I've justed tested 4 Athlon XP CPUs (2 Palominos and 2 T-Breds) with this board, and here are the disappointing results -

OK:

T-Bred 1700+ AXDA1700DLT3C green. Works OK. (However, BIOS POST says it is an "unknown CPU type")

No POST:

Palomino 1700+ AX1700DMT3C brown.
T-Bred 2200+ AXDA2200DKV3C brown.

Significant problems:

Palomino 1900+ AX1900DMT3C green. Windows 98 goes "mad" at the desktop: double-clicking a desktop application yields mad error message of "not enough memory to run this app". Retested: reached desktop OK, but Prime95 instantly fails.

I am testing an Abit KT7A non-Raid board, PCB revision number 1.1. The BIOS POST datestamp is 02/14/2001 ....YH

These CPUs POST OK -

1700 t-bred
1900 palomino (then inside the BIOS set up area, I increase the FSB to 133. This works.)

These CPUs do no POST -

2000 palomino (I thought I had this working earlier, with 100speed RAM, but it won't work now. Edit: sometimes it does, sometimes it doesn't.)
2000 t-bred

Please flash the BIOS to the latest one and see if the t-bred 2000+ works 😀

My Palomino 2000+ was properly recognized when I flashed to this BIOS on my old now dead board

http://file.abit.com.tw/pub/download//fae//kt7a9.exe

Create the bootdisk with this

http://www.techsp.net/Files/drdflash.exe

Seems like the Abit FTP site died again

FTP now works 😀

Here's the direct link to the latest bios

ftp://ftp.abit.com.tw/pub/download/bios/kt7/kt7a9.exe

Thanks a lot for the BIOS! 😀 The flash operation was succesful. The BIOS datestamp now reads: 07/11/2002 ... A9. With the Palomino 1900 CPU still installed, all seemed OK. I then removed it, and tried both a Palomino 2000 and a T-Bred 2000. No POST. 🙁 I tried clearing the CMOS each time. No joy. As soon as the Palomino 1900 went back in, all was OK. That's a pity! 😕

Edit: The Palomino 2000 which I am using (and only occasionally POSTs on this board) can be seen here, on cpu-world -

http://www.cpu-world.com/CPUs/K7/AMD-Athlon%2 … X2000DMT3C.html

The mystery deepens. I just put in a barton 2400 CPU, and switched the mains power on. The PSU power came on, without me having to press the power button on the mobo. I then see the BIOS POST screen! The CPU is reported to be "Unknown type". The speed is reported as 1667mhz. I get to the windows 98 desktop OK. I run Sandra, and see that the FSB is 266 and the memory is 133. The multiplier is only 12.5x. Ideally, it should be 13.5x for this barton 2400 CPU. There is no 13.5x option inside the BIOS. I chose the best matching option, which was "12x higher".

I wonder if the Palomino and T-Bred 2000 CPUs are not working because of an "issue" with the way that the mobo deals with these CPUs power/voltage requirements?

retro games 100 wrote:

Thanks a lot for the BIOS! 😀 The flash operation was succesful. The BIOS datestamp now reads: 07/11/2002 ... A9. With the Palomino 1900 CPU still installed, all seemed OK. I then removed it, and tried both a Palomino 2000 and a T-Bred 2000. No POST. 🙁 I tried clearing the CMOS each time. No joy. As soon as the Palomino 1900 went back in, all was OK. That's a pity! 😕

Edit: The Palomino 2000 which I am using (and only occasionally POSTs on this board) can be seen here, on cpu-world -

http://www.cpu-world.com/CPUs/K7/AMD-Athlon%2 … X2000DMT3C.html

Try and get into the BIOS and "use failsafe defaults" save the settings and reboot...this solved the issue you're having on my old KT7A 1.1 board

I'm almost certain this was the case. However, I will retry both 2000 CPUs, with "failsafe" BIOS settings. However, while the barton 2400 was installed, I tried this -

BIOS settings = "Optimized" settings, not "failsafe".
OC'g settings = 140 FSB

I get to the desktop and run Sandra:

(real) Speed of CPU = 1750mhz
FSB = 280, RAM = 140
Sandra CPU scores = fractionally better than a 2200+ CPU, which is to be expected.
CPU temp = 24C. That's not bad!
Prime95 (max FPU stress) = doesn't fail and CPU temp is still 24C!

