Asus P4T-E not capable of 133MHz FSB?

imgod2u

Senior member
Sep 16, 2000
993
0
0
I just read the GamePC overclocking article that was posted in General Hardware and in there it stated that the Asus P4T-E has some incompatibility in its clock generator that wouldn't allow a 2/4 AGP/PCI divider at 133MHz FSB. Is this true? Can someone else confirm this? I was planning on getting this board but not if I can't overclock it to 133MHz and beyond. Also, what are usually the acceptable limits of PCI and AGP clocks? I have an Adaptec 29160 SCSI controller, an SBLive!, a Netgear FA312 ethernet card, and an ATI TV-Wonder, would any of these devices have any problems with a 150MHz FSB? (37.5MHz PCI). I'm also planning on a Geforce 4 Ti4600, would that have problems with a 75MHz AGP bus? I've heard somewhere that the Geforces seem to handle AGP bus speeds below 80MHz fine but what about my PCI devices? If not then I may have to get a 2.2 GHz when a new (2.4GHz maybe) comes out and prices drop in order to reach my goal of 3GHz (water cooled of course). But my question is, is 133MHz FSB on an Asus P4T-E possible?
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
37.5pci and 75agp are nothing to worry about at all.
start to worry at the 41pci 91agp levels.

make it too 133fsb and you really have nothing to worry about ;)
everything would be in spec except for your cpu and the 533 system bus.

i dont know about your p4t-e question - that would suck tho.
 

StanFL

Senior member
Dec 30, 1999
697
0
76
My P4T-E will do 130fsb but won't do 133 under any cirumstances. It's really odd. The 130 is with the Rambus running at full speed though (PC 1040).
 

augiem

Senior member
Dec 20, 1999
746
0
76
Hi,

I ordered a northwood 1.6A and an Asus P4T-E a few days ago (before that GamePC article came out) and I am also worried about not being able to reach 133MHz bus.

Does anyone know if you can change the PCI/AGP divider? I downloaded the manual for the P4T-E from Asus, but I see no option for PCI/AGP divider. There is a RDRAM divider setting (set to PC600 for 3x) which would get the RAM running back in spec at 133 fsb (400MHz)

Dang, I really hope there is a way. I was all set to crank this thing up to 2.128 GHz while leaving the RDRAM and 400MHz.

Augie
 

KenAF

Senior member
Jan 6, 2002
684
0
0
Here's a useful bit on the Asus P4T-E from the site that brought us the first PC1200 benchmarks...

From: Sander Sassen - Hardware Analysis [mailto:removed@hardwareanalysis.com]
Sent: Saturday, February 09, 2002 3:05 PM
To: removed
Subject: Re: PC1066/PC1200 on Asus P4TE...Are you sure?


Hi Ken,

Yes, it was just a retail, off-the-shelf, P4T-E, nothing special, no special BIOS or anything else. However, the possible reason why GamePC couldn't get it higher is because the BIOS only allows for FSBs up to 133MHz, you'll have to use the jumpers to go beyond.

The manual however does not list these so you'll have to figure those out yourself, or use those for the P4T which uses the same clock generator as the P4T-E. Rambus recently demoed a similar configuration as I tested and benchmark to show the performance delta of PC1200, and they too used a P4T-E.

Have a good weekend,

Sander Sassen
 

KenAF

Senior member
Jan 6, 2002
684
0
0
There's also this thread over on the Asus forum where they got their RDRAM up to PC1066 speed. Apparently, the program CPUFSB works well to manipulate the FSB on this board from Windows.

The same user also has his Sandra results posted right here. Although is it just me, or do his RDRAM results look a bit low for PC1160 speed?

 

KenAF

Senior member
Jan 6, 2002
684
0
0
StanFL,

Please let us know if you can get higher than 132 using the mainboard switches or the CPUFSB program.

You can download CPUFSB right here. Remember to use the P4T for the PLL selection.
 

StanFL

Senior member
Dec 30, 1999
697
0
76
When I was trying the 133fsb setting in the bios it would post fine but give me a "unmountable boot volume" error while booting WinXP. So today I did a clean install of XP Pro while at the 133fsb setting. Things are running fine except I don't know much about the sys yet as far as PCI speed etc, I misplaced my Sandra Professional install file and am still looking for it.

I had tried CPUFSB and had less than stellar results. I did choose the P4T pll but regardless of how I set it, every boot up thereafter I got a "system has recovered from a serious error" message in XP. Even after uninstalling the proggie. It was an annoying nag screen and part of the reason I reinstalled XP (using sys restore didn't get rid of it either). I don't think I'd be happy long term with software overclocking even if it HAD worked for me so I doubt if I revisit it.

If I get around to playing with the dip switch overclocking I'll update this thread. Everything I've done to date has been thru the bios. I saw a post somewhere forum maybe, that linked to a 1006 bios at Asus' ftp site. The bios was gone when I got there but hopefully Asus is working on this.

