REBEL
HBM3e chiplet-based scalable Al accelerator
for hyperscale workloads
GenAI for All,
Maximum Impact at Minimal Cost
Designed to power large-scale AI systems, REBEL achieves the highest level of energy efficiency while delivering top performance. Equipped with the latest HBM3E, REBEL demonstrates outstanding computing speed and throughput per Watt, making state-of-the-art AI services more accessible. REBEL gives companies the extra edge to stay ahead of the competition.
Performance at Any Scale:
from a Single Card to a GenAI Datacenter
Combine four dies onto a single IC chip, connected through the state-of-the- art advanced chiplet packaging. Scale beyond a single card to multi-rack configurations, able to support any multi-modal models for years to come. Achieve greater sustainability of accelerating AI workloads for the upcoming power-hungry GenAI datacenters.
See also...
ATOM™
Cost-efficient, Powerful AI Acceleration for Small-sized Data centers
ATOM™-Max
Boosted Performance for Large-Scale Inference
ATOM™-Lite
Low-power, Yet Highly Powerful AI Inference at the Edge
System Foundation
Start Lean, Scale Green
ATOM™ SoC
Inference AI Accelerator for Data centers
With international recognition from the industry-standard benchmark MLperf™ v3.0, ATOM™ can be scaled to accelerate state-of-the-art AI models of various sizes.
ION™ SoC
