ClusterX™GPU Cloud Rating & Ranking System

After evaluating a provider against the ten criteria (Security, Lifecycle, Orchestration, Storage, Networking, Reliability, Monitoring, Pricing, Partnerships, Availability) we assign one of these ratings. Gold and Platinum providers rise to the top by introducing features and functionality that others do not have.

Six of these definitions were written for the 2.x releases; the Participation Ribbon was added in 3.0. The board above shows current standings.

Platinum

The best GPU cloud providers in the industry. They consistently excel across evaluation criteria, are proactive and innovative, and maintain an active feedback loop with their users. In practice they command a pricing premium because their total cost of ownership is better even when the raw $/GPU-hr is higher.

Gold

Strong performance across all evaluation categories with some opportunities for improvement. Gold-tier providers may have small gaps or inconsistencies but are responsive to feedback. A great choice that generally wins deals, especially with the best $/GPU-hr and availability on the customer timeline.

Silver

An adequate offering with noticeable gaps compared to Gold or Platinum. Some users will not consider a Silver provider despite attractive pricing and availability. There is clear room for improvement, and we encourage adopting industry best practices to catch up to peers.

Bronze

Fulfills our minimum criteria, and the last tier we directly recommend. Common issues include inconsistent support, subpar networking, unclear SLAs, limited Kubernetes/Slurm integration, or less competitive pricing. Several providers here are making real effort to catch up.

Participation Ribbon

Introduced in 3.0 as a tier between Bronze and Underperforming. It more accurately describes our opinion that these providers do the bare minimum to get by.

Underperforming

Based on hands-on testing, these providers can quickly rise to Bronze or Silver by fixing one or more critical issues, such as offering only older GPUs, missing basic security attestation (SOC 2, ISO 27001), misconfiguring key server features (leaving PCIe ACS enabled, or failing to enable GPUDirect RDMA), or charging for GPU hours during cluster creation or hardware downtime.

Unavailable

An interesting service we are excited about but cannot verify yet: not launched publicly, sold out with no plans to add capacity, government-only, or otherwise untestable. We keep a “trust but verify” approach until we can complete hands-on testing.

1. Security

   Data protection, compliance, and infrastructure security measures including encryption, access controls, and audit capabilities.

2. Lifecycle

   Resource provisioning, scaling, and management capabilities including automation and self-service features.

3. Orchestration

   Container and workload management features including Slurm scheduling and Kubernetes support.

4. Storage

   High-performance, scalable shared storage that mounts reliably and protects data, including throughput, durability, and backup guarantees.

5. Networking

   Network performance, latency, and connectivity options including RDMA and high-speed interconnects.

6. Reliability

   Uptime guarantees, fault tolerance mechanisms, and disaster recovery capabilities.

7. Monitoring

   Observability, alerting, and diagnostic capabilities for infrastructure and workloads.

8. Pricing

   Cost transparency, value for money, and flexible pricing models including spot/spot-like instances.

9. Partnerships

   Ecosystem and integration support with major AI/ML frameworks and tools.

10. Availability

    Geographic reach, service accessibility, and capacity planning capabilities.

• Slurm

  Slurm is an open-source job scheduler and the de facto standard for HPC for over 20 years. Used on over 60% of TOP500 supercomputers and over 50% of AI training clusters.

• Kubernetes

  Kubernetes is the industry-standard container orchestration software. It is the de facto standard for inference and growing in popularity for training.

• Standalone

  Individual GPU machines provide direct hardware access for training, inference, data processing, and research workloads. Standalone systems offer maximum flexibility and performance without orchestration overhead.

• Monitoring

  A complete monitoring dashboard includes high-level and low-level views of the cluster, providing comprehensive visibility into system performance, resource utilization, and potential issues before they impact workloads.

• Health checks

  Proactive health monitoring identifies issues before they impact workloads through both active diagnostic testing and passive continuous monitoring that automatically remediates common problems.

1. ClusterX 3.0

   September 2026Current

   Managed GPU clusters evaluated through audit, performance, reliability, and fault-tolerance testing, covering 77 providers, with the market view widened to 323 providers and well over 200 end users interviewed. Only 19 neoclouds achieve a medallion rating this round, and a Participation Ribbon tier is introduced between Bronze and Underperforming.

2. ClusterX 2.1

   April 2026

   Added a small set of newly tested providers rather than a full re-test. Four moved into Bronze on initial evaluation, several serious providers moved into Unavailable ahead of launch, and the GPU cluster TCO and goodput expense calculators were published.

3. ClusterX 2.0

   November 2025

   Our second major evaluation, conducted from August to September 2025. Rated 84 neoclouds, up from 26, and tracked 209 providers. Interviewed over 140 end users, clarified and expanded all ten criteria with itemized requirements, and extended testing to Kubernetes clusters and standalone machines.

4. ClusterX 1.0

   March 2025

   The inaugural comprehensive evaluation of the GPU cloud market, conducted in Q1 2025. It established baseline ratings for major providers and defined the evaluation methodology.