Winchester 90nm and Prime95

SuperPI still fails at 2T command rate, running Prime95 now.

Edit: Failed after 4 mins, lol. Gonna load optimised defaults, if that fails then something is very wrong.

Edit2: Trying optimised defaults, SuperPI still fails. Running Prime95 now. BTW memtest86+ 1.4 runs fine, ran the extended test (7hrs 30 mins, ugh :) in 1.3 and that was fine as well.

Edit3: Prime95 has been running for 3 hours now without a problem, which is the longest yet. Not holding my breath though ;).

Edit4: Failed after 2 hours 23 minutes. I'll probably be buying a Newcastle soon...

Quote:

---

Originally Posted by IvanAndreevich

Badge56
Can you test the highest your combo will run 1:1? It seems that people having this glitch are all 1:1!

---

With the Corsair 220 was the highest..

Testing with Mushkin now..

I think that the ones that can not run Prime95 at stock speeds are either not timing the RAM correctly... to agressive. Relax those timings !!

With 2 sticks timings MUST be relaxed a lot with winnies.

I have a freind that was having a hell of a time getting his system stable and was using a newcastle. He owns a store so has access to a lot of RAM.
I told him it was a RAM issue and shure enough it was..

I had tried a whole bunch of cheap sticks but they where always timmed to agressively....

yes like george said try to get media involved if people start complaining about the chips amd will notice it you just have to make sure it hits all around the same time not just randomly make a spred like wildfire sort one place starts talking about it then 2 then 3 then 5 so amd notices that there is soemthing wrong it will draw a lot more attention.

first post...sup everyone.

no problems with my week 33 3200+ Winchester OEM in rig #2 and prime95. no problems with rig #1 either.

specs of my 2 rigs:

Rig #1:
- AMD A64 SH7-C0 3200+@2.2ghz (220x10@1.5v)
- TT Venus 7+Cu
- ASUS K8V Deluxe (1009.003)
- Corsair TwinX1024-4000PRO (2x512 DDR440 2.5-3-3-7@2.6v)
- Western Digital WD360GD Raptor (2x36gb Via RAID-0)
- Western Digital WD2500JD (1x250gb Promise SATA)
- Samsung SD-816BRPS
- Plextor PX-PREMIUM
- ASUS Radeon 9800XT
- Gateway VX900 19in CRT
- Creative SB Audigy2ZS Platinum
- Klipsch ProMedia Ultra 5.1
- OCZ 520ADJ PowerStream
- Kingwin KT424SWM Aluminum Case

Rig #2:
- AMD A64 DH8-D0 3200+@2.4ghz (240x10@1.475v)
(board overvolts by .025-.05 according to cpuz)
- Thermalright XP-90 / 92mm Antec SmartFan
- ASUS A8V Deluxe (1009.005)
- OCZ Platinum Rev2 (2x512 DDR480 2.5-3-3-5@2.8v)
- Western Digital WD740GD Raptor (2x74gb Via RAID-0)
- Western Digital WD2500SD (1x250gb Promise SATA)
- Plextor PX-716A
- Plextor PX-708UF
- eVGA 6800LE w/ThugRook's 1.4v GT BIOS@425/1150
- Viewsonic VP201s lcd
- OCZ 520ADJ PowerStream
- Lian-Li V1000B Aluminum Case

Quote:

---

Originally Posted by drpino

first post...sup everyone.

no problems with my week 33 3200+ Winchester OEM in rig #2 and prime95. no problems with rig #1 either.

specs of my 2 rigs:

Rig #1:
- AMD A64 SH7-C0 3200+@2.2ghz (220x10@1.5v)
- TT Venus 7+Cu
- ASUS K8V Deluxe (1009.003)
- Corsair TwinX1024-4000PRO (2x512 DDR440 2.5-3-3-7@2.6v)
- Western Digital WD360GD Raptor (2x36gb Via RAID-0)
- Western Digital WD2500JD (1x250gb Promise SATA)
- Samsung SD-816BRPS
- Plextor PX-PREMIUM
- ASUS Radeon 9800XT
- Gateway VX900 19in CRT
- Creative SB Audigy2ZS Platinum
- Klipsch ProMedia Ultra 5.1
- OCZ 520ADJ PowerStream
- Kingwin KT424SWM Aluminum Case

Rig #2:
- AMD A64 DH8-D0 3200+@2.4ghz (240x10@1.475v)
(board overvolts by .025-.05 according to cpuz)
- Thermalright XP-90 / 92mm Antec SmartFan
- ASUS A8V Deluxe (1009.005)
- OCZ Platinum Rev2 (2x512 DDR480 2.5-3-3-5@2.8v)
- Western Digital WD740GD Raptor (2x74gb Via RAID-0)
- Western Digital WD2500SD (1x250gb Promise SATA)
- Plextor PX-716A
- Plextor PX-708UF
- eVGA 6800LE w/ThugRook's 1.4v GT BIOS@425/1150
- Viewsonic VP201s lcd
- OCZ 520ADJ PowerStream
- Lian-Li V1000B Aluminum Case

---

Both nice rigs there drpino and nice to hear that you not having any problems with p95.
Have you tried overclocking your 3200+ past 2.5Ghz? If yes was it prime stable?

