Raid'n speed

[Vag The Duke]

I have a few questions regarding raid0, and I'm hoping someone in here is able to answer them:

1. Does windows boot up faster?

2. Do I get higher framerates in games?

3. Does games or applications install faster?

4. Does games or applications load faster?

5. Does the defrag end up faster?

6. Does the installation of windows xp go faster?

The reason I ask these questions is whether it's better to lay the money on game-related stuff if you're a gamer.

Thanks in advance Vag...

[Wombat]

Y
shorter pauses when you swap
depends if your CD is the bottleneck
"
hell yes.
dunno
didn't care

How much money are you talking about?

[Vag The Duke]

About 400$...

[Wombat]

No, for the raiding ability?
I got two D740s because I wanted that much (160GB) space and they are inexpensive. The Promise card ($70) was so I could get a little speed while I was at it, and not be tied to a MB solution.

[RichL]

I've got 2 Maxtor 30gb drives, and took care to benchmark them with SiSoft Sandra before and after RAID0-ing them together.

Pre-RAID HDD Benchmark (1 drive only) was 19925.
RAID 0 HDD Benchmark was 40785.

But more importantly -
System stability before RAID - 6 months, no crashes.
System stabilty with RAID 0 - 36 hours, BSODs cumulating in fried registry and finally totally unrecoverable system failure, requiring a full reinstall.

And that's why I'll never use a Highpoint RAID controller ever again.

[Vag The Duke]

Use only the Promise Cards....

[Greebe]

I'm sure Doc will chime in shortly

[Indiana]

I have a Highpoint controller and this one is working just fine.
A benchmark on a really cluttered partition (didn't really have time to defragment) with two IBM HDs - they're nice as long as they're working


No stability problems at all, not even during a two hours uncompressed video capture (which is stressing the RAID and HDs a lot). Most of those reported Highpoint problems are with older Highpoint chips prior to the HPT370.
Last time I rebooted was because I installed new leaked Radeon drivers.

[Paulr]

Any chance of you RAID owners giving us some benchmarks in something different, HDTach or ATTO for example?
Sandra is a tad.....ummmm.....well lets just say some people find it gives RAID 0 arrays a lower score than a single HD.
Also most Sandra benchmarks are "one off's" run it enough times and everybody will eventually get that "killer score" even if only once.
HDTach & ATTO seem to be a lot more consistant in their results.

[Dr Mordrid]

I'll use a Fasttrak100 for comparison as the TX4 is in a whole different class in that it has a master for each drive instead of master/slave's like the Highpoint.

This test was done with the RT-2000's HDD benchmark program and the NTFS filesystem on a PIII/850 system I'm no longer using. The setup is four IBM 75GXP drives (before I got rid of 'em) in a RAID0;



For compartison: the TX4 delivers over 94 mb/s single writes with four Maxtor D740X drives in a RAID0, mainly because of its different archetecture.

The dual stream test may be interesting for those who don't do realtime video editing. It is the throughput when reading two video streams off the same drive simultaneously. Talk about something that really tests your setup

HDTACH & co. may be OK if you're testing for database use, but I've seen them deliver numbers that are too low. This is really apparent when HTTACH says I shouldn't be able to do something I've been doing every day for months

Dr. Mordrid

[xortam]

I found the synthetic benchmarks to be worse than useless. The information was many times contradictory to the results I received with the ZD Disk WinBench application benchmark when testing various SCSI RAID 0 arrays. You really need to test your system according to how you're going to be driving the array. You may find that you have to find the right compromise because various applications have very different RAID needs.

[Technoid]

I actualy got a big boost when I put the Swap file and Temp dir on the Raid!
Nowdays I have W2K on the normal drive and XP on the Raid and the speed diferences is between having XP on the primary drive and the Raid is small....

[Indiana]

All those benchmark programs have their flaws, but I find Sandra not to be worse than HDTach or others.
Here is an ATTO score of my RAID0 with two IBM IC35L040 drives:

[Paulr]

I owned a TX4 for a whole two days before it had to be returned.
Appears Promise has an issue with them on Pentium 4 based systems - Windows installation will not see the RAID array even though you go through the normal F6 setup procedure and specify the drivers.

It was bloody annoying as I really wanted my 4 HD's on seperate channels, but until Promise sort that problem out I'm having to run my RAID 0+1 array off of a TX2.

[GT98]

In Reguards to the Boot up...I think my Work machine (P3-700, no raid, Win2000) boots up quicker then my home machine in my sig, after it goes through the BIOS on the Mobo then on the RAID card. OS loads are about the same I guess.