• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Any A64 Mobile Users Out There?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

TC

Senior Seti Addict
Joined
Jan 15, 2001
Location
Denver, CO
Aside from Guatam that is... I'm having major issues with my 3200+ DTR and I'm about to get rid of it for a regular desktop version. I wanted to bounce some ideas/comments around and see if anyone else has something to say. I first paired the chip with the GB K8N pro. It was a real flaky setup. It took me a half dozen power cycles, resets, insert key held, etc just to get it to post most of the time. Rarely it would fire up normally like most boards. I couldn't reboot the machine without having to babysit it - reset, power off and on, hold insert, etc. I finally got fed up and ordered one of the new Abit K8V Pros. I hoped it would behave a little better. Well got it all hooked up tonight and it will not even post - not once. I've cleared the cmos over and over, swapped out ram, video cards, removed and replaced the cpu, yadda yadda, etc etc. It flat out does nothing. So here's my thinking - these boards must be having a hard time because they don't recognize the cpu string - sound reasonable? The A64 platform can't be this bad most of the time I hope. I was about to sell all of it lock stock and barrell, but I think I'll get a regular desktop chip and see if this mobo will act normal that way. Any thoughts about this? Of course the way my luck has been going I'll get a new chip and it will turn out to be a bad board. Doesn't explain why the GB board acted so retarted though.
 
TC there is a known problem with DH7-CG cpus (was reading about it earlier on in the week) it is down to cpuid getting the wrong model instruction number possibly causing a failure to load specific cpu drivers. Not sure what your cpu is however?
 
It's the CG stepping alright. Not sure what cpuid would have to do with it though - I can't even get the mobo to post. I was thinking maybe the mobo can't identify the cpu so it doesn't have a clue what voltage to default to, so it doesn't post.
 
Looks like that's the problem - says bios may fail to recognize the cpu and take a fail safe default action. Looks like the GB board was doing that intermittently while the Abit board won't boot at all. Guess it's time to order a new chip.
 
Hmm...

I was about ready to get the same processor for the upcoming MSI motherboard. Are there any new revisions of the BIOS available that would maybe fix your issues?

I've spoken to Gautam and he's using the same exact board you are in his setup. Maybe he has a newer BIOS?
 
I don't have any of those problems. The only problem that I have is that if I'm using a core voltage of over 1.5, I have to go through the BIOS twice before it saves it, if I decide to change anything. Normally, I just have it set at 1.7v and have no problems ever starting up, unless I change stuff around. Then it takes two cycles, but its not enough of a deal for to bother changing anything in the setup...wastes at most 10 seconds maybe once or twice or a month if I ever decide to change anything. Older revisions of some boards may have trouble recognizing it, but I can't figure out why the Gigabyte would have any issues.
 
Abit doesn't list any bios updates - of course if they had one I would first have to boot the board with a different chip in order to flash it. The GB board was an absolute pain in the butt. Very often it would appear to be in a continuous state of reset when I turned it on. The ide light and cdrom drive light would flash over and over exactly like it would if the board was soft reset. It got to where I could tell if the board was going to finish posting or not by the length of the beep during post. Most of the time it would take about 60 seconds before it would start the post sequence, which would start with a long beep. It would then go through post and after displaying the bios options for both raid controllers it would either reset itself, or hard lock. Once in maybe a dozen power on attempts it will post immediately with a short beep and it would always load windows properly those times. Other times I would hit the power and it would just sit there for upwards of 2-3 minutes doing absolutely nothing. When it did that I would have to shut it down and clear the cmos. I never could get it to restart correctly even after a successful boot into windows. If I asked windows to restart the machine would reset and never post. Of course that's better than this Abit board. It never has posted a single time. I placed an order for a regular 3200 chip - hope that fixes the problem.
 
TC said:
Aside from Guatam that is... I'm having major issues with my 3200+ DTR and I'm about to get rid of it for a regular desktop version. I wanted to bounce some ideas/comments around and see if anyone else has something to say. I first paired the chip with the GB K8N pro. It was a real flaky setup. It took me a half dozen power cycles, resets, insert key held, etc just to get it to post most of the time. Rarely it would fire up normally like most boards. I couldn't reboot the machine without having to babysit it - reset, power off and on, hold insert, etc. I finally got fed up and ordered one of the new Abit K8V Pros. I hoped it would behave a little better. Well got it all hooked up tonight and it will not even post - not once. I've cleared the cmos over and over, swapped out ram, video cards, removed and replaced the cpu, yadda yadda, etc etc. It flat out does nothing. So here's my thinking - these boards must be having a hard time because they don't recognize the cpu string - sound reasonable? The A64 platform can't be this bad most of the time I hope. I was about to sell all of it lock stock and barrell, but I think I'll get a regular desktop chip and see if this mobo will act normal that way. Any thoughts about this? Of course the way my luck has been going I'll get a new chip and it will turn out to be a bad board. Doesn't explain why the GB board acted so retarted though.