BTW Welcome to XS :toast: :banana: :banana4: :banana:

Quote:

---

Originally Posted by [ r2 ]

Both nice rigs there drpino and nice to hear that you not having any problems with p95.
Have you tried overclocking your 3200+ past 2.5Ghz? If yes was it prime stable?

BTW Welcome to XS :toast: :banana: :banana4: :banana:

---

thanks for the welcome [ r2 ]. :toast: :D

i've taken stabs at 2.5ghz overall but nothing too too in-depth haven't primed above 247 which was my ceiling for superpi at 1:1 with these timings.... im satisfied for now but the bug will get me again sooner or later. when it does, i'll be sure to post back.

Hi


I just let my rig run P95 custom test 2048-4096 test heres the result after 14 hours
http://www.upit.dk/files/2048-4096.JPG

Funny thing some have trouble running stockspeeds, makes u wonder :).


Menty

Don't know if this has been posted, but I've heard of amd reps saying they were aware of the issues with the memory controller and they are going to be addressing it in the next revision of the winchesters or vinice. Well I saw this on the inquiror, so apprantly amd is bsing us. But hey, they'll be glad to sell you a new chip with a bunk memory controller on it as well.

http://www.theinquirer.net/?article=20262

Hmmm, why does AMD insit on relating the issue to overclocking? We here are specifically talking about Prime95 failures at stock speeds. Another thing; only the 90nm 3500? lol

At least the media coverage, that was spoken about earlier has begun :)

Everyone should try and rma their winnies to amd, claiming they fail at stock speeds, which they likely do, and maybe then they'll notice.

Well since replacing my winchester with newcastle ALL IS GOOD NOW.. Regardless of what AMD may say i can run stable speeds of 10x255 @ 2-2-2-10!!! and i couldnt with my winnie.. THis is a fact

The memory controller on the 90nanos is tragic

.. and guys, we made it in "The Inquirer" LOL

http://www.theinquirer.net/?article=20262

I find some of the statements in that article hillarious:

Quote:

---

An AMD representative said that the firm's techies say that because the Athlon 64 chips are locked and cannot be overclocked, an attempt to increase the mobo's system clock is causing all the components to respond, rather than just the microprocessor.

---

What is he talking about here? AGP/PCI lock or what? AFAIK, most good boards can do more than the CPU can handle.

Quote:

---

He said: "A small increase in the signal clock speed - say 5MHz on the 200MHz signal clock - is probably within the tolerances of all of the components".

---

So, AMD64 CPUs can only do about 5mhz OC on the FSB... Yeah, right :stick:

To me, if anything of the above was true, AMD wouldn't have been as succesful as they are today. Also, some have failed Prime95 at stock speed IIRC.

Sound like :bs: to me.

Quote:

---

Originally Posted by Orker

Well since replacing my winchester with newcastle ALL IS GOOD NOW.. Regardless of what AMD may say i can run stable speeds of 10x255 @ 2-2-2-10!!! and i couldnt with my winnie.. THis is a fact

The memory controller on the 90nanos is tragic

.. and guys, we made it in "The Inquirer" LOL

http://www.theinquirer.net/?article=20262

---

I saw this when I opened firefox this morning. My browser home is the inquirer. I was laughing after reading this. http://www.theinquirer.net/?article=20262

This has got to be a joke... "the firm's techies say that because the Athlon 64 chips are locked and cannot be overclocked, an attempt to increase the mobo's system clock is causing all the components to respond, rather than just the microprocessor." So the A64 is locked, but the FX is not so you can overclock the FX, but you can't overclock a regular A64. lol. This sounds like the level 1 newb $hit you hear when you call AMD on the phone LOL. I've gotten these kinds of responses before, not directly related to overclocking, but that same Level 1 (just graduated from Technical college kind of dudes on the phone. I think these may be the same dudes that pick up the phone at Asus.

LOL, I guess they spoke to the biggest retard at AMD. OCing the FSB of a CPU causes other components to respond?? YOU DON'T SAY!! My memory overclocks too???? NO WAY!!! You'd think referencing a forum called "Xtreme Systems", you'd expect a little more than total newbiness. But this is where this particularly ignorant AMD "techie" missed the point:

Quote:

---

Originally Posted by dumb AMD guy

He said: "A small increase in the signal clock speed - say 5MHz on the 200MHz signal clock - is probably within the tolerances of all of the components".

