General Compute to Offer Positron AI's Memory-Optimized Inference Systems Through Cloud Partnership

General Compute has agreed to purchase and resell Positron AI's Atlas inference systems, creating a heterogeneous cloud infrastructure that splits AI workloads between GPUs and specialized memory-optimized accelerators. The multi-year deal, announced October 20, 2026, positions General Compute to differentiate itself in an increasingly commoditized GPU-dominated cloud market by offering faster and more efficient token generation for large language model inference.