Sequential Performance – HDTach, HDTune, File Copy, YAPT (sequential)

We have shifted over to combining our results into two groupings for consumer reviews. First up is sequential performance:

HDTach:

HD Tach will test the sequential read, random access and interface burst speeds of your attached storage device (hard drive, flash drive, removable drive, etc). All drive technologies such as SCSI, IDE/ATA, 1394, USB, SATA and RAID are supported. HDTach tests sequential performance by issuing reads in a manner that was optimized more for HDD access, but this unique method has proven useful in evaluating the sequential response time of SSDs. The accesses are relatively small in size (2k), and are issued with a single working thread (QD=1). The end result is that devices with relatively IO high latency will not reach their ultimate rated speed.

hdtach-avg.png

Good QD=1 sequentials from the Vector 180, with slight drop at the 240GB capacity.

HDTune:

HDTune tests a similar level of features as compared with HDTach, but with a different access pattern. Thus provides us with an additional set of benchmark numbers to compare between storage configurations. CPU utilization has proven negligible with modern processing horsepower, and is no longer included. Additionally, we do not include write performance due to HDTune's write access pattern not playing nicely with most SSDs we have tested it on.

hdtune-read.png

The lowest capacity Vector sees the lowest read speeds and most variance during those reads (compare directly against the equivalent capacity Crucial M550 256GB).

PCPer File Copy Test

Our custom PCPer-FC test does some fairly simple file creation and copy routines in order to test the storage system for speed.  The script creates a set of files of varying sizes, times the creation process, then copies the same files to another partition on the same hard drive and times the copy process.  There are four file sizes that we used to try and find any strong or weak points in the hardware: 10 files @ 1000 MB each, 100 files @ 100 MB each, 500 files @ 10 MB each and 1000 files at 1 MB each.

pcperfc-create.png

The tool that does the file creation does so in a single threaded manner, so that that into account when considering the above results. You may note a steadily longer duration for the increasing capacity Vector 180 samples in this test – and those increses do not happen as badly at the largest transfer size. This is because the Vector 180 essentially stops all writing where a file deletion has recently taken place, and the size of the files deleted determines how long it stalls. In the sequence above, the 100@100MB test starts after deleting the 10x 1000MB files from the prior test, and the Vector 180 halts all writes as it TRIMs the data, causing the abnormally long times to complete each subsequent test. The 100@100MB test is bad for the 960GB Vector 180 just about as slow as a 1TB VelociRaptor (yes, a HDD) in that test and all three capacities of the Vector 180 are all *slower* than the HDD in the 500@10MB test. More on that later in this article.

pcperfc-copy.png

File copies uses a batch copy from within Windows, and sees added threading compared to our file creation test. These take place long enough after the TRIM operations that all SSDs turn in reasonably consistent figures here.

YAPT:

YAPT (yet another performance test) is a benchmark recommended by a pair of drive manufacturers and was incredibly difficult to locate as it hasn't been updated or used in quite some time.  That doesn't make it irrelevant by any means though, as the benchmark is quite useful.  It creates a test file of about 100 MB in size and runs both random and sequential read and write tests with it while changing the data I/O size in the process.  The misaligned nature of this test exposes the read-modify-write performance of SSDs and Advanced Format HDDs.

yapt-seqreads.png

yapt-seqwrites.png

Back when we looked at the Radeon R7, we figured those few dips in performance were just an anomaly. When we add in the results of the Vector 180s, we see more seemingly random drops in performance. We will explain these anomalies later in this article.

« PreviousNext »