11

In conjunction with the Super User blog, we will be getting three SSD's from Kingston. We want to perform tests on these drives and blog about the results, and need your help. What testing, comparisons, or how to guides, would you as a Super User like to see related to SSD technology?

Update:

The SSD's have come in and I have a total of 3 128GB drives. Here is the tentative plan for them:

  1. Test speed differences between a single SSD and a RAID 0 array:

    • File transfer from single drive to RAID 0 array and back
    • File transfer from single drive or RAID 0 array to a regular HDD 7200 rpm
    • File copy onto same disk
    • In all of these tests I'll be testing three different file sizes (small, medium, large, x-large: 4kb or smaller, ~12MB, ~2GB, and ~10GB)
  2. Test Speed differences between a regualr HDD, Hybrid HDD and a single SSD (clone a current copy of laptop HDD to Hybrid and SSD)

    • Boot Up times - Boot and measure difference in time (at least 5 boots each)
    • Application startup - Measure how long it takes for an app to start
    • File creation - a "copy" will be made onto the drive and tested for speed [In all of these tests I'll be testing three different file sizes [(small, medium, large, x-large: 4kb or smaller, ~12MB, ~2GB, and ~10GB)]

Let me know if this is good and/or if there are any other changes that you see should be done. I will be proceeding within a couple of weeks.

Testing begins this weekend! [26 Feb 2011] Cast your finals votes, or comments before then.

4
  • You lucky bugger - or are they expecting them back?... Right, time to tweet... "Testing workstations ~£5,000+ - anyone want me to test theirs?" Commented Jan 26, 2011 at 18:13
  • 2
    @Wil... My understanding is that this is a donation. :) Commented Jan 26, 2011 at 18:15
  • @Wil @Kronos gets to keep them :P @Kronos - you're welcome :P
    – Sathyajith Bhat Mod
    Commented Jan 26, 2011 at 18:46
  • 1
    "Will mailing SSDs in a jiffy bag to [my address] get them there in a working state?". I will complete the second part of the test with a long-term burn in using real world situations.
    – Linker3000
    Commented Jan 26, 2011 at 22:32

5 Answers 5

6

Depending on whether we get more than one drive, we do want to perform a test of a single SSD vs. a RAID0 setup of SSD's. We want to focus on read/write speed, and also management of the drives (i.e. TRIM) when they are in a RAID configuration.

4

Actually, clean install doesn't indicate much - try boot timings on clean install & clean install + Firefox + Chrome + VLC + MSFT Office on both standard HDD + SSD. That ought to show some real world testing.

For benchmarking:

  • EVEREST
  • HD Tune
  • Sandra

For logging boot times

Also, related questions

4

HDTach and ATTO can be useful for pure statistics, I often see reviews use those.

As for benchmarks, do not rely on the results of a single test, for example, you could use XBootMgr from the Windows Performance Toolkit which is available in the Windows SDK or at my mirror. This can automatize the process of doing multiple reboots and it will leave you with traces of each shutdown and boot and allows you to use XPerf to analyze these boots in detail, but basically you could just look where the logon phase ends. All this is explained in a handy on/off transition analysis document by Microsoft...

When using benchmarks that are also dependent on other things, try to minimize the impact. (eg. net I/O)

Also, be careful in which order you run tests, some tests might have an effect on subsequent results...

1
  • 1
    If order is an issue you do three in a row of the same condition and use the middle one or an average of the three. Or you just mix up all trials randomly and then take an average or the fastest boot time :-)
    – Ivo Flipse Mod
    Commented Feb 1, 2011 at 23:25
4

For ideas, I would check out Tom's Hardware for sure to see what they have compared and benchmarked in the past.

http://www.tomshardware.com/charts/ssd-charts-2010/benchmarks,110.html

I would throw a standard hard drive in the mix there just for comparison. I also like when sites include "high performance" versions of the other technologies. (WD Raptor for example).

With SSD's, it is all about those access times and potential high read/write speeds. I don't think people realize the vast difference between SSD'd out there and how they are not all the same. You might also consider mentioning SLC and MLC drives and some of the technical differences generally found between SSD's. There is also some focus on power savings, but see what Tom's Hardware found out.

I also like the way they do price per watt and price/performance comparisons to help determine what is the best for your money.

2

Definitely test boot times and read/write speeds on Windows 7...How much of an improvement, say on a clean install of Windows 7 (if possible, if not, or no machines free, no worries).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .