The article also tested the extraction of 108.000 files to the same volume, not exactly a sequential task i would figure. The AAE disk timed over 30 seconds faster than AAK in this task, and that includes many head movements. This is real time gained and will be quite noticeable since these are times when you as a user really 'feel' the performance of a drive. Waiting half a minute longer is a noticeable performance difference.
It's not just about sequential performance, if that was all...
That is not entirely true Christian Smith. The sequential I/O rate is very dependent on both read-ahead and write buffering. If you disable the write-back buffer on the disk you will get much lower throughput.
The firmware in question is probably suited to server environments where STR is not important at all, and IOps define real performance. Thus a sacrifice is made which makes the drive less suitable to desktop users. The article used real benchmarks measuring a stopwatch to complete tasks like copying, re-copying and extracting a RAR-archive with 108.000+ files. In all three instances, the AAE proved faster than the AAK disk. Sometimes the difference was quite big.
AAK was probably never meant to be sold to consumers, the fact that Seagate is withholding information is simply inexcusable.
Slashdot Mirror
The article also tested the extraction of 108.000 files to the same volume, not exactly a sequential task i would figure. The AAE disk timed over 30 seconds faster than AAK in this task, and that includes many head movements. This is real time gained and will be quite noticeable since these are times when you as a user really 'feel' the performance of a drive. Waiting half a minute longer is a noticeable performance difference. It's not just about sequential performance, if that was all...
That is not entirely true Christian Smith. The sequential I/O rate is very dependent on both read-ahead and write buffering. If you disable the write-back buffer on the disk you will get much lower throughput. The firmware in question is probably suited to server environments where STR is not important at all, and IOps define real performance. Thus a sacrifice is made which makes the drive less suitable to desktop users. The article used real benchmarks measuring a stopwatch to complete tasks like copying, re-copying and extracting a RAR-archive with 108.000+ files. In all three instances, the AAE proved faster than the AAK disk. Sometimes the difference was quite big. AAK was probably never meant to be sold to consumers, the fact that Seagate is withholding information is simply inexcusable.