 

StanFL

Senior member
Dec 30, 1999
697
0
76
omfg, this is too easy!! Just download the pdf manual for the regular socket 423 P4T from Asus and use the same dip swich settings for the P4T-E. They go as high as 150fsb. I'm trying to be conservative on both voltage and PCI/AGP speeds but now have my 1.6A at 2.2 gig. Core voltage is 1.55, 138 fsb, PCI is 34.5 hz, AGP 69.0 hz.

All this with RDRAM running at full speed which equates to about PC1100.

Life is good!
 

augiem

Senior member
Dec 20, 1999
746
0
76
>> omfg, this is too easy!! Just download the pdf manual for the regular socket 423 P4T from Asus and use the same dip swich settings
>> for the P4T-E. They go as high as 150fsb. I'm trying to be conservative on both voltage and PCI/AGP speeds but now have my 1.6A at
>> 2.2 gig. Core voltage is 1.55, 138 fsb, PCI is 34.5 hz, AGP 69.0 hz.
>> All this with RDRAM running at full speed which equates to about PC1100.

Wow, that's good news! So when you have it in jumper mode can you still select your voltages and memory divider in the bios? (In case my RAM doesn't make it that high... I hear 256MB sticks are not as hearty as 128MB.)

I can't believe your RAM is running that fast! Congrats! :)
 

KenAF

Senior member
Jan 6, 2002
684
0
0
augiem,

Below is a recent post from Hardocp on the P4T-E...

The Asus P4T-E has a ICS clockgen on it capable of running up to 156mhz FSB (manufacturer guaranteed) and also it has 2 ICS DRCG chips that are capable of running the rdram up to 600mhz DDR (PC1200, manufacturer guaranteed).

The only problem is these speeds are not able to be gotten through BIOS, you will need to refer to the ICS clockgen datasheet to be able to get the extended speeds through the onboard dip switches. The dip switches are especially important as there are 3 different 133mhz settings on that clockgen.

133.33mhz with 3/4 AGP, PCI/3
133.33mhz with 1/2 AGP, PCI/4
133.60mhz with 1/2 AGP, PCI/4

The best setting would be the bottom 2 as they run PCI and AGP within spec.

When running in dipswitch mode, you will not be able to increase the core voltage, so you will need to either mod the chip or mod your board to get higher voltages. Joining cpu pins with wires is probably better.
----------


and this information:

----------
The current BIOS disallows the proper 133mhz settings that run AGP and PCI within spec, it seems that the Asus board uses 2 types of clockgens. I've gone through a few boards to have a browse and some of them only come with the cypress semiconductor DRCG that will only allow up to pc800. So its both a BIOS issue and the clockgen on the board. Both boards i've seen are the SAME revision, that is 1.00.

Seems the newer Asus P4T-E boards are coming out with the pc1200 capable clockgens. Looks like they are doing "trial" runs for the new 533mhz rdram on existing boards. I reckon when the 533mhz P4 comes out, asus will make a slight revision and call their board a P4T-133E or something like that. It will be identical to the P4T-E, but with the newer clockgen and updated BIOS to allow 133mhz bus operations within spec.
----------
 

augiem

Senior member
Dec 20, 1999
746
0
76
>> When running in dipswitch mode, you will not be able to increase the core voltage, so you will need to either mod the chip or mod your
>> board to get higher voltages. Joining cpu pins with wires is probably better.

Ugh... I knew I should have gotten the Abit (it was cheaper too.) Well, hopefully the 1.6A will get to 2.1GHz without voltage mods.

>> The current BIOS disallows the proper 133mhz settings that run AGP and PCI within spec, it seems that the Asus board uses 2 types
>> of clockgens. I've gone through a few boards to have a browse and some of them only come with the cypress semiconductor DRCG
>> that will only allow up to pc800. So its both a BIOS issue and the clockgen on the board. Both boards i've seen are the SAME revision,
>> that is 1.00.

Hmm, well I hope I get one of the PC1200 capable DRCGs (otherwise it's going back to be replaced by an Abit -- yeah, I know, restocking fees.)

Thanks for all the help!

Augie
 

KenAF

Senior member
Jan 6, 2002
684
0
0
aug,

Don't give up hope...there are a number of people with the 1.6A that have been able to hit 2400MHz with the P4T-E at default voltage with their RDRAM @ PC1200 speed (although most people can't...they need the increased voltage that you lose with the mainboard switches) . StanFL was able to hit 2400MHz with a 1.6A on his P4T-E, but he had to reduce the clock of his RDRAM back to PC800 to get that...cause his Mushkin RDRAM couldn't reach PC1200 levels (although it could reach PC1100).

Stay tuned to this thread for his results with the new Samsung RDRAM that he is getting...
 

KenAF

Senior member
Jan 6, 2002
684
0
0
Another important update on the P4T-E....

There are two versions of the P4T-E, both of which share the same revision. The newest version uses ICS brand 9212-13 DRCGs, while the older uses CYP 632081 DRCGs. Only the boards with the ICS chips can achieve stable operation at 533FSB and above. There are two DRCG chips on each board, located near the RAM slots.

