My Raid 0 Experiment

[LoD]

I was reading this article over at overclockers.com

http://overclockers.com/articles1297/index03.asp

That article inspired me to see how much of an impact Raid 0 makes on regular 7200 RPM hard drives. I have two identicle drives and needed to reformat and flush my system clean anyways.

The results are not that surprising for those that follow hardware news, but for anyone interested in raid this might be of interest.
**********
Hardware
**********

• A64 3200+ @ 2.6ghz
• DFI nF4 SLI-D
• eVGA 7800GT
• 4x 512MB OCZ PC3200 2-2-2-5 2T
• 2x Seagate 7200.8 250GB SATA NCQ Enabled 16KB Stripe Size

**********
Drivers and Software
**********

• Fresh Install XP Pro SP2, with all windows updates that can be installed without activating the OS.
• nVidia 6.70 Motherboard drivers, including nvidia IDE drivers. Only excluded the firewall part of the install.
• nVidia 81.98 Display forceware drivers
• WinRar
• Alcohol 120%
• Photoshop CS 8.0
• BF2 Patched 1.2
• HD Tach

The system was fully defragmented and rebooted before and after each test. I used a stopwatch to calculate times and took an average time with multiple tests when possible. (I didnt install BF2 3x times for instance, but for most of the remaining tests I ran them multiple times)

**********
Testing in Seconds
No Raid / Raid
**********
Bootup: 36.1 / 35.7
Shutdown: 11.5 / 11.7

Photoshop CS 8.0 Load: 10.2 / 9.2

1GB MPEG-2 File Copy: 51.3 / 47.9
700MB Xvid File Copy: 45.7 / 29.1
367MB BF2 1.2 Patch File Copy: 18.3 / 17.3

5.73GB Intense MP3/AAC Folder Copy: 317 / 287

1GB BF1942 CD Image Winrar Extraction: 53.6 / 42.2

BF2 Hard Drive Image Install Time: 125 / 92
BF2 1.2 Patch Install Time: 260 / 192

*********
Gaming
*********
BF2 LAN Server Host Creation 64 Man Daqing Map: 62.5 / 60.5
BF2 Single Player 16 Man Gulf of Oman Load Time: 54 / 55.5

*********
Conclusion
*********

I think its fairly obvious that Raid 0 has little impact on BF2 game load times. I probably could have tested more maps, but I did both the 64 man and 16 man single player tests multiple times and all the results came out within a half second or two. Those are the averages.

Human error of course played part in all this, as I used a stopwatch and relied on my eyes and reaction times, but hopefully this gives everyone an idea of the expected performance gains (losses) of using Raid 0.

File copying, and installing patches and imaged program installs were significants faster with a Raid array, but outside of that not much of an increase. Gaming definately does not benefit with standard run of the mill 7200RPM drives, but other windows activities do receive a small boost.

I can post some HD Tach screenshots later if anyone is interested.

thx

[NoSoup4You]

Re: My Raid 0 Experiment

Well duh, anyone could have told you this. RAID 0 has little to no impact on gaming applications or game load times. Where does it have an impact? Large file transfers. Its well known.

[[GU]LIZARD{GEN}]

Re: My Raid 0 Experiment

It may help in the verifying client data department.

[Espion]

Re: My Raid 0 Experiment

The most significant impact I found it had was on the windows load time and when unrar'ing files. Windows loads in under two scrolls of the blue bars on my machine, where as before RAID0 it took 20. 700mb files also unrar in ~5s.

[Heavy552 AWG]

Re: My Raid 0 Experiment

I use RAID for video work and can almost keep up with the Firewire in real time.

My current RAID is ok for gaming, the new AF no longer loads first into a map, in fact I load last with a extra 40 seconds verifying data despite a 6 down/700 up internet and a well maintained computer.

I will probably install the new Raptor 150's starting with two in RAID and beef up some other componets on the computer as I go until I achieve fast load times. I might have to move on to another board and CPU to crunch the data.

But the SATA RAID has been stable on my system with only one rebuild in two years (MBR's were intact) on the controller level.

[Roger Smith]

Re: My Raid 0 Experiment

whats the very best solution to bf2 load time, and game stability? i run an amd 3000+ which is like a 2.4 and a gig of ram, with a geforce 6600. is iram a possible solution?

http://www.giga-byte.com/Products/Storage/Products_Overview.aspx?ProductID=2180&ProductName=GC-RAMDISK

[Sir. MatDef]

Re: My Raid 0 Experiment

whats the very best solution to bf2 load time, and game stability? i run an amd 3000+ which is like a 2.4 and a gig of ram, with a geforce 6600. is iram a possible solution?

http://www.giga-byte.com/Products/St...ame=GC-RAMDISK

I hope you mean your 3000+ is equivalent to a p4 at 2.4ghz? In reality it runs at 1.8 (s939) or 2.0(s754), but is at least equal to a 2.6ghz o4.

[Roger Smith]

Re: My Raid 0 Experiment

heres a better question, iram or more ram?

[Komet.nl]

Re: My Raid 0 Experiment

Doesn't BF2 loads a 64 Player map when you will even play a 16 Player map I decided to take a look at the out of bounds with a chopper in Dragon Valley and what did I see? I saw the whole 64 Player map without vehicles in a 16 Player map what's the difference then between a 16/64 Map? I don't know the only thing I saw was that the vehicles weren't there..

Just a little experiment I don't know if this is actually true or not?

~ Komet

[Roger Smith]

Re: My Raid 0 Experiment

yes thats what i meant, either way it should be fine for bf2.

[imported_Wreck]

Re: My Raid 0 Experiment

Doesn't BF2 loads a 64 Player map when you will even play a 16 Player map I decided to take a look at the out of bounds with a chopper in Dragon Valley and what did I see? I saw the whole 64 Player map without vehicles in a 16 Player map what's the difference then between a 16/64 Map? I don't know the only thing I saw was that the vehicles weren't there..

Just a little experiment I don't know if this is actually true or not?

~ Komet

Yes, it's the entire map, but with fewer spawn points, vehicles, and a lower ticket count.

[Coderedpl]

Re: My Raid 0 Experiment

I have a Raid 0 setup with those 2 raptors