The hardware behind
every reservation.
ARO deploys current-generation NVIDIA GPUs in 8-GPU servers from Dell, Supermicro, and MSI, in liquid-cooled racks inside U.S. data centers, with 100 GbE fiber out to you. The NVIDIA B300 leads. The H200 is available where its memory profile fits better. The detail below is what we hand a tenant's technical lead.
One reference design, four building blocks.
Compute, network, cooling, and power. Each is engineered to data center thresholds, and each is tuned to the site and the workload before anything is ordered.
01 / Compute
NVIDIA B300 and H200
B300 with 288 GB HBM3e per GPU leads. H200 with 141 GB HBM3e where it fits. RTX, H-series, and B-series supported. Eight GPUs per server, NVLink-connected.
02 / Network
100 GbE backhaul
Redundant fiber paths out of every site. NVLink inside the server, high-speed Ethernet or InfiniBand between servers for multi-node clusters.
03 / Cooling
Liquid-cooled racks
Rear-door heat exchange and direct liquid cooling, sized to rack density. Motivair as the cooling partner.
04 / Power
415 V three-phase
Data center power with UPS and generator backing. Rack densities sized to the server configuration, not the other way around.
Current-generation NVIDIA, chosen for the workload.
ARO's lead GPU is the NVIDIA B300, the Blackwell Ultra data center GPU with 288 GB of HBM3e per GPU. Eight B300s in one server put more than 2 TB of GPU memory behind a single NVLink domain, which is what lets large models and long contexts run on fewer GPUs with less sharding overhead.
Where the memory profile or the budget fits better, we deploy the NVIDIA H200, with 141 GB of HBM3e per GPU and the Hopper software ecosystem most production teams already run on. Both ship as 8-GPU servers with NVLink connecting the GPUs inside the server.
ARO is not limited to these two. We also support RTX, H-series, and B-series GPUs for workloads that fit them, including smaller single-server deployments at our Micro Edge Data Hubs, and we keep pace with new NVIDIA generations through our hardware partners. The server vendor is chosen per deployment from Dell, Supermicro, and MSI.
Per-server configuration
- GPUs: 8 x NVIDIA B300 (288 GB HBM3e each) or 8 x NVIDIA H200 (141 GB HBM3e each)
- Interconnect: NVLink between GPUs inside the server; high-speed Ethernet or InfiniBand between servers for multi-node clusters
- CPU, memory, storage: specified per deployment and documented in your data sheet
- Vendor: Dell, Supermicro, or MSI, chosen per deployment
What you can run
- Production LLM inference and agents, large models and long contexts across a single NVLink domain.
- Multi-modal and vision, image, video, and audio inference with memory headroom.
- Fine-tuning and RAG, tenant-specific adaptation on isolated servers.
- Multi-node training, on dedicated clusters with a high-speed fabric between servers.
100 GbE out. NVLink in.
Dedicated capacity only matters if the network gets out of the way. Every ARO deployment lands on a 100 GbE tenant backhaul with redundant fiber paths, in multi-carrier facilities, onward to the public internet or to your private connection.
Inside each server, the eight GPUs are connected over NVLink, so tensor-parallel workloads scale across the server without saturating the host bus. Between servers, multi-node clusters use a high-speed Ethernet or InfiniBand fabric sized to the cluster.
For tenants with their own networks, we support direct connect and private VLAN handoff, useful when egress costs and data-residency posture matter more than internet-routed throughput. For regional Micro Edge Data Hubs, we design for sub-10 ms tenant-to-hub round-trip latency targets.
Liquid where it matters. Sensible where it doesn't.
Modern GPU racks run far beyond what air alone can carry. ARO deploys in liquid-cooled racks, with rear-door heat exchange as the baseline and direct liquid cooling where the server configuration calls for it. Motivair is our cooling partner.
At data center sites, the facility provides conditioned power, chilled water, and the mechanical plant. At Micro Edge Data Hubs, ARO installs a closed-loop system fed by the property's water supply, and the hub runs silent inside the room.
Power profile
- 415 V three-phase service per rack, with UPS and generator backing at data center sites.
- Rack power sized to the server configuration.
- Behind-the-meter generation or microgrid integration at qualified sites, where utility queue times are blocking grid-tied AI capacity.
- Sub-metered at Micro Edge Data Hubs so the property's electricity is reimbursed cleanly.
Water profile (rear-door cooling)
- Closed loop, 55–65°F supply temperature.
- Coolant chemistry within published Motivair limits (chloride, calcium, magnesium, sulfate ≤ 25 ppm; pH 7–10.5).
- 50-mesh inline filtration in the closed loop.
- Total water draw is per site; we share specifics during the design pass.
Sized to the site, from one rack up.
There is no fixed GPU ceiling. A reservation starts at one rack and grows with the rack space available at each site, across multiple sites when a single facility is not enough.
Data center deployments
Standard racks at 750 mm × 1200 mm, liquid-cooled, in professionally operated U.S. facilities with multi-carrier fiber. Capacity scales with the rack space contracted at each site.
Tier III-class facility redundancy. Facility specifications, available racks, and delivery windows are shared in writing during the reservation conversation.
Micro Edge Data Hubs
Smaller single-tenant deployments inside host properties, starting at a single 8-GPU server. The property contributes a secure room, three-phase power, water for the cooling loop, and a fiber path.
For workloads that need to sit next to their data. How the property program works →
Selected for what holds up in production.
No single-vendor risk. Each component is sourced from the supplier whose product is best for what it does, and the facilities and networks are run by partners who do that for a living.
B300 and H200 lead. RTX, H-series, and B-series supported. Hardware ecosystem reference; no other affiliation implied.
8-GPU PowerEdge servers with manufacturer warranty and support.
8-GPU server platforms with manufacturer warranty and support.
8-GPU server platforms with manufacturer warranty and support.
Channel partner for procurement and integration through Xerox's data center practice.
Facility operations and managed services behind the ARO NOC at data center sites.
Fiber connectivity to and between ARO sites.
Business connectivity at ARO sites.
Rear-door heat exchangers and direct liquid cooling for dense GPU racks.
Every reservation gets its own design.
Generic specs go only so far. Your deployment is sized to the racks available at the site, the power in the room, the fiber on the floor, and the workload your team intends to run. We put it in writing before you commit.
Request a deployment-specific data sheet- GPU and server recommendationB300, H200, or another NVIDIA GPU, and the server vendor, sized to the workload.
- Power and thermal envelopekW per rack, cooling method, facility specs.
- Network designFiber paths, inter-server fabric, VLAN handoff, direct connect, latency targets.
- Site, tier, and SLAFacility, delivery window, uptime targets, and response times.
- Compliance posturePartner audit reports available to you, isolation guarantees for the workload type.