Yotta NM1 Navi Mumbai
- Uptime Institute Tier IV certified
- Dual utility feeds from separate substations
Blackwell · SXM
NVIDIA’s Blackwell flagship: more memory and much higher throughput than Hopper, at up to 14.3 kW per server.

Allocation Lead time, October 2026: 8–16 weeks with an OEM allocation; 30+ without · All GPU lead times
Good fit for

The hardware
8-GPU HGX server, 1.4 TB of GPU memory. 180 GB HBM3e per GPU, up to 1,000 W per GPU, 4u liquid to 10u air of rack space.
Hosting
At up to 14.3 kW a server, B200 needs a high-density placement. We qualify each B200 deployment individually, including whether air cooling is enough for your configuration.
How managed hosting worksNVIDIA rates the DGX B200 at up to 14.3 kW, roughly 40% more than an H100 server.
8-GPU B200 servers are available air-cooled. Denser racks and some OEM designs use liquid cooling; we confirm what your configuration needs.
Supply moves quickly. We quote lead time and availability per request rather than publishing stock.

Rack, power, cooling, network and remote hands for servers you own.
Learn more →
Lender-ready quotes and asset records for your bank or NBFC.
Learn more →
kW, amps and tons of cooling for every GPU class.
Learn more →
Tested, valued and sold, or traded up for newer GPUs.
Learn more →