Abstract:
Apple’s latest iPhone 18 Pro Max was recently discovered to have a controversial storage feature. Under extreme and sustained large-scale data writing loads, its internal storage speed will drop significantly, even to less than 100 megabits per second in some test scenarios. However, analysts believe that the vast majority of ordinary users will hardly encounter this situation during daily use.
This discovery came from a test by Bilibili technology channel Homolab, and subsequently attracted the attention of many hardware media. The test objects were the 1TB version of iPhone 18 Pro Max and the 512GB version of iPhone 18 Pro Max. The results show that the performance difference between the two is not obvious in ordinary reading and writing environments, but in some sustained high-intensity load tests, the 1TB version shows significantly different characteristics.
Researchers believe that the difference is mainly related to Apple's use of QLC flash memory in high-capacity versions. QLC is a four-layer cell flash memory. Each storage unit can record four bits of data, while traditional TLC flash memory records three bits of data. Since more data can be stored per unit area, QLC is beneficial for achieving higher capacities, but at the expense of write performance and endurance that are generally lower than TLC solutions.
In the 4K reading test, the overall performance of the 1TB version is quite close to that of the 512GB version, and even has a slight advantage. The gap between the two is about 11%. Therefore, in ordinary reading tasks, users will basically not feel any obvious difference.
However, when testing shifted to a continuous queue load scenario, performance differences began to show.
In the Q1T1 test project, which simulates a continuous single-threaded sequential task environment, the 512GB version scored 11,285 points, while the 1TB version only reached 8,168 points, with a performance gap of about 38%.
Even in the Q4T4 test with a higher degree of concurrency, the 512GB version still leads the 1TB version with 35,992 points and 32,219 points, a lead of about 12%.
What really caused concern was the sustained high load write test results.
During a long period of continuous writing, the storage speed of the 1TB version will gradually decrease. Test data shows that its average writing speed eventually dropped to 79.4MB/s, and even touched an extremely low level of 25.6MB/s at individual moments.
For a flagship smartphone, this data looks amazing. For comparison, some mid-to-high-end memory card products can also achieve similar or even higher continuous writing performance.
The researchers pointed out that this phenomenon is not a storage hardware failure, but a typical feature of QLC flash memory.
Because QLC needs to record more information in a single memory cell, the data writing and erasing process is more complex and the speed will be affected. Additionally, QLC is generally inferior to TLC flash in terms of long-term endurance.
In order to alleviate this problem, Apple introduced a high-speed SLC cache mechanism into the system. The SLC cache uses a single-layer unit storage method and has extremely high data throughput capabilities. When the cache space is sufficient, the write speed can reach approximately 3000MB/s.
In daily scenarios, whether taking photos, recording videos, installing applications, or processing ordinary files, most data will enter this part of the cache first, so users can usually get a very fast response experience.
The problem occurs after the cache is continuously exhausted.
When users perform very large write operations for a long time, such as continuously importing large amounts of material, executing long-term professional video workflows, restoring massive backup data, or continuously generating large volumes of content, the cache will eventually fill up. At this point, the system has to write data directly to the QLC flash memory, and performance degradation will appear.
Apple has not yet publicly responded to this test result.
However, many analytical institutions believe that this situation is mainly a result of laboratory-level stress testing. For ordinary users, even if they use iPhone 18 Pro Max to shoot high-specification videos, install large games, or process daily files, it is difficult to continuously trigger such extreme writing pressure.
In other words, most consumers will barely notice the difference during normal use.
For professional user groups, however, the situation is different. Creators who shoot ProRes videos for a long time, frequently process very large material files, perform large-scale data migrations, or rely on continuous high-speed writing workflows may need to pay attention to the changes in storage architecture between different capacity versions.
This test has once again triggered industry discussions on the application of QLC flash memory in mobile devices. As the storage capacity of flagship mobile phones continues to expand to 1TB or even 2TB, manufacturers need to find a new balance between capacity, cost, performance and durability.
For Apple, iPhone 18 Pro Max still has a powerful A20 Pro processor, advanced imaging system and record-breaking sustained performance. But the latest test also reminds the outside world that even on flagship products, there are still some easily overlooked technical trade-offs behind ultra-large-capacity storage solutions.
Comments