iPhone 18 Pro Max write speeds plummet to 1.1 MB/s as QLC NAND fills up, testing reveals
Apple's shift to QLC NAND storage in the iPhone 18 Pro Max enables higher capacities but introduces significant performance penalties under sustained writes, according to benchmark testing by Homolab.
The iPhone 18 Pro series launched last week with Apple deploying QLC NAND technology in higher-capacity variants, a choice that appears to exact a performance cost during intensive write operations. Testing conducted by the Bilibili channel Homolab on a 1TB iPhone 18 Pro Max revealed substantial speed degradation when the device handles sustained heavy data transfers. QLC architecture stores four bits per memory cell compared to three bits in TLC designs, enabling manufacturers to pack more storage into the same physical footprint. The trade-off manifests as reduced throughput, especially during large-scale data operations.
Homolab's comparison between the 1TB QLC-equipped iPhone 18 Pro Max and a 512GB TLC-based iPhone 18 Pro showed comparable performance in 4K read scenarios, but the QLC variant lagged by 38% in low-queue-depth mixed workloads, achieving 8,168 compared to 11,285 for the TLC model. Under more demanding conditions, the performance gap narrowed to a 12% disadvantage for QLC. The most dramatic differences emerged during sustained write testing. The iPhone 18 Pro Max initially leverages a rapid SLC cache delivering speeds up to 3,000 MB/s. Once exhausted, a secondary TLC-based buffer assumes the load, maintaining approximately 578 MB/s throughput. When both buffers deplete, direct writes to the underlying QLC NAND occur, plummeting to 79.4 MB/s on average, with occasional dips to 25.6 MB/s.
Performance deteriorates further as storage fills. At 60% capacity, the SLC cache shrinks from 250GB to merely 58GB. This contraction forces the secondary buffer to slow to 396 MB/s, while direct QLC writes average just 45 MB/s and occasionally hit 1.1 MB/s—speeds comparable to many budget microSD cards rather than flagship smartphone storage.
These measurements represent synthetic benchmarks designed to simulate worst-case scenarios. Typical usage patterns including web browsing, messaging, gaming, and photography should encounter no noticeable slowdown. Performance degradation would likely surface only during specific demanding tasks such as 4K or ProRes video recording, substantial file transfers, or data restoration operations. Homolab's findings nonetheless underscore the inherent compromise in deploying QLC NAND for high-capacity flagship devices, particularly for users whose workflows involve storage-intensive operations.