---

At no increase in the "signal clock speed" (does only AMD refer to FSB like this?), the CPU fails stability testing!

I guess that means we really have no idea of the concept of "stable" memory timings and clocks... or we have no idea what a good PSU is, or are totally mistaken that our $150 motherboards can't run a CPU @ stock with stable results!

Quote:

---

Originally Posted by dumb AMD guy again

But, he continued: "These overclockers are trying to get all teh components to respond to a 25 per cent increase in the speed of the signal clock, which would cause multiple components on the board to become unstable".

---

No Mr. AMD Technie, we don't know how to keep the RAM stable. God.

let me add a few things here (been ages since I last did that)

I came across A64 (FX) prime95 problems over a year ago already working for nVENTIV. At the time I was trying to specify the highest overclocked speeds "every" A64 FX would be able to do.

Prime95 kept giving problems. But i wouldnt straight out say that the systems were unstable because prime95 failed...because its a little more complicated than that.

First off I noticed that older versions of prime could complete just fine, whereas the latest version (at the time) would fail within seconds or minutes.

I had the luxury of using something like 7 or 8 different chips, and so I could compare. SOME batches did Prime better than others. The batches which were better at prime were not necessarily the best overclocking chips however. It all seemed fairly random.

But to give you an extreme example : I came across a few chips which would do absolutely EVERYTHING I threw at them @ 2800 MHz (phase change cooled of course) but would fail prime within seconds @ 2350 MHz same cooling. At the time I also talked to AMD about this issue, but they didnt really have anything meaningful to say about it. I didnt expect that either because if there was an explanation it would bound to be a highly complicated technical one.

Anyway I wasnt very confident in chips failing prime at such a low speed running much higher speeds so I did weeks of testing with any program I could find, and the result was that only prime95 kept messing up. I concluded that if an A64 fails prime95 at a given speed it doesnt necessarily mean its not completely stable. It means nothing more or nothing less than that it simply doesnt want to run prime95 at that speed. Since the problem CAN be eliminated by underclocking the CPU, it is definately an issue related to the CPU itself. It has absolutely nothing to do with FSB overclocking as I could play with the FX' multiplier.

So, do Athlon 64 chips have issues with prime95 : in my opinion that's a big YES. But somehow I'm not worried about it, I never managed to catch the chips making errors in real usable programs.

If your chip is failing Super Pi however that would be a serious concern. But prime95 alone will never dictate what is stable, for me. I also wouldnt blame Winchester too quickly, as I noticed huge differences in prime95 between different weeks and batches. One could run it at nearly every speed others would fail quickly.

But I can definately tell you that every member of the K8 family has issues with Prime. Some batches are better and some are worse but if you look hard enough you always find it. The "better" batches tended to be voltage hungry chips by the way. The ones that didnt stand voltage very well (though running cooler than the others) had a tendency to do worse at prime, but would most likely pass 3Dmark/Pi at the same speeds as the others

never did I meet any K8 though which would not run prime95 at default speed & voltage with the stock cooler.

that's my experience.

I had to send in an email to the Inq on that article...

Quote:

---

Originally Posted by Me

You should call AMD back and ask to speak to someone who actually has a relative clue as to what he's talking about, preferrably one whose knowledge is not so draconian that they refer to FSB as "signal clock speed". I know they'd love to sound more technical than they actually are, but the information you received from AMD is so tragically far from the truth that I'm surprised it was published in the first place. If I were AMD's PR department, I'd be jumping on this like white on rice trying to make it look like their "techies" actually have a real idea of how a CPU relates with the other system components.

You'd think if they read the thread on XtremeSystems, they might have a glimpse into the fact that these people might actually understand the concept of what it means for a system to be stable, especially seeing as that's the ultimate bottom line in overclocking components. Maybe AMD needs to fire these people who are hardly passable as "technies" and hire some of the forum members to do be their technical representatives, at least that way their responses wouldn't sound like they came from someone who thinks the only possible overclocking comes from raising the multiplier.

---

Damn, I was hoping Inquirer would do a better job covering this issue. They should SPECIFICALLY say that chips are failing at STOCK and that Prime95's author said that he's confident it's a hardware error.

Wazzupp :p:

My Winchester 3200+ is not stable in the PRIME95 nor when it is in stock.
Why?
I leave cpu vid in startup I can be this the problem?
E in superpi after 225mz it gives error. Why can be? I'm Use á bios 1.5x

Thanks, Mickey Mouse.

Quote:

---

Originally Posted by JCviggen

let me add a few things here (been ages since I last did that)

I came across A64 (FX) prime95 problems over a year ago already working for nVENTIV. At the time I was trying to specify the highest overclocked speeds "every" A64 FX would be able to do.

Prime95 kept giving problems. But i wouldnt straight out say that the systems were unstable because prime95 failed...because its a little more