2CPU.com » News » December 2002 » Seagate ST380023AS Serial ATA Drive Reviewed

Hexus.net has a review of Seagate's ST380023AS and they are impressed with the drive, even after noting the drive's first-generation serial ATA design and the use of a serial ATA controller on the PCI bus.

The benefits of ultra fast data writing would make this drive ideal for write hungry tasks like video rendering or data backup. The Seagate drive itself is very well made and seems to be very robust. Its quiet operation makes it ideal for inclusion in a system where quietness is of benefit.

12/10/2002 11:13 AM: 2CPU.com reviews the Seagate 15K.3 by dadmin
I finally found the time to write up some thoughts on Seagate's flagship drive, the 15K.3. Looks like a hard drive; feels like a hard drive; goes like stink....on a whim I threw it in the machine, ins...

06/19/2002 03:24 PM: VIAHardware + Seagate Barracuda ATA IV = Decent IDE RAID Article by Hooz
IDE RAID. What is there to say? It seems that anymore everyone has it in one flavor or another. I remember when the first Promise RAID controller hit the market (expensive too), now they're giving IDE...

02/01/2001 10:57 PM: IBM joins Seagate at the 15K rpm Plateau! by J0rdan
This just in from the good chaps at The Register. Apparently IBM will soon be releasing a 15,000 rpm UltraStar hard drive. Of course IBM says it also will out perform its competition in certain benchm...

Murdock Captain Posts: 1398 Joined: 2001-07-02

#20518 Posted on: 01/01/2003 03:08 AM That bit about the "SATA interface located on the PCI bus"... Are they trying to say that it's bridged from the native PCI-PATA controller? Are there any SATA controllers that aren't on the PCI bus??? Other than that confusion, a decent read. Looks like a nice drive. The pathetic state of our government will never change unless we stop electing politicians and start electing public servants. Remember: There was once a time when the term "politician" had a very negative connotation.

Pomp Huffs Thermal Adhesive Posts: 397 Joined: 2000-02-17

#20519 Posted on: 01/01/2003 05:27 AM They also say "It is common for performance to drop the further into the drive the test goes. This is due to the sectors at the end of the disk being physically further from the drives starting point." What the hell does that mean. Could also be that the sectors at the end of the disk spin slower. Or maybe that's what they were trying to say? Dunno. #/> Pomp

questionlp the Cowardly Tech Posts: 323 Joined: 2002-02-14

#20520 Posted on: 01/01/2003 05:41 AM Originally posted by Murdock That bit about the "SATA interface located on the PCI bus"... Are they trying to say that it's bridged from the native PCI-PATA controller? Are there any SATA controllers that aren't on the PCI bus??? Other than that confusion, a decent read. Looks like a nice drive. I think it means that the Serial ATA controller is attached to the PCI bus rather than integrated into the southbridge. 32-bit/33MHz PCI has a maximum throughput of ~133MB/s whereas Serial ATA currently has a maximum throughput of 150MB/s. Does that really matter right now? Probably not since you would need two of the fastest hard drives sending data out at it's peak to even breach 80MB/s. But it also means that it has to share the bandwidth with all of the other devices on that bus. Instead, if it was integrated into, say the nForce 2 or the ICH4 southbridge, it would have a separate connection to the bus connecting the southbridge back to the northbridge (800MB/s and 266 or 533MB/s respectively). [http://closedsrc.org/] My Rig: 2x 2.4GHz Xeon, 1GB PC2700, Supermicro X5DAL-TG2, ATI AIW 9000, SB Audigy 2 Plat/EX, 2x 80GB 7200.7's, Plextor 40x and 8/20, Pioneer 106D, Antec 1000AMG, Enermax 460W EPS-12V

Murdock Captain Posts: 1398 Joined: 2001-07-02

#20521 Posted on: 01/01/2003 07:13 AM Ahhh.... I was under the impression that even the devices that were integrated into the southbridege (physically) were still (electrically) connected VIA the PCI bus... demonstrated by standard IDE controllers showing up on the PCI device list during POST. But I could be wrong. The pathetic state of our government will never change unless we stop electing politicians and start electing public servants. Remember: There was once a time when the term "politician" had a very negative connotation.

