Tom's Guide | Tom's Hardware | Tom's Games
![]() |
![]() |
![]() |
Hi All,
I recently purchased a CompUSA Brand ATA/100 controller card to speed the performance of a 500 Mhz AMD K6 II system (FSB 100MHz).
The controller card bears a "CMD PCI-649" controller chip that reads:
"Silicon Image
SiI0649CL160
A003
0138
Taiwan"
FCC# MP649From my reading, CMD Technology is purported to be a "leader in controller technology", and the CMD PCI-649 a "good controller chip".
After installing the controller card I noticed that the system was significantly slower. Initially, the computer performs the "POST" test at normal speed. Next, the OEM logo appears as usual, and then the CMD PCI-649 information screen appears and begins to detect the ATA/IDE attached devices.
The appearance of the Windows 98 logo screen immediately follows. However, the logo stays on the monitor a long while (60-90 secs). When the logo screen finally disappears it is replaced by the computer's wallpaper, and the "hour-glass icon" appears for an equally long time. Finally, the system begins to load it's desktop and application icons ( these operations also take an unusually long time to complete).
I've installed the latest drivers provided by CMD Technology. When I check the CMD PCI-649 Primary IDE Controller icon in Device Manager, it says the PCI bus speed is 33 MHz. The BIOS shows the following values for the onboard IDE controllers:
"PCI BUS Mastering .......Enabled"
"Ultra ATA 33/66 .......Enabled"
"Primary IDE .......Disabled"
"Secondary IDE .......Enabled"Question #1: Can anyone offer suggestions on how can I can overcome this problem, and get this system to experience the ATA/100 maximum speed of the card? And if not possible, how I can at least regain the speed the system has lost? (Preferably while retaining the card).
Question #2: Also, is it possible that the PCI Bus Mastering and Ultra ATA 33/66 functions on the motherboard might conflict with those same functions on the ATA/100 controller card?
Please advise.Thanks All

I hate to break this to you, but the new ATA-100 controller won't do anything to improve the performance of your machine. By far the biggest factor in disk access, is the hard drive itself. A faster controller can only allow a really fast drive to achieve its potential. So, unless you already upgraded to a blistering ATA-100 drive that runs at 7200 RPM or better, the controller won't have an impact.
Your post suggests your drive runs at ATA 33, not 66 or 100. It's the maximum speed of the drive that limits your overall disk access, so you shouldn't expect better than 33.
The boot-up delays are common with IDE controllers. First, since they are add-ons, they typically have self-detecting BIOS'es, which take longer to detect than regular IDE busses. As far as slow Win98 pickup is concerned, it sounds like your drive doesn't have the correct drivers on it. Assuming your card is similar to Promise's, I do know that there are detection issues if you don't install the right drivers.
There are other Win98 issues with controllers, that you can look up in Microsoft's Knowledge Base. I suggest you spend some time searching there.

Hi JeffJ,
Thanks for your response.
Firstly, the drive is a Maxtor Ultra ATA/100. It is compatable with all former data transfer rates.
The motherboard has "Ultra DMA 33/66" Enabled in the BIOS, which leads me to believe that the motherboard can suport a maximum transfer rate of 66MHz.
Thanks for clarifying the fact that add-on controllers will takes longer at boot-up. But, is it normal for boot-up to increase from 1.5 minutes(without the card)to 3.5 minutes(with the add-on card)?
What I gather from your response to my post, I should expect the boot-up to be slower - but applications and processes should be faster. Is this true? Also, do you know of any "free" and "spyware-free" Benchmarks or utilities that can determine the speed of a system's memory and PCI Buses?
Lastly, do drivers make a difference? Do you think that if I remove the drivers that come with the add-on controller card, the system might operate faster?
Thanks for all your insight.
Any further help is appreciated.

I'm greatly relieved you have an ATA-100 drive, as I never got that from the original post. You do, in fact have a really fast drive, so things should go faster, as long as you keep your drive defragmented.
No matter what, the presence of the card will add some time to boot up, though I think it's more like under a minute more for my Promise controller, so something doesn't seem right with the extra 2 mins yours needs. That's a long time to auto-detect. I really think spending some time searching Microsoft's Knowledge Base might be worthwhile, as well as your card manufacturer's site.
FWIW, the longest delay for a Win98 boot, is the registry backup time. Check out these articles:
"http://support.microsoft.com/default.aspx?scid=kb;en-us;Q183887"
"http://support.microsoft.com/default.aspx?scid=kb;en-us;Q183603"
I used to love Norton Utilities' benchmark, but they discontinued it since v3.0. I also use the popular SiSoft Sandra, which is free for drive benchmarking. Look at some articles over at www.tomshardware.com, and you'll see the latest gizmos.
Regarding drivers, I didn't mean to say that having the drivers slows things down, but rather that not having them stalls the controller's detection process, and may delay Win98's loading too. Controller cards are SCSI devices, so they need the right drivers. I had a drive connected to a controller on a previous machine, and it took enormously long to load, since it didn't have the right driver installed. I put the right one on each drive, and bootup was much faster. There are some Plug-n-Play issues when Windows can't get to the controller right.
And don't worry about your motherboard's controller chip only being able to reach 66; you're bypassing that with your controller card anyway. I misinterpreted your first post to mean an actual speed limit of 66 was detected. Aparently not, since that was refering to your unused on-board chip.

