Sign In
Register
Help
Nothing is set in stone and we are still working out some of the kinks. If you have a suggestion and it is something we could add to the current testing scheme feel free to let us know. We would love to have longtime StorageReview members help shape the site going forward. So far out reviewing scheme includes the following items.
Performance Tests:
- HDtune: Read test, write test, random access test
- HDTach: Long bench
- CrystalDiskMark: 100MB test on all parameters
- Iometer: Server scenario and workstation model
- Passmark V7
- PCMark05/Vantage
Usage Scenarios
The step which we feel one of the most important and also helps keep us ahead of the curve is working on usage scenarios and recording the I/O performance during that run.
HTPC with TV Tuner: Measure IOps of a predefined scenario of a movie playing, HDTV tuner recording one or more video streams, and serving media files to another computer over a network.
Gaming System: Measure IOps of predefined scenario of multiple games played for a set period of time. Manually triggering new map load sequences.
Power User: Still working on this one
All Drives: Transfer of 30-40GB of mixture of documents, media files, etc.
Recording the disk I/O performance during these scenarios is not a problem. Xperf is great for this. What we need is help finding a suitable application that will let us record various scenarios and play them back bit for bit in the sequence we define. This is the area we are having trouble with as it’s not exactly “off the shelf” software. Suggestions you guys (our readers!) would be GREATLY appreciated.
Power Consumption Tests:
- Idle: 5v average watts, 12v average watts
- Peak Startup Current: 5v peak startup, 12v peak startup (First stage is using slow reaction multimeters and averaging out the results, second stage is using a current sensing circuit and oscilloscope for high resolution measurements)
- Sequential Read: 12v average and 5v average watts over the duration of the HDTune Read Test (logging starts after test starts, logging stops before the section that measures random access starts)
- Sequential Write: 12v average and 5v average watts over the duration of the HDTune Write Test (logging starts after test starts, logging stops before the section that measures random access starts)
- Random Read Access: Sequential Read: 12v average and 5v average watts over the duration of the HDTune Random Access Test (logging starts after test starts, logging stops once test stops)
- Power Consumption Metric: Details are still being honed out but we feel confident we will be able to work out a metric that works in “normal usage” weighting for each category to give each drive a score based on real worked power situations over a given day. We feel this is important with the current Green trend
Thermal Tests:
IR Temp Readings: this is secondary to power reading, but stuff some people still want to know. Readings will be taken with an IR gun in multiple locations and to keep results consistent all drives will be tested in an external drive bay with a known volume and known airflow CFM with a room temperature reading listed with each test.
Thermal Probe: Measuring how hot the chamber gets through heat dissipation. Useful to know if the drive is shedding its heat through the air or through the chamber walls.
Hottest Component: Using the IR gun to locate the hottest part of the drive, be it the case, the cover, or some random electronic component on the board.
Noise Tests:
Localized Sound Readings: We are not trying to reinvent the wheel or compete against other sites that go incredibly in-depth in this area. We will be working with our own subjective sound measurements and including a sound recording taken through a device placed directly in contact with the drive. As the site grows we may consider investing more in this area or branching out.
Metrics:
Since we will be cataloging the same data for hundreds of drives… what better way to use this information than making helpful metrics to compare drives on various levels. These will not be listed until we have enough drives in each category to give useful ratings. Once we hit the five or ten drives reviewed mark this will be included. We will use that time to calculate a useful weighting system to get useful ratios between the given categories we feel are important for each drive type
Best HTPC Drive: Performance, power consumption, noise
Best Storage Drive: Performance, Price
Best Workstation Drive: Performance
Cheapest Drive: Price/Capacity
Green Drive: Power
Most Reliable Drive: Putting the reliability database to some good use for editorial content. We will have to work out a method to disregard legacy drives, drives with too few reviews, and also work in sample of user reviews from retail sites. If one particular drive is taking heat on NewEgg or in our forums we want to make sure this isn’t overlooked.
Overall Testing Methodology
The *MOST* important part of our review process is a consistent test bench used for every single drive review we publish. We will post specs once the system is finished being added to, but at its core it’s a Dell XPS 9000. We will be working with quite a few freelance writers for content going forward but one aspect that will never change is where the drives are benchmarked. Any and all equipment will be shipped to where our test bench is located and have tests performed on it before content is published.
Something tells me this XPS isn't stock...
The last and final part is what drives and hardware we will be reviewing. Going forward everything is fair game but some items will be overlooked in the short term while we work on getting our review process honed out and content generated on a daily basis. Items are ranked in terms of priority:
1. Internal Consumer 3.5” Spinning Hard drives
2. Internal Consumer 2.5” SSDs
3. External Storage Media (USB 2.0, USB 3.0, eSATA)
4. NAS/Windows Home Server Units
5. Flash Storage
6. RAID Storage (This is currently at the bottom while we work on getting our bread and butter reviews out first. It’s a chicken or egg situation as well since until our site traffic goes up it will be difficult to acquire RAID cards for review and most are well out of a useful price range if we had to buy every single one ourselves to test. )
We hope this is the stuff all of you wanted to hear. We have been going over every recommendation you guys have mentioned since we announced the transition to new ownership and tried to work in most of it. We can’t promise we will be able to include everything but our goal is to get very close in the next few months. Some of this will also require the help of our readers (You guys), so please contribute to make this site successful going forward!
MultiQuote
