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

I am getting very frustrated with this PQI!

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

BC|Wulf

Member
Joined
Feb 24, 2004
See sig below for system specs.

Ok I have about had it with this PQI memory I have. It may be time to RMA it for something different.

Here is my frustration. I have been playing with different overclock settings to try and get my computer 100% stable. At my current settings (220x11) I can run Prime95 using the small FFT test with no errors. As soon as I run the blend test I error out in under a minute. I have not run Memtest yet with these settings, but I would error out in test 5 when my settings were 220x10.

Is my memory crappy? It kinda pisses me off as I bought PC-4000 so I could keep the timings at stock and be able to run the FSB up to 250. So why does it seem that my memory is unstable at stock timings even though I am underclocking the memory?

Unless somebody can give me some pointers to solve this instability problem, I will probably RMA it and get something different. Anyone have suggestions for good memory. I am thinking I should have bought PC-3500 instead of going for PC-4000.

Thanks!
 
Last edited:
Which PQI did you get and what are the "stock" timings that you are using?

For some reason I can't see your sig, what are your system's particulars?
 
I have the PQI Turbo Series Dual Channel Platinum Kit 1GB(512MB x 2) PC-4000 running at 3-4-4-8 timings, 1:1 with the FSB and at 2.6V. If you need any more info let me know.

Here is the rest of my system:
AMD Athlon XP 2600+ Mobile (AQYHA 0421), 2.42 Ghz (220 x 11) at 1.60V
Abit NF7-S V2 with Abit Bios 27
Antec TruePower 550
 
Last edited:
Will the value of Vcore affect memory stability. Forgive the newbie questions, but my only other OC experience was on a crappy Abit SE-6 with a P3.

PQI's website lists the test voltage of the memory at 2.8V so I will up that from my current setting. My board only goes up to 2.9V for the RAM and I do not plan on doing any volt mods to my MB.

Whats the easiest way to test for memory stability while removing as many factors as possible? Should I set my CPU back to stock settings and voltage, reset the BIOS to factory settings and run memtest looping test 5?
 
If you can RMA it, RMA it for a PQI Turbo 2-2-2-5 ( TCCD )
Awesom stuff, I'm seeing 277 2.5-3-3-7 @ 2.75v

Suma.
 
I sent an email to PQi to pick their brain, and here is their official remark! :rolleyes:

"You are run that memory under clock!If you don`t want to run Max speed(depend how ratio between CPU:DRAM)the system will more stables.Your CPU can run at 250MHz or not?M/B can handle or not?....?As you know AMD plateform is not so good to over clock...!!!
We have been tested our memory and it can run 500MHz at 2.7-2.8V.
If you want to test the memory,you can get the memtest86 at www.memtest86.com
Our memory is lifetime warranty.!!!!"

So obviously PQI tech support is useless! :bang head
I am going to play around with some fo the settings in the BIOS to see if I can get this memory stable.

On a side note I read somewhere that Memtest can give incorrect errors! Is that true?
 
i wouldnt say PQI tech support is useless but they are defintly hard to understand
I think what they were tring to say is try to set your memory divider to a lower value to slow it down, if your still failing memtest or prime95 then its most likely not ur ram but either board or CPU limitaion, Im not taking there side in any means!!! if its rated at 250 then it better run at that, and hopefully yr rig can run those speeds?? also the hole intels overclock better i think they ment the are KNOWN (not in all cases) for higher FSB

But seriuosly try another ram didder just to eliminate some varibles Aloso some ram has trouble running at lower speeds
 
Last edited:
BC|Wulf said:
I have the PQI Turbo Series Dual Channel Platinum Kit 1GB(512MB x 2) PC-4000 running at 3-4-4-8 timings, 1:1 with the FSB and at 2.6V. If you need any more info let me know.

Here is the rest of my system:
AMD Athlon XP 2600+ Mobile (AQYHA 0421), 2.42 Ghz (220 x 11) at 1.60V
Abit NF7-S V2 with Abit Bios 27
Antec TruePower 550

They have a PC4000 with 2.5-3-3-7 timings which is TCCD based ram, just like the 2-2-2-5 stuff. If you can't get your RAM working right, then consider swapping for the TCCD stuff.

Is this a new system? In other words, is your board and cpu combo known to be capable of 220 plus while using the 11 multi?

Drop the multi down to 9 or so and see what you can do. In reality, few NF7s hit 250, many won't hit 240 so you may be directing your fustration in the wrong direction.

Let us know how things proceed.
 
You were right Reefa! I was directing my frustration at my memory and once I dropped my multi down to 10 I ran memtest for 8 hours without an error. I am running the blend test in Prime95 right now.

Granted I am only running 1.6v on the Vcore, so I should have more head room available for overclocking. I think it also helped that I disabled the FSB and AGP spread spectrum in the BIOS.

Thanks for everyones input. I am going to ride out this memory for awhile and see how it works out after I do the L12 mod and get the heatsinks on the mosfets and chipset.
 
I am not supprised, you have an athlon xp, those arent known to do very high fsb, also the mobos often dont like high fsb. Have you done the L12 trick? this sometimes gives people higher fsbs. I could never get over 220fsb stable back on my xp2100 with any ram but the same ram did 240 easy with my first 2800+ a64 so its not the ram but your cpu and mobo
 
OC550 what are u on about Athlon XP can run higher FSB's i could run mine at 227FSB with 1stick of my Adata in there but one stick wasnt stable at that even on its own

i would RMA it and get the TCCD ram from PQI
 
That doesnt mean every A-XP is like that maybe it was just your board or the cpu it doesnt mean just because your couldnt hit high FSB no ones elses can take alook in the Abit NF7-S thread people have hit 250+FSB
 
Steven4563 said:
That doesnt mean every A-XP is like that maybe it was just your board or the cpu it doesnt mean just because your couldnt hit high FSB no ones elses can take alook in the Abit NF7-S thread people have hit 250+FSB

Yes, there are people that have hit those speeds on NF7s, but for everyone of those there are probably 20 or more that don't. I think that most people would agree that high fsb on NF7s are a rarity, not the norm. Heck, even 240 is a very good OC for those boards. I believe that 220-235 is probably more along the lines of reasonable expectations.
 
One thing that can help with the "wall" is the 26d beta bios from tic tac...was able to run the M2600 at 245FSB stable when I had the M2600...this might also help your prime 95 issues also...

BTW...245FSB was on Kingston 2x256 DC value ram at 1:1 timings and 2.9v...
 
220 FSB is a good OC on the NF7-S with the official bioses. I run tictacs Manta Ray XT cpc off. I run 250 fsb 2.5-2-2-7 @2.9.(2x512) This is with OCZ TCCD chips.
Sandra bandwidth is running mid 3600's.
I have a bud that has a DFI LP 250 with the same OCZ TCCD chips runing low 3100's on Sandra.
Now some will say I'm just lucky, because this board is bone stock. I have seen many people on other forums getting this high. If you are a daring person, go ahead and run modded bios with cpc off. You will not be let down.
 
running cpc off takes such a big hit its not worth it. and I had to vmod my nf7-s in every way you name it, I did it just to hit 220fsb stable. Maybe I got a dog? but point in case, not all will do 240+
 
Overclocker550 said:
running cpc off takes such a big hit its not worth it. and I had to vmod my nf7-s in every way you name it, I did it just to hit 220fsb stable. Maybe I got a dog? but point in case, not all will do 240+
Same thing with my old rig, 240fsb with the modbios3 cpc off was like 220fsb with the d23 subzero cpc on.
 
Back