Intel X25-M 80GB SSD, Intel Ups The Ante

Our Test Methodologies:  Under each test condition, the Solid State Drives were installed as secondary volumes in our testbeds, with a standard spinning hard disk for the OS and benchmark installations.  The SSDs were left blank without partitions wherever possible, unless a test required them to be partitioned and formatted, as was the case with our ATTO benchmark tests. Windows firewall, automatic updates and screen savers were all disabled before testing. In all test runs, we rebooted the system and waited several minutes for drive activity to settle before invoking a test.

The IOMeter Question:
As we noted in our recent SSD round-up article, though IOMeter is clearly thought of as a well respected industry standard drive benchmark, we're becoming increasingly uncomfortable with it for testing SSDs, as well as comparing their performance to standard hard drives.  The fact of the matter is, though our actual results with IOMeter appear to be accurate, it is debatable whether or not certain access patterns, as they are presented to and measured on an SSD, actually provide a valid example of real world performance, at least for the average end user.  Here's a sampling of our test runs with IOMeter version 2006.07.27 on our SSD sample lot.  We should note that these tests were performed on a Windows XP installation (separate OS drive) to eliminate any variability related to Windows Vista.  All other tests in this article were performed with a Windows Vista installation (also on a separate OS drive).

In the tables above, we're showing two sets of access patterns; one with an 8K transfer size, 80% reads (20% writes) and 80% random (20% sequential) access and one with IOMeter's default access pattern of 2K transfers, 67% reads and 100% random access.  What you see in the table above is an example of how random write operations kill I/O throughput of SSDs in IOMeter.  There is no question random write performance is the Achille's Heel of MLC SSDs, though SLC-based SSDs have a much easier time with it.  However, as you can see in the table, Intel's dynamically adjusting SATA to Flash controller on the X25-M SSD, alleviates this bottleneck completely and allows the SSD to adapt to the workload and access pattern of IOMeter.  As such, Intel's SSD blows all the other SSDs we tested right out of the water.

As we mentioned earlier in our discussion of Intel's adaptive memory controller technology, after these IOMeter tests, the Intel SSD had to be "re-conditioned" back to a factory-shipped state with a secure erase procedure, since these rigorous tests flood the drive with a very specific access pattern.  Regardless, the question is, do these results equate linearly to a real-world performance advantage of this magnitude?  As you'll see in our forthcoming benchmarks, Intel's new SSD certainly puts up very impressive numbers in all of our tests, but not by a factor of 100 over the next fastest SSD.

HotHardware Test System Intel C2E Powered

In our SiSoft SANDRA testing, we used the Physical Disk test suite. We ran the tests without formatting the drives and both write and read performance metrics are detailed below.  Please forgive the use of these screen captures and thumbnails, which will require a few more clicks on your part.  However, we felt it was important to show you the graph lines in each of the SANDRA test runs, so you are able to see how the drives perform over time and memory location and not just an average rated result. 

In terms of write performance, the Intel SSD again offered very flat, clean spike and valley-free performance across its entire 80GB of space, but it couldn't quite keep up with OCZ's standard SLC-based (Samsung) drive.  Intel's SSD fell behind the leader by about 17% or so but kept a nice linear pace with the others in our test group.