Before buying a board, attempt to confirm with the vendor that the two chips located near the RAM slots have "ICS" (and not CYP) listed on them; that will ensure that 533+MHz memory is supported.

Note even if you have the version with the superior ICS chips, you will still need to use the mainboard DIP switches in order to run RDRAM at 533+MHz. Through the bios options, it is not possible to access the 1/2 AGP and 1/4 PCI dividers that are necessary to attain stable operation at high memory speeds; the only way to access these dividers is through the use of the mainboard DIP switches. Download the P4T manual from Asus' web site for info on how to correctly set the DIP switches for various memory speeds between 533MHz (PC1066) and 600Mhz (PC1200).
 

augiem

Senior member
Dec 20, 1999
746
0
76
>> Don't give up hope...there are a number of people with the 1.6A that have been able to hit 2400MHz with the P4T-E at default voltage
>> with their RDRAM @ PC1200 speed

:) That's very good news. My original intention was to get to 2.1GHz w/400MHz RAM so as long as I can get that I'll be happy. I would like to see how far I can go, but I would probably not leave the system at 2.4 and PC1100 speed RAM -- I'm generally a bit conservative in my overclocking. (PPro 180-200, P2 333-400, Athlon 500-650)

>> StanFL was able to hit 2400MHz with a 1.6A on his P4T-E, but he had to reduce the clock of his RDRAM back to PC800 to
>> get that...cause his Mushkin RDRAM couldn't reach PC1200 levels (although it could reach PC1100).

I wonder what the performance difference would be -- I'd guess it would probably be faster running at PC1100 with a somewhat lower CPU clock. But then again, you may be pushing that PC800 RAM pretty hard...

When the proc is running at 2400MHz wouldn't the AGP/PCI be pretty far over spec? (even with the larger divider)

>> There are two versions of the P4T-E, both of which share the same revision. The newest version uses ICS brand 9212-13 DRCGs,
>> while the older uses CYP 632081 DRCGs.

Hmm, that's a bit strange. That makes getting a P4T-E more of a hassle than a TH7-II. Ahh well, at least now I know what to look for when it comes. Thx!
 

KenAF

Senior member
Jan 6, 2002
684
0
0


<< When the proc is running at 2400MHz wouldn't the AGP/PCI be pretty far over spec? (even with the larger divider) >>

At 150x4=600MHz FSB (PC1200 memory), the AGP bus would be at 75MHz and the PCI at 37.5MHz when using the appropriate motherboard DIP switch settings. Just about every graphics card will work on a 75MHz AGP bus, and there are very few PCI boards that would have trouble at 37.5MHz PCI. Generally, people start to experience problems starting in the mid 80s for AGP, and in the low 40s for PCI.



<< That makes getting a P4T-E more of a hassle than a TH7-II. Ahh well, at least now I know what to look for when it comes. Thx! >>

Well, at least when you get the P4T-E, you'll know if it will do 533MHz by looking at the chips on the motherboard...many, if not most, of the TH7-II's won't do 533+MHz memory at all without modification, and nobody has yet figured out why.

Look at this P4T-E shot (see green area) to see the chips to look for...in determining whether your board is capable of running the memory at 533+MHz.
 

josh358

Junior Member
Feb 11, 2002
2
0
0
I've seen several posts that say that USB can fail if you overlock the P4B266 or P4T-E. Has anyone had any experience with this?
 

KenAF

Senior member
Jan 6, 2002
684
0
0
josh,

The Microsoft Intellimouse Explorer USB has problems in an overclocked PT4-E....other USB mice are fine, I believe. Also, I am not certain, but I believe you will only experience this problem if you use the bios to overclock; I have not heard of anyone experiencing this problem that used the mainboard switches. Again, when you use the mainboard switches, you gain access to the appropriate AGP/2 and PCI/4 dividers.
 

thegamer36

Member
Jun 5, 2001
35
0
61
So does anyone know where we can buy the newer ASUS P4T-E boards with the ICS brand 9212-13 DRCGs?
If anyone has ordered one recently, can you please post where you got it. Thanks.
 

bmg

Senior member
Mar 18, 2000
243
0
0
Anyone know what RDRAM clock generators are used on the TH7II?

BTW, got my P4T-E around Nov. 1 and it has the 9212-13 DRCG. I wonder if the newer boards are the ones with problems?
 

thegamer36

Member
Jun 5, 2001
35
0
61
I have access to an Abit TH7-II (No RAID) board and here is the value I received for the DRCG:

ICS 9212DF-03
0129

Thats exactly whats on the chip.

If you want more info on these chips with PDF files...do a search at yahoo on 9212-13 or 9212-03.

From what I can tell, the 9212-13 is a high speed clock generator providing up to 600MHz differential clock source.
The 9212-03 is a high speed clock generator providing 400MHz clock source.
So it looks like the 9212-13 is capable of higher speeds.