Jeff, Hi look I'm not familar with your controller or chip, but I purchased two Promise Ultra TX2 100 cards to put into my kids machines.
My son has in his Epox EP-MVP3G5 mobo AMD K6-2 350 3.5 @112FSB (soon to be a K6-III+ 450) and Seagate Barracuda IV ATA-100 40gb drive and everything works fine.
It came with a disc to run in the bios etc.
What I found was:
Leave the BIOS setup to AUTO detect the drives.
Turn off the on board IDE disk controller as it conflicted with the sound card using the same IRQ (they shared but what the hell it free'd up a IRQ and resources.)
I stuck his burner/rom drive on the other channel on the card. His machine boots up fine, rock solid, no blue screens In fact his drive benchmarks better than my IBM 75GXP Deskstar on my Intel EEAU2L with PIII 1 ghz.
The other promise is in my daughters machine.
Fordlain S7 Mobo (Who) yeah I know with a Intel P100 running 83FSB x 1.5 = 125 and my two old drives. Quantum CR Fireball 4.3 66 and Fujitsu 4.3 66. The reason is the mobo is 33 so at least the run the best they can.
Her rom drive uses a channel on the board since it's only a hex speed.
I'm sorry I cant offer any more advise than that.
Have a look for conflicts in the device manager. Enable DMA in the properties for your drive.
Stick the card in another slot ?
Don't stick it in PCI 1 if you are running a AGP card. Check the BIOS to see if you have pnp set or is the config auto, have a play.I think Jeff is right and you don't have the right drivers installed. Look at the Maxtor site for there drive .inf file if any and maybe a firmware??
All the best Mike.

Ej, Hi look I'm not familar with your controller or chip, but I purchased two Promise Ultra TX2 100 cards to put into my kids machines.
My son has in his Epox EP-MVP3G5 mobo AMD K6-2 350 3.5 @112FSB (soon to be a K6-III+ 450) and Seagate Barracuda IV ATA-100 40gb drive and everything works fine.
It came with a disc to run in the bios etc.
What I found was:
Leave the BIOS setup to AUTO detect the drives.
Turn off the on board IDE disk controller as it conflicted with the sound card using the same IRQ (they shared but what the hell it free'd up a IRQ and resources.)
I stuck his burner/rom drive on the other channel on the card. His machine boots up fine, rock solid, no blue screens In fact his drive benchmarks better than my IBM 75GXP Deskstar on my Intel EEAU2L with PIII 1 ghz.
The other promise is in my daughters machine.
Fordlain S7 Mobo (Who) yeah I know with a Intel P100 running 83FSB x 1.5 = 125 and my two old drives. Quantum CR Fireball 4.3 66 and Fujitsu 4.3 66. The reason is the mobo is 33 so at least the run the best they can.
Her rom drive uses a channel on the board since it's only a hex speed.
I'm sorry I cant offer any more advise than that.
Have a look for conflicts in the device manager. Enable DMA in the properties for your drive.
Stick the card in another slot ?
Don't stick it in PCI 1 if you are running a AGP card. Check the BIOS to see if you have pnp set or is the config auto, have a play.I think Jeff is right and you don't have the right drivers installed. Look at the Maxtor site for there drive .inf file if any and maybe a firmware??
All the best Mike.

I did forget all about the BIOS, thanks Mike. I do remember that both my old Ultra66 and my newer Ultra100 have custom BIOS'es, so updating either the controller card's or the MoBo's BIOS might help (though that's delicate). Looking in Device Manager is a really good idea.

JeffJ and Mike:
Thanks alot. You both clarified some concerns I had.
JeffJ: I'll check the resources you suggested over the next few days.
Mike: I've checked Device Mgr and no conflicts are present. Also, DMA is Enabled on the drive, and PnP is functioning. There is no AGP slot since the system has video-on-board. Out of curiosity, does the AGP/PCI1 slot rule still apply when you have built-in video?
One more question guys:
1- When I installed the Maxtor ATA/100 drive, I used the MaxBlast Plus software - which installs EZ-BIOS on the drive. Could one possible explanation for the sluggish system start-up be due to the THREE bioses(system bios, controller card bios, and EZ-BIOS) vying for control?Any additional insights are appreciated.
In the meantime, I've disconnected that ATA/100 controller card from the system. I will look into obtaining a Promise card. Although I think the system may be fine at 66MHz(maximum burst speed) for now.
Once again guys,
Thanks For All Your Help.

I only used EZ-BIOS once, and I didn't leave it activated for long, but I somehow suspect that's not a problem. It's only a BIOS-extender, intended for use with older systems that do not support really large drives. If your machine does, you don't need it, but I don't think it's significantly impacting boot-up (it's one of the first things loaded, way before the "Starting Windows 98..." prompt). If removing your card solves things, then it definitely isn't a problem.
The controller card's BIOS is completely separate from your motherboard's, so I don't think they're in conflict. However, if Windows can't communicate with the card, if its BIOS or SCSI driver aren't setup right, then drive data would be harder for Windows to interpret, which could cause delays. Mikes and my cards' obviously are setup as desired, but that doesn't necessarily mean yours isn't (isn't problem-solving fun!).
Using the regular IDE busses at 66 will be fine, as even current drives sustain only about 30-40 anyway. It's actually a newer drive's huge RAM cache and better algorithms, that make it so blazing fast, not so much the burst rate (the 100 part of the spec). If your on-board controller was a 33 I'd say definitely stick with your card, but the diff between 66 and 100 burst is small and rarely hit. Ideally, the card only adds around 15-30 secs to startup (what I measure on my sys), and that's way before Win98's logo shows. However, I wouldn't bother buying a Promise; it likely will behave the same, as I suspect an unresolved Windows issue between [your] outboard controller and Win98.
Regarding slot1, I believe most onboard AGPs are hard-wired to where an external AGP slot would normally be anyway, so the issue remains unchanged for you. Avoid slot1 if you can, and play it safe.
Keep digging; I really think you can resolve the extra delay, and have the best of all worlds :)

![]() |
![]() |
![]() |

This post is quite old and has been locked from receiving new replies. Please create a new posting instead.
| Ads by Google |