Everspin Technologies announced that it demonstrated what it calls the world’s first Compute Express Link (CXL)-connected MRAM platform at the SNIA Developer Conference 2026. The platform creates a persistent memory tier between traditional DRAM and NAND storage, targeting a key computing bottleneck.
CXL connectivity allows Everspin’s MRAM to reduce the time processors and accelerators spend waiting for data, potentially helping AI systems run at higher utilization.
The demonstration used an AMD EPYC processor and an AMD Alveo FPGA platform. Everspin said the open standards-based solution is ready for customer workload testing.