is designed for AC-input via traditional PDUs. Ecosystem partners (Delta, Flex, CoolIT) determine actual deployment configurations more than the accelerator vendor.
- Update cadence. Quarterly. Between-cycle updates when NVIDIA/AMD publish new accelerator generations or significant ecosystem announcements.
04Change log
| Date | Version | Change |
|---|---|---|
| 2026-08-21 | v1.0 | Initial public release. |
Definition A living index of accelerator generation rack power envelopes across the AI compute buildout, from the H100 baseline (~40 kW/rack typical) through Blackwell GB200 NVL72 (~120 kW/rack) and forward to Rubin and Rubin Ultra deployments where per-rack power crosses 500 kW. The table is refreshed as hyperscaler specifications and public disclosures update.