Actually, it sounds like what's happening has nothing to do with the CPU. If the BIOS doesn't recognize the CPU string it should lock down to a 4x multiplier for a 800mhz total speed. But outside of that it should work fine. You can then either have your mobo manufacturer update their BIOS (mine did at my request - very nice of them) or I've heard you can use clockgen with the nForce3 150 mobos to change multiplier and FSB.

On the other hand, your mobo could flat out recognize the CPU like the Asus K8V Deluxe and work fine.

To me at least it sounds like you've just had a run of bad mobo luck. For example, I had never had a motherboard fail before and in the span of one month I had 3 mobos die and one arrive DOA. Since then, knock on wood, everything's fine. Might want to try the cheap new Chaintech board. Everything works pefectly on my setup.
 
I don't know - the amd advisory makes it sound like the bios should go into a "fail safe" mode and shutdown because it can't determine what voltage to supply. That's what this abit board does. It powers up for about 5 seconds and the post code goes through a range of numbers, then it shuts off. At first I thought maybe it was a cpu fan fail alarm, but I tried two different fans with rpm sensors to no avail.
 
I take it you're referring to the Abit board? Unfortunately, you're the first I've heard of that's tried it so I can't really comment. I suppose it could be the BIOS especially considering how raw the board is. I wouldn't really take that board as any sort of indication of the general state of board support for the DTR.

As to the Gigabyte, Gautam's got his working so the flakiness seems to be the board itself I would think rather than a BIOS recognition issue.

I've gotten the DTR to work fine on both a SiS 755 chipset board and a VIA K8T800 board while I've heard two seperate accounts of DTR working fine (but locking down to 4x multiplier) on nVidia nForce3 150 boards so I'm really doubtful it's a big issue. I just wouldn't put too much into that Abit board which is obviously beta given how early it came out and the lack of AGP/PCI lock.

If you're willing to give up the extra headroom and reduced temps the DTR gives you then the desktop would definitely be a good choice but I really think you've just had a bad run and should give the new nForce3 250 boards with AGP/PCI lock a shot but then again I'm unnaturally attached to my DTR ;)
 
Well I've already decided to swap it out with someone else. I wasn't too impressed with it really. The best it would do on air was 2400 with 1.7 volts. I was getting about the same with an fx51 I bought back last October, although it took 1.8 volts and definitely ran hotter. I was hoping this one would do 2500-2600 on air.
 
Well the mystery is over. I got a regular desktop 3000+ in today and have tried it in both boards I was having trouble with. They run flawlessly now. The abit board had never posted with the DTR in it, but it fired up instantly with the new chip. The GB board starts up instantly as well, and behaves normally through power cyclings.
 
I thought everything was running smooth, but actually this Abit board is doing one odd thing. This is a 3000+ newcastle which should be running at 2.0GHz. When the bios is set to Auto for some reason it runs 204X10 and reports 2040MHz. When I manually set the bios to 200X10 it reports the cpu speed as 1.8GHz. Multiplier control is locked as has been reported. There appears to be no agp/pci lock as has also been reported.
 
The Abit Pro board does clock the referance clock 4MHz higher in the boards which have been reviewed, and presumably in all boards. Abit might have assumed one wouldn't notice this while benchmarking motherboard comparisons.
 
It isn't running at the correct speed when manually set though. I locked it in at 200x10 and it's running 1800 :rolleyes:
 
TC, I think I mentioned this earlier, but I always do all of my overclocking within ClockGen. You may want to give it a try. Using a normal config script doesn't seem to let me control the fsb properly, so I use a batch file that calls several of them to keep it in check. Have it set to startup with Windows, and you're all set. I also managed to bypass the voltage saving problem, which you won't have anymore, by never restarting the computer, but just shutting it off and then turning it on manually. Guess its too late for this, but ClockGen may still help you.
 
Problemo - running redhat at moment. Has got me awfully curious though. It's reporting the same clock speed regardless of what the bios is set to.
 
Last edited:
Back