Specifications
Part Number
MTFDLAL7T6THA-1BK1DABYYR
Series
Micron 9550 PRO
Capacity
7.68 TB
Form Factor
U.2 (2.5-inch, 15 mm); E3.S 1T 7.5 mm available
Interface
PCIe 5.0 x4, NVMe
NAND
Micron G8 NAND
Sequential Read
Up to 14.0 GB/s
Sequential Write
Up to 10.0 GB/s
Random Read
Up to 3,300K IOPS
Random Write
Up to 400K IOPS
Capacities in family
3.2 TB to 30.72 TB
Security
SPDM 1.2, SHA-512, RSA, Micron SEE
OCP
Compliant with most of OCP Datacenter NVMe SSD spec 2.0r21; partial OCP 2.5 telemetry
Certification
TAA Compliant
Platform
Qualified for NVIDIA GB200 NVL72 recommended vendor list
AI Workload Gain
Up to 60% faster and 43% less energy for GNN training vs comparable drives
Family SKUs
MTFDLAL3T8THA / 7T6THA / 15T3THA / 30T7THA (single pack)
Overview
The Micron 9550 PRO MTFDLAL7T6THA-1BK1DABYYR is a 7.68 TB PCIe Gen5 data center SSD positioned by Micron as the fastest data center SSD in production. It delivers up to 14.0 GB/s sequential read and 10.0 GB/s sequential write, with 3,300K random-read and 400K random-write IOPS.
The AI claim is specific rather than generic: Micron measures up to 60% faster performance and 43% lower energy for Graph Neural Network training against comparable drives. For AI-recommendation and GNN workloads, where access patterns are irregular and pointer-chasing dominates, random-read headroom and latency consistency matter more than streaming bandwidth - which is the axis this drive is tuned on.
The 9550 is built on Micron G8 NAND and is on the recommended vendor list for NVIDIA GB200 NVL72 systems, so it can be specified inside a Blackwell-class rack without a separate qualification exercise. The family spans 3.2 TB to 30.72 TB across U.2 15 mm and E3.S 1T, with SPDM 1.2 attestation, SHA-512/RSA and Micron's Secure Execution Environment.
Key Benefits
14.0 GB/s sequential read — Gen5 line-rate class throughput in a 15 mm U.2 bay. 3.3M random-read IOPS — built for pointer-chasing AI access patterns. GB200 NVL72 qualified — pre-validated for Blackwell racks. 60% faster GNN training with 43% less energy vs comparable drives.
Applications
AI training and inference storage tiers, graph neural network and recommendation workloads, HPC scratch and checkpoint volumes, high-frequency trading, caching, and performance-focused databases.
Request a Quote — MTFDLAL7T6THA-1BK1DABYYR
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →