You truly know that your a computer geek when you hollar “Woot” to your first RAID results and have to call up the wife and kids to see. Quite frankly I don’t think I will ever forget the look of confusion on her face as much as she won’t forget the look of amazement on mine. She felt a bit out of place when my son started telling me how amazing the speed was.
I locked myself in my office this past weekend in order to finally tackle something I have always been curious about, performance in a RAID environment and how easy it would be to construct such a thing.
After spending endless hours locked in the room, building, testing and ripping apart only to rebuild again, I have put together an article that I believe will show, not only the benefits and absolute amazing performance increase possible, but also how really simple it is to build.
To give this story the proper spin, I have to concede that I knew absolutely nothing about RAID before this experiment. Not only did I not know the basics of RAID, but also, I had never even once worked with a system that had RAID running. Now I know that RAID means Redundant Array of Independent Disks but as I was soon to find out in testing RAID 0, RAID is not really RAID at all.
In trying to explain RAID to me, a close colleague had to laugh in stating that, although RAID0 is RAID, technically there is no data redundancy so it really doesn’t qualify as RAID. Ok, I’m confused.
Rule # 1 – Lets keep it simple!
Imagine that a you had drive two cars from Point A to Point B. Now, imagine that there were two of you and you both drove the cars from Point A to Point B. It would take half the time right? Did you not technically just double your performance?
This is how RAID0 works. Rather than the entire file moving from Point A to Point B, its actually split into two and each part written to a different disk at the same time. Conversely, if it had to be read, it would be both disks being read simultaneously. Thats RAID0.
In just under 15 hours, far too many coffees and 70 performance tests that include Crystal DiskMark, ATTO Disk Benchmark, HDTune Pro and PCMark Vantage x64, I discovered enough about RAID0 to create a RAID0 build in less than 20 minutes.
A WORLD OF THANKS TO CRUCIAL
Our sincere thanks goes out to Kelly Sasso and the Crucial Team for lending 2x256GB Crucial C300 and 2x64GB Crucial C300 RealSSDs for our RAID Guide and subsequent testing of these drives. As you will soon find out, there truly is something special about these solid state drives.
TOOLS AND EQUIPMENT
- Printer to Print off Our Directions;
- 2x HDD or SSD (Preferably Crucial C300 RealSSD);
- Latest Intel RST Package found here;
- Windows7 on Disk or USB; and
Very little patience is required for this procedure as I will detail my activities step by step and explaining the performance enhancing tricks that extended my 20 minute installation to just under 15 hours. You get the best of it without the heartache and trust me, it is quick and easy.
STANDARD RELEASE OF LIABILITY
This is not a RAID instruction manual or guide.
This is a compilation of my experiences in building and testing RAID0 and contains a compilation of tips that were learned through RAID usage. The SSD Review doesn’t recommend anyone follow this in the building of their own RAID0 environment and doing so is, of course, done at their own risk.
OUR TEST BENCH
TEST BENCH, SOFTWARE AND PROTOCOL
TEST BENCH CONFIGURATION
|Asus Rampage III Gene SATA 6Gb/s USB3
|Intel i7-930 CPU@2.80Ghz (OCd to 3.2Ghz)
|6Gb G.Skill 1600 Triple Channel
|MSI N460GTX DDR5 768MB
|OCZ ModXStream-Pro 600w
|Cooler Master Hyper 212+
|Windows 7 64 Bit Ultimate
|Belkin Surf & Share Wireless USB Adapter
|Dell E228WFP 22? Flat Panel
|BitFenix Survivor Mid-Tower ATX Gaming Case (x)
|Front Loading Mobile SSD Rack2x64GB Crucial C300 RealSSD
CTFDDAC256MAG-1G1 (rev 0006) (x)
Software used for testing by The SSD Review consists of Crystal Disk Mark, HDTune Pro, ATTO Benchmark, along with FutureMark PCMark Vantage. The first three do a great job of showing us the numbers that we want to see, or don’t want to see in some cases, while PCMark Vantage x64 is an excellent program which recreates tests that mimic the average users activity, all the while providing a medium to measure each.
In order to ensure accuracy of tests throughout the evaluations, tests may be repeated or conducted in safe mode to confim consistency throughout and eliminate any possibility of a wild card result. Safe mode may be utilized as well as any other number of set criteria to ensure that tests conducted at present do not differ from those conducted in the past. Where comparative SSDs are still in the possession of The SSD Review, we will follow every opportunity to gain fresh test evaluations from sample media.
Pg2 – TRIM and Building RAID0