Buub Old SMP codger Posts: 1276 Joined: 2002-11-19

#20522 Posted on: 01/01/2003 07:46 AM They also say "It is common for performance to drop the further into the drive the test goes. This is due to the sectors at the end of the disk being physically further from the drives starting point." What the hell does that mean. Could also be that the sectors at the end of the disk spin slower. Or maybe that's what they were trying to say? Dunno. Almost all drive made today use what's called "zone recording". Basically, a track on the outside of a disk is longer, physically, than a track on the inside. So the drive is broken up into zones, having different numbers of sectors. The inside-most zone might have 30 sectors per track, where the outside-most zone might have 60 sectors per track. Since the drive always spins at the same speed, this means that there is twice as much data throughput on the outside tracks. CD-ROM drives do basically the opposite. They speed up and slow down the physical rotation to keep a constant data rate coming in, depending on where you are on the physical disc.

HEMI Old SMP codger Posts: 2467 Joined: 2001-12-18

#20523 Posted on: 01/01/2003 08:53 AM Originally posted by questionlp I think it means that the Serial ATA controller is attached to the PCI bus rather than integrated into the southbridge. 32-bit/33MHz PCI has a maximum throughput of ~133MB/s whereas Serial ATA currently has a maximum throughput of 150MB/s. Does that really matter right now? Probably not since you would need two of the fastest hard drives sending data out at it's peak to even breach 80MB/s. But it also means that it has to share the bandwidth with all of the other devices on that bus. Instead, if it was integrated into, say the nForce 2 or the ICH4 southbridge, it would have a separate connection to the bus connecting the southbridge back to the northbridge (800MB/s and 266 or 533MB/s respectively). Yeah, that's what I was thinking when I posted the link after reading the article. The author was factoring in the 133m/s PCI throughput limit in to the drive's overall speed versus, say, having the serial ATA controller hooked directly in to (in theory) the Hypertransport bus on AMD boards or whatever system bus the Intel boards are using these days. Theoretically, the shared PCI bus could be a bottleneck down the road. Unix is user-friendly; it's just picky about its friends.

Mr Bill two by two, hands of blue Posts: 2946 Joined: 2002-02-16

#20524 Posted on: 01/01/2003 01:13 PM 5% slower reads to get 30% faster writes and its hot swappable. This is hot stuff! My SMP rig [URL="http://personal.palouse.net/billshan/ghost.htm#A_Merlin"]Merlin[/URL]

Nuclear Aspiring Duallie Posts: 92 Joined: 2001-10-01

#20525 Posted on: 01/01/2003 08:31 PM buub Old cdrom where clv (constant linear velocity) which is how you described it, the cdrom would spin at faster speed on the inner track and slower spped on the outher so to have the same transfer rate. But modern cdroms (higher thant 16x i think) use cav (constant angular velocity) which is the same as hard drive, the cd always spins at the same speed. But the difference in the speed lies in the fact that hd record from outher to inner and cdrom are the reverse (inner to outer). So a hard drive start with high transfer then slow down, cdroms are the opposite, starts with a slow transfer then speed up (for cav, for clv, it's always the same). (the exception to this would be dreamcast cd, which were outher to inner, so to burn them using a normal writer, you needed a program that would reverse the image) Thunder K7 with 2 athlon 1.6g , 512megs and Radeon 9700 pro