My early conclusion is: forget the stove hot Palomino 2000. Stick in a cooler Barton (with 512k cache) instead, and overclock it.

retro games 100 wrote:

I'm almost certain this was the case. However, I will retry both 2000 CPUs, with "failsafe" BIOS settings. However, while the […]
Show full quote

I'm almost certain this was the case. However, I will retry both 2000 CPUs, with "failsafe" BIOS settings. However, while the barton 2400 was installed, I tried this -

BIOS settings = "Optimized" settings, not "failsafe".
OC'g settings = 140 FSB

I get to the desktop and run Sandra:

(real) Speed of CPU = 1750mhz
FSB = 280, RAM = 140
Sandra CPU scores = fractionally better than a 2200+ CPU, which is to be expected.
CPU temp = 24C. That's not bad!
Prime95 (max FPU stress) = doesn't fail and CPU temp is still 24C!

My early conclusion is: forget the stove hot Palomino 2000. Stick in a cooler Barton (with 512k cache) instead, and overclock it.

Can't find any info of a Barton 2400+ been made....sure it's not a 2500+?

http://www.cpu-world.com/CPUs/K7/AMD-Mobile%2 … H2400FQQ4C.html

retro games 100 wrote:

http://www.cpu-world.com/CPUs/K7/AMD-Mobile%2 … H2400FQQ4C.html

Ahh..it was a "mobile" versiom that explains it

Desktop Barton's starts at 2500+

I have a 3200+ Barton CPU in my Asus A7N8X deluxe mobo..use to run Amithlon/windows XP on that machine but a fast Pentium 4 eats it for breakfast 😀

Here's an interesting update. I just tried a T-Bred 2400. It POSTs OK, with a BIOS POST speed reading of 2000mhz. That's good, but the downside is that I cannot get the mobo to POST after I have adjusted the BIOS FSB setting to 133. Everything is OK, so long as I keep the FSB set to 100. Curiously, Sandra reports that the multiplier is 20x! So, at the windows desktop I run these benchies -

Sandra: dry = 5471, wet = 2711. These are good scores. They comfortably beat the 2200+ comparison scores.

I go back to the BIOS, and meddle with it once more. This time, I set the multiplier to "above 12x", and instead of selecting 133/33 for the FSB, I pick something a bit more cautious: 107/35. The mobo POSTs OK, and I get to the desktop. Time for another benchie -

Sandra: dry = 5945, wet = 2949. Not bad eh?

More meddling in the BIOS. This time I set the FSB to 112/37, which gives me a BIOS POST speed of 2240mhz. Hehe, this is fun! One last benchie. I run Sandra, but the poor thing goes mad and I don't get any scorez. Oh well.

I've just realised why those temperature readings of approx 24 were so low. When I installed one of the 2000 CPUs, it had "funny looking lumps" on its underside that pressed up against the mobo's CPU temperature sensor. I decided to push this temp sensor downwards, to give the CPU some more room to sit in the CPU socket more comfortably. Consequently, I think there's now a gap between the newly installed 2400 CPU and this sensor. This may explain why the temperature is being reported as only 24C.

Interesting update #2. I've got another KT7A board, but it's PCB revision is 1.3. It's non-raid. Unfortunately its BIOS is corrupt, and it's been flashed with a BIOS file suitable for a PCB revision number less than 1.3. It still works though. Anyway, I put the 2400 t-bred in to it, and set the BIOS settings to "failsafe". The CPU speed is reported on the BIOS POST screen as a speedy 2200mhz. Sandra tells me that the multiplier used is 22.5x! Curiously, the FSB is reported as 98, not 100.

Sandra's CPU scores are 6055, 2985. These are great values, IMHO.
Quake2 @ 1024x768 = 176.9 fps (using "crappy" nVidia 440 MX)
3DMark 99 = 7274 3Dmarks, 25959 CPU 3Dmarks (440 MX)

I've just noticed something a bit unusual about an Abit KT7A non-raid rev 1.2 board. I'm testing a t-bred 2000+ rated CPU in it, and sometimes when I cold boot, the keyboard lights momentarily flash but I get no POST. The way to get around this, is to just press the Reset button on the mobo when you see no POST activity on the screen, and it will reboot then POST normally.