Sign in with
Sign up | Sign in
Your question

My WD 7.2k hdd is faster than my Raptor?!?!

Last response: in Storage
Share
October 10, 2006 2:31:16 PM

Heya,
OK so I was noticing slow map loadtimes in WoW on my new Raptor vs. my old WD 7.2k Caviar. So I downloaded HDTach 3.0 to compare them and what the devil???...The burst-speed of my 7.2k HDD is FASTER than the Raptor. They are both SATA 150 and my old Caviar has 8mb of cache vs. the Raptor's 16mb. Why is the burst-speed of the Caviar almost twice as fast as my Raptor? I haven't switched the jumper on the Caviar to slave or cable select and I think my computer thinks its the primary (because its listed above my Raptor which has the OS on it)...Could that be a cause? This doesn't make any sense to me...

More about : hdd faster raptor

October 10, 2006 2:56:51 PM

Something sounds MIGHTY fishy. Even without running a benchmark, I could tell a difference moving from a 7200 WD drive to a 10k Raptor (74GB). Perhaps it has something to do with paging... I'm assuming both drives are installed at the same time. Assuming your page file is located on your Raptor, your 7200 drive would have a huge advantage as paging wouldn't degrade its performance. When you installed WoW on the Raptor, the game must load from the same drive that is handling paging duties.

Of course, this is just a theory based on what information you gave... but hey... no one else has replied so I figured what the hell.

To repeat... it's just not possible that Raptor is slower.
October 10, 2006 3:01:38 PM

You got something wrong there...
I have a 150gig Rapter, two 74 gig ones, and two 7200 WD SE16's, and performance is in order I listed them...
Related resources
October 10, 2006 3:01:49 PM

I do have a small page file set up, but mind you its small as I already have 2gigs of RAM. In addition to that, when I DID have WoW installed on my old HDD, it loaded maps faster than the Raptor as well.
I will report that this is not the case in all apps...The Raptor definately installs, reads/writes faster, etc. But the burst is slower and my WoW map loads are slower. So yes there definately is something arye. HDD Sentinel reports it's in 100% perfect condition. I'll check the pagefile when I get back home...Maybe give it NOTHING and compare them again. Maybe I'll switch the jumper to Slave on the Caviar too.

*Edit* If it helps at all, I'm using the on-board SATA controller on my mobo. Would using a 3rd party controller card (a budget one) help?
October 10, 2006 3:19:40 PM

Quote:
Heya,
OK so I was noticing slow map loadtimes in WoW on my new Raptor vs. my old WD 7.2k Caviar. So I downloaded HDTach 3.0 to compare them and what the devil???...The burst-speed of my 7.2k HDD is FASTER than the Raptor. They are both SATA 150 and my old Caviar has 8mb of cache vs. the Raptor's 16mb. Why is the burst-speed of the Caviar almost twice as fast as my Raptor? I haven't switched the jumper on the Caviar to slave or cable select and I think my computer thinks its the primary (because its listed above my Raptor which has the OS on it)...Could that be a cause? This doesn't make any sense to me...


I guess you checked DMA mode? changing things can sometimes default a drive to PIO. Or the wrong driver may be mapped to a particular drive. Are the drives being run AHCI or IDE? An AHCI drive might become mapped to an IDE driver. There are many variables. The easiest solution is usually to UNinstall the chipset drivers (ICHR6 etc.) then WITHOUT shutting down "Search for new Hardware" and let the installation run its course. If you shut down or reboot meanwhile you can lose all drives and not be able to get back in to Windows without a real hassle.
October 10, 2006 3:34:52 PM

Aye well here's the trick. I WAS running them off of the ICH8 intel ports which I guess are inferior to the JMicron ports. So then, I switched them, and in BIOS turned it to AHCI mode. But my Device manager is kind of a mess because I believe that the drivers are still there from when they were in the Intel ports using Legacy/IDE mode. Perhaps I'll post a screen shot when I get home.
So you say if I uninstall all the hard-disk controllers, then reinstall them with Find New Hardware all without rebooting that it should
A) Not shut down my comp, and will be able to reboot sucessfully
B) Purge out all the drivers I didn't need, and using only the ones that I need.

I believe the JMicron drivers are on the CD that came with my mobo.
October 10, 2006 4:05:26 PM

Quote:
Aye well here's the trick. I WAS running them off of the ICH8 intel ports which I guess are inferior to the JMicron ports. So then, I switched them, and in BIOS turned it to AHCI mode. But my Device manager is kind of a mess because I believe that the drivers are still there from when they were in the Intel ports using Legacy/IDE mode. Perhaps I'll post a screen shot when I get home.
So you say if I uninstall all the hard-disk controllers, then reinstall them with Find New Hardware all without rebooting that it should
A) Not shut down my comp, and will be able to reboot sucessfully
B) Purge out all the drivers I didn't need, and using only the ones that I need.

I believe the JMicron drivers are on the CD that came with my mobo.