questionlp the Cowardly Tech Posts: 323 Joined: 2002-02-14

#20526 Posted on: 01/02/2003 01:23 AM Having everything from Gigabit Ethernet, USB 2.0 (for those motherboards that include additional USB 2.0 controllers outside of the integrated one in the southbridge), FireWire, ATA/133 RAID, 4+ channel audio, and who knows what else is being integrated into today's motherboards are most likely sharing the same 133MB/s PCI bus. The more you use all of those components, plus trying to do heavy disk I/O, the key limitation is definitely going to be the shared bus. But... doesn't the i84x/i850 chipsets limit the maximum PCI throughput to around 90MB/s? I guess for most people (business users, home and home office users) that the architecture is more than what they need. But for use performance users, it's becoming a problem. Now... for the dual processor users (dual Athlon and dual Xeon anyways) there are other issues to worry about [http://closedsrc.org/] My Rig: 2x 2.4GHz Xeon, 1GB PC2700, Supermicro X5DAL-TG2, ATI AIW 9000, SB Audigy 2 Plat/EX, 2x 80GB 7200.7's, Plextor 40x and 8/20, Pioneer 106D, Antec 1000AMG, Enermax 460W EPS-12V

Nuclear Aspiring Duallie Posts: 92 Joined: 2001-10-01

#20527 Posted on: 01/02/2003 02:30 AM questionlp Most of the newest generation of dual Xeon boards (using the i7500 generation) have 2 pci-x channel, so you have at least 2 time 133 megs for 32 bit cards Most of the board using integrated gigabit and scsi are using one channel for the onboard device and the other for the cards. (like the iwill we saw before on 2cpu So for power user running dual xeon, things are getting better, for dual athlon, well using the mpx, you have a 32 bit and a 64 bit channel, so it's still better than normal pc's Thunder K7 with 2 athlon 1.6g , 512megs and Radeon 9700 pro

Zyggman SMP Lurker Posts: 336 Joined: 2002-01-22

#20528 Posted on: 01/02/2003 02:56 AM I'm just wondering why it needs a different power connector? What is the point of that? Full Specs

HEMI SMP Lurker Posts: 2467 Joined: 2001-12-18

#20529 Posted on: 01/02/2003 03:59 AM Originally posted by Zyggman I'm just wondering why it needs a different power connector? What is the point of that? Maybe serial ATA communicates with lower voltage. There could be a few reasons. If the new connectors are easier to unplug I'm all for it. I hate trying to unplug a four-pin drive Molex in a tight spot. Unix is user-friendly; it's just picky about its friends.

questionlp the Cowardly Tech Posts: 323 Joined: 2002-02-14

#20530 Posted on: 01/02/2003 04:03 AM I think the reason for the different power connector is to standardize the connector for desktop and mobile hard drives, plus to allow for hot-swappable hard drives. The other reason might also be to have a thinner or more flexible power cable that is easier to route in a tiny case. Dunno... just a couple of things that popped into my mind [http://closedsrc.org/] My Rig: 2x 2.4GHz Xeon, 1GB PC2700, Supermicro X5DAL-TG2, ATI AIW 9000, SB Audigy 2 Plat/EX, 2x 80GB 7200.7's, Plextor 40x and 8/20, Pioneer 106D, Antec 1000AMG, Enermax 460W EPS-12V

jives BP6 User Posts: 2419 Joined: 2001-05-18

#20531 Posted on: 01/02/2003 08:56 PM Wonder why they didn't test it against an 80G ATA Seagate drive? Seagate might just be the first out the door with SATA drives but I've never looked to them as the top of the line in performance WD and Maxtor drives (the older IBM as well) always seem to be at the top. The cabling alone is worth the switch for me 1 meter and only about 0.5

zaphod78 SMP Newbie Posts: 36 Joined: 2002-10-15

#20532 Posted on: 01/03/2003 10:18 AM The new power connector has longer ground pins so that when you plug the drive in (particularly during a hot-swap), the drive is grounded first, then current is passed. Zaphod78 -Iron Programmer

Tanj FORK ME BABY! Posts: 395 Joined: 2000-02-05

#20533 Posted on: 01/04/2003 01:30 AM zaphod78 and questionlp are correct about the new power connector. The new connector will be standardized between desktop and mobile applications. For mobile applications there is a 3.3v rail that will be used for power saving features. -Tanj! My F@H Stats My System Stats

2CPU.com » News » December 2002 » Seagate ST380023AS Serial ATA Drive Reviewed