GenStorAIGE introduced its AI90 inference acceleration platform at WAIC 2026, using a storage-centric approach that incorporates PCIe Gen5 solid-state drives into the memory hierarchy. The platform combines HBM, system DRAM, and SSD into a unified three-tier memory structure, offloading KV Cache data onto SSDs to reduce pressure on GPU memory. According to GenStorAIGE, this cuts first-token latency from several seconds down to sub-second response times, representing up to a 50x improvement, with throughput gains reaching 5.1x and a roughly 39% reduction in GPU memory usage. Combined with intelligent peer-to-peer GPU communication, the company states AI90 can accelerate inference by up to 5.8x on systems running eight Nvidia GeForce RTX 5090 cards, effectively allowing an eight-card setup to behave closer to a 46-GPU cluster during sustained inference tasks. The architecture supports context windows exceeding 128,000 tokens. GenStorAIGE paired AI90 with its new PT200Z AI SSD, built using pSLC NAND flash and connected through a PCIe Gen5 x4 interface, delivering sequential read speeds reaching 14.8 GB/s, random read performance of approximately 3.1 million IOPS, read latency at 54 microseconds, and write latency at 10 microseconds, with endurance ratings reaching up to 100 drive writes per day.