Micron 9650 NVMeTM SSD now enters mass production; World’s first PCIe Gen6 drive to achieve this milestone
Yesterday, Micron officially announced that its Micron 9650 NVMeTM is now in mass production. This is the world’s first PCIe Gen6 SSD to reach this new shipping milestone.
High‑performance storage just reached a new milestone!
The Micron 9650 NVMe™ SSD is now in mass production, the first PCIe® Gen6 drive to achieve this landmark. In an AI‑driven world, storage is no longer a background component — it’s central to system performance, efficiency,… pic.twitter.com/6iwH5hpgJj
— Micron Technology (@MicronTech) February 13, 2026
Micron 9650 NVMeTM SSD
This SSD enters the market as the world’s first PCIe Gen6 data center SSD, delivering massive performance increases compared to prior generations. Check out the tablet below.
| Performance type | Micron 9650 | PCIe Gen5 SSD | Micron 9650 advantage |
| Sequential read | 28,000 MB/s | 14,000 MB/s | 100% higher |
| Sequential write | 14,000 MB/s | 10,000 MB/s | 40% higher |
| Random read | 5.5 MIOPS | 3.3 MIOPS | 67% higher |
| Random write | 900 KIOPS | 720 KIOPS | 22% higher |
Apart from this, the Micron 9650 is engineered to deliver industry-leading performance and strong performance per watt. At the same 25W power state, it delivers twice the performance of PCIe Gen5 drives. It achieves better energy efficiency for both random reads and writes, while faster data transfers at the same power result in lower total energy consumption.
This latest SSD supports both air-cooled and liquid-cooled configurations, recognising that storage can no longer be thermally isolated from the rest of the platform. It is revealed that over the past 18 months, Micron has conducted extensive interoperability testing across the PCIe Gen 6 ecosystem, paving the way for the rapid adoption of the Micron 9650.
Overall, the Micron 9650 represents a fundamental shift, from storage that tries to keep up to storage that keeps GPUs fed, removing bottlenecks, improving energy efficiency and enabling AI infrastructure to operate closer to its theoretical potential.
