Intel's Optane memory technology, discontinued in 2022, would have been ideally suited to the write-intensive AI inference workloads that emerged shortly after its demise, according to analysts.
Optane, based on 3D XPoint memory co-developed with Micron, offered sub-10-microsecond latencies and write endurance of up to 100 drive writes per day — specifications that remain unmatched today. However, the technology struggled commercially due to high pricing relative to NAND flash and was discontinued by Intel in 2022.
The same year, OpenAI launched ChatGPT, triggering a surge in AI development. The shift from training to inference has since driven demand for memory and storage, with modern inference engines caching key-value pairs to track model state — a write-heavy workload that strains conventional NAND flash.
Analysts note that Optane's endurance and low latency for small random writes would have made it a perfect choice for such caching, potentially easing the cost and supply pressures on DRAM and NAND.