Without the Jmicron drivers the hard drives on their ports will still work as standard IDE with reasonable speed, but there's no Raid capability on those ports. There is a small element of risk in the Controller reinstall I suggested but overwhelmingly it is safe. Before doing that first of all try clicking on each controller in Device Manager and selecting "update Driver," in some cases that will actually correct a driver error, a particularly effective method in Vista.
October 10, 2006 4:30:04 PM

You should always install ALL the motherboard drivers, unless you are Sure you are not going to use the device, in which case you should disable in the BIOS.
Mike.
October 10, 2006 8:21:11 PM

I think I found some answers.
My board has two different Sata ports...Gigabyte (JMicron) and Intel. When I installed them, I used the Intel ports which are defaulted in BIOS as IDE Emulation (which was why I didn't have to F6 for the floppy). So in fact, they hadn't been running at full SATA150 speed. Unfortunately I've found a whole slew of theories and driver updates and crap to try and fix the issue. So I think I'm going to have to re-install Windows, and use an F6/floppy for the JMicron ports. Only problems are:
A) Some people had weird issues with that plus an IDE CD-ROM and it was causing major problems

and B) Even if this all were true, BOTH of my SATA-150's are running in this Emulated mode, which still wouldn't explain why the slower hdd is faster burst than the Raptor.

Ugh,....I think Ill be up late tonight again.
October 10, 2006 10:01:34 PM

Misunderstaning One:

Burst speed is not a measure of hard drive performance.

What it measures is the actual bandwith of the interface and bus through which the hard drive is connected.

It is only relevant when the burst speed is nearly equal to the peek transfer rate of the hard drive.

In such cases it means the hard drive could be running faster if it had a better interface or was connected to a faster bus.

For example I have a hard drive that starts at about 65 MBps and ends at about 35 MBps with an average of 54 MBps

But when its in my USB enclosure it gets 32 MBps. I look and see the interface speed is 34 MBps.

The interface speed is slower than the hard drive so it matters.

When I connect the same drive in the same enclosure via the SATA cable I again get 54 MBps. With a burst speed of 90 MBps.

Since the drive never reached the actualy upper limit of its interface, the burst speed was completely irrelevant to its peformance

The HD Tach results that ARE measures of a hard drives performance are t the Random Access Time, Average Read Rate and the transfer graph showing how the read rate changes across the drives surface.


Miisundertanding Two: (which may not apply to you at all)
Average read rate is only one of many factors that affect hard drive performance and in most real world conditions it is often the least important.

Check out storage review database and compare the 74 GB Raptor with the 750 GB Seagate (without TCQ NCQ). The Seagate has higher read rates but loses every real world performance test except noise level and power usage.

http://www.storagereview.com/php/benchmark/suite_v4.php?typeID=10&testbedID=4&osID=6&raidconfigID=1&numDrives=1&devID_0=323&devID_1=306&devCnt=2
October 10, 2006 11:27:45 PM

Quote:
You should always install ALL the motherboard drivers, unless you are Sure you are not going to use the device, in which case you should disable in the BIOS.
Mike.


You miss the point, by default the Jmicron port will run in windows, the drive isn't "hidden" and inaccesible, even if there isn't native OS support for Jmicron. The drive is shown as Master on port 0, the first one.. and it is recognized during install without a floppy. And it does transfers at 45 to 60 Megabytes/sec from and to the 500 gig RE2 connected to the rear Jmicron port on the P5W-DH board with Windows IDE drivers, so that's not shabby. The other point is I'm running 64-bit Vista 5744, and there aren't working JMicron drivers for Vista yet (the XP X64 drivers are buggy.) The Jmicron port is set to IDE ("Basic"), the others are AHCI.
October 11, 2006 6:03:05 AM

When all said and done, I ended up using the JMicron ports because I got better Hd Tach scores using them. I finally figured out that the Caviar is SATA II (although I swore it wasn't...oh well) I think I've done everything I can but yet my mind is still boggled. The Raptor surpases the Caviar in all scores (except burst, which Codesmith pointed out is almost negligible in performance), yet I still have horrible load times, and such. I also have no idea what was changed those two times or so that everything worked like butter either. One time it happened when I was messing with an in-game addon/mod, and the other after I had done some initial HDD toying.
Eithetr way I've tried reinstalling the game, all the drivers, configuring HDDs to best performance, and everything else under the sun. WHo knows it could be a game bug of some sort.
October 11, 2006 6:16:18 PM

Master/Slave relationships only exist for drives sharing a PATA cable. The master is faster because the slave needs to ask the master's permission to use the shared bus.

Many people mistakenly confuse master/slave with hard drive boot priority, leading some to thing boot order affects performance.

(Again I am not assuming this applies to you).

--

Many factors can affect performance. If you are getting good numbers in HDTach its probably not your hard drives fault.

Startup could be slowed due to the motherboard. Non-chipset storage controllers often take a long time to detect hard drives, they also pause to allow you to press F whatever and enter thier setup.

If you are not connecting any devices to such a controller its best to disable it in BIOS.

Another factor could be hard drive fragmentations.

Another factor could be spyware, or an excessive number of program running at startup or in the backgroud whether you are using them or not.

Also I noticed that not having the correct drivers for a particular controller can influence performance on other controllers, even if no devices are connected to that controller with the wrong drivers.

Basically the OS keeps r-trying to internact with the hardware using the wrong drivers rather than just giving up and ignoring it after the first attempt.

You might want to look at your event viewer to see if you system is throwing any errors.

Hope some of this is usefull
!