• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Adaptec AHA-2940UW

T

TwiceSliced

Gawd
Joined
May 1, 2002
Messages
525
I often find myself receiving donations of older systems in my line of work, and the other day I found the following in one machine:

Adaptec AHA-2940UW
Western Digital Enterprise 9100 9.1GB HDD
IBM DDRS-34560 4.3GB HDD

Are these of any use? Ive got this old 800MHz Celeron machine I've been meaning to turn into a Linux box, and I'm wondering if those drives would offer any sort of benefit over a standard IDE drive (e.g. 7200RPM, ATA100).

Thanks.
 
I can't make as easy a blanket statement about 5400RPM ATA drives now, can I? :p
Definitly depends on what drive we are talking about, but really anything recent, like the last 2-3 years, will outperform those drives. they are dinosaurs
 
Thanks again, and just one last question if you don't mind (sorry; I know very little about SCSI, and new questions keep coming to me after the fact)... Is the Adaptec controller card worth hanging-on to? Here's the link to Adaptec's spec sheet.
 
I have one built into my Tyan S1862 Dual PIII Server system. Works solidly and reliably with my Fujitsu MAP, 2x 73GB Seagate Cheetahs, and cast off 18GB IBM SCSI drive. As an added bonus, drivers for it come on Windows 2000, and presumably XP and 2003, CDs out of the box - no F6 disk required :) Coupled with some drives off Ebay, it would be a great way to try SCSI if you ever get the itch.
 
I've got that card in all my machines and it works great.
I use SCSI for my system drives and IDE drives for mass storage.
 
One last question: I've come across a Seagate Cheetah ST39102LW (9.1GB, 10000RPM, 5.4ms, "Ultra2 SCSI Wide")... How would that compare to a 7200RPM ATA100 drive? It has a much shorter seek time, but what about the data transfer rates? (potential 40MB/sec SCSI versus a potential 100MB/s ATA100? Or do I have that all wrong?)
 
This, unfortuantely, is a very common myth - "Lower seek time - It's faster, right?"

On the desktop, seeks of more than a few tracks are not all that common. On an ATA drive, the drive's firmware and buffer work to suck up data before the user requests it. While a SCSI drive does have a (smaller) buffer, the drive's firmware is not nearly as smart with it. It will not read ahead as well as a recent consumer drive does.

When running programs on the desktop, the majority of the files being accessed for a particular program are in a very small area of the disk. Many games and programs will acess very small files - My UT2004 directory has over 1000 files of less than 64KB in it. If you have an 8MB buffer drive with 6 of it for read ahead, you can have 96 of those files entirely in the buffer.

Just like SCSI drives use command queuing to mask the mechanical latency of the drive, ATA drives use read ahead buffer strategies to achieve the same goal - grab the data when it's convenient for the drive, squirrel it away in the buffer, and have it there for fast access when the system does request it. However, unlike command queuing, an ATA drive that is reading ahead doesn't know which files will be requested - it has to guess. The people that design the firmware have to analyze typical desktop access patterns and train the drive to know what to look for. That's why a drive with a fairly high seek time (such as the 16MB Maxtor drives, at over 9ms) can school a SCSI drive with a little over half the seek time, and even ATA drives with lower seek times but buffers that are smaller and/or not as smart. That's why there is no substitute for real world, application level tests of hard drive performance on the desktop. You simply can't benchmark buffer strategy and performance, and it is by far the biggest factor in hard drive performance on the desktop. How big of a difference? Well, check out this SR performance comparison: of note is the Cheetah 15K.4, which ships with two firmware modes, 'server' and 'desktop.' I have included the 73GB Raptor and 250GB Hitachi 7K250 as reference. After simply switching to 'desktop' mode, ie turning on the read ahead algorithms and other buffer tricks, the 15K.4's Office score rockets up 46%, High-End goes up 27%, Boot-up gains 35%, and the Gaming score increases by an astonshining 53% - with zero mechanical changes made to the drive, only firmware changes. Yes, buffer performance really is that big of a deal, and old SCSI drives simply don't have it.
 
Thanks for the info everyone. It looks like my dreams of adopting SCSI technology were never meant to be... Yet, anyway.

I figured saving the expense of the controller card would make things a little more accessible, but even though this one "supports" newer drives, it seems that it'll still come up a little short in the end.

Thanks for the lessons though :)
 
You must log in or register to reply here.
Back
Top