RebelServer™
Powering AI Inference Efficiently and at Scale
Why RebelServer™
Sovereign AI
Deploy anywhere, without compromise. Run fully on-prem in existing air-cooled data centers with full control over data, security, and operations.
Ease of Use
Built for seamless integration. Works with open-source frameworks and industry-standard tools — no vendor lock-in, no new skills required.
Optimized for
Production Tokenomics
Designed for real-world AI economics. Delivers strong performance-per-watt, optimized for large-scale inference workloads.
Production-Proven
Solution
Proven at scale. Deployed across enterprise and government environments, supporting real-world AI workloads.
Built for
Efficient AI Inference
Designed to deliver high performance with strong energy efficiency, reducing power consumption and operational costs for large-scale inference deployments.
High-Performance System Architecture
Up to 2 PFLOPS FP8 performance with 8 RebelCards™ per server, each powered by a Rebel100™ chip with advanced chiplet architecture and high-bandwidth memory.
Highest Flexibility for AI Workloads
Supports a wide range of AI workloads, including large language models, MoE architectures, and multimodal applications across language, vision, and speech.
Ease of Use with Production-Ready
Software
Integrated with leading open-source frameworks such as vLLM, PyTorch, and Triton, enabling seamless deployment and optimized distributed inference.
Specifications
1x 100G 2-Port QSFP56 1x
10G/25G 2-Port SFP28 (OCP 3.0)
* Theoretical maximum power consumption based on specifications. Actual power consumption will not exceed 7 kW, typically hovering around 4-6kW at most under practical workloads.
Net Weight: 65.6 lbs (29.7kg)
Non-operating Temperature: -40°C to 60°C (-40°F to 140°F)
Operating Relative Humidity: 8% to 90% (non-condensing)
Non-operating Relative Humidity: 5% to 95% (non-condensing)