Bestest RAID card?

Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
edited June 2003 in Hardware
For ~$100 or less, that is. I'm not willing to pay insane amounts of $$ for a RAID card- $110 max probably... so what's the best ATA-133 RAID card for around $100 or less? It has to handle 4 drives in RAID 0 (4x 160gb) and I've got a K7D-L it's going in, so if it can do 64 bit or 66MHz PCI, that's a plus. I'd also like to avoid ordering on line 'coz I want it NOW (yea I'm impatient), which means my choices are as follows:

HighPoint RocketRaid 133- $78
HighPoint RocketRaid 404- $98
Promise FastTrak TX2000 -$108

OR if I go to Newegg and wait a few days for shipping
My choices are:
HighPoint RocketRaid 133- $70.99
HighPoint RocketRaid 404- $90.99
Promise FastTrak TX2000- $85.00
Promise FastTrak TX4000- $103.00

So, what should I get? I COULD go with newegg and wait, if the TX4000 is worth it...

opinions anyone?

//EDIT
I'm going to run 2, 2-drive RAID 0 arrays, one primary and one backup, so I think I could get away with running one array as the 2 masters, the other as the 2 slaves, and not have a performance hit except in array -> array transfers... but I might be wrong...

Comments

  • TexTex Dallas/Ft. Worth
    edited June 2003
    The kd7 does not do 64bit or 66mhz. You need a dual cpu board with 64 bit slots to do that.

    The hpt is far better then the promise. Get the 4 port hpt if you can afford it.

    tex
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    Tex, it's a MSI K7D Master-L, not an Abit KD7
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    so.... who's better- promise or highpoint?
  • SlackerSlacker CA, USA
    edited June 2003
    wtf.. there's something seriously wrong with the logic here. Tex just posted (4 times) that the "hpt is far better then the promise" and Geeky1 then asks "so.... who's better- promise or highpoint?" Is it just me or do I need more sleep. :confused:
  • EnverexEnverex Worcester, UK Icrontian
    edited June 2003
    Geeky - Highpoint are better in every way way apart from one. Promise can run ATAPI devices (CD-ROM, ZIP etc) whereas the highpoint cant.

    Tex - Just need to press it once :p

    NS
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    Slacker, no actually it's me that needs sleep... I wasn't thinking when I posted that... sorry...
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    I just heard that the Highpoint 404 CAN'T run @ 66MHz... is that true?
  • Mt_GoatMt_Goat Head Cheezy Knob Pflugerville (north of Austin) Icrontian
    edited June 2003
    Friends don't let friends buy Promise controllers.

    I would go for HPT. I don't think you will gain that much on an IDE based array with 64 bit over 32 bit as you are closing on the theoretical max of the drive's design. I have been wrong before but think the HPT 404 would serve you best. That Texan tried talking me out of a Promise 4000 but I didn't listen and ended up returning it after 2 weeks.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    mtgoat:
    1. why did u return it
    2. we're talking about 4 drives with sustained data transfer rates of ~40-50mb/s, which is 160-200mb/s, and the only way to get that kind of speed is using 66MHz cards...
  • TexTex Dallas/Ft. Worth
    edited June 2003
    Chcek the HPT site. They have recently started making several cards in 66mhz mode bit not ALL cards. So when you buy you have to verify you are getting a 66mhz capable version also.

    Tex
  • Mt_GoatMt_Goat Head Cheezy Knob Pflugerville (north of Austin) Icrontian
    edited June 2003
    Originally posted by Geeky1
    mtgoat:
    1. why did u return it
    2. we're talking about 4 drives with sustained data transfer rates of ~40-50mb/s, which is 160-200mb/s, and the only way to get that kind of speed is using 66MHz cards...
    I returned it because I wasn't happy with the performance I was getting overall. I wanted to run 4 Maxtor 740's in 2 independent raid 0's (couldn't get anything higher than 67,000ish in ATTO). I ended up swapping mobos and did the same on an onboard HPT 372 with the results I was looking for (76,000ish in ATTO). Now I'm using a version of the same on an add on card (see sig) and almost hitting 100,000.

    BTW
    I thought you joined the SCSI club and went zoom zoom???
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    I've got SCSI... I use 2 IBM Ultrastar LVD-160/SE 10,000RPM 4MB hard drives @ 36.7 & 18.2GB. But they're slower than my 80gb Maxtor 7200rpm 8mb drives are, and the Maxtors aren't RAIDed (at least according to SiSoft's disk benchmark- I just D/Led ATTO and I'll try that on the 160gb after I get one of 'em installed)

    But here are the ATTO results for my 18gb SCSI:
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    BTW... I forgot to mention- the drives I'm RAIDing are 160GB, 8mb, 7200rpm Maxtor ATA-133 units... 4 of 'em.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    However, I'd love an explanation for these scores- 500MB/s? on a LVD-160 drive? WTF?!?
  • Mt_GoatMt_Goat Head Cheezy Knob Pflugerville (north of Austin) Icrontian
    edited June 2003
    Damn, that looks like a job for Tex support!!! He has been trying to get me into it but I just don't have the resources these days. I still think that you could get your SCSI shaped up you would be better off if you are looking for all out performance. If you are going for the 2 160GB Maxtors then I would do the 404.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    I may be able to get the scsis performance up, but compare that last atto run to this one, which is off ONE of the Maxtor drives that I threw in the K7D just to benchmark it:
  • EnverexEnverex Worcester, UK Icrontian
    edited June 2003
    Yeah, but as was mentioned in the other thread, thats not quite real drive performance as the memory on the controler is just caching the ATTO test data, so you are basically seeing the RAM of the board speed.

    NS
  • TexTex Dallas/Ft. Worth
    edited June 2003
    He also changed the options on the right side of atto.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    ah... so, what settings should Atto be run with? I realize it's not real-world performance, but...
  • Mt_GoatMt_Goat Head Cheezy Knob Pflugerville (north of Austin) Icrontian
    edited June 2003
    Originally posted by Geeky1
    ah... so, what settings should Atto be run with? I realize it's not real-world performance, but...
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    thanks... sorry for the delay, I've been having "technical difficulties" :mad:

    anywho, with those settings, compare my 18gb scsi (the 36gb is essentially identical in terms of scores) to one of the 160gb drives...

    here's the scsi:
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    versus ONE of the IDEs...
  • Mt_GoatMt_Goat Head Cheezy Knob Pflugerville (north of Austin) Icrontian
    edited June 2003
    WTF!!!

    I think you need something on that SCSI Atto! There is something seriously wrong.
Sign In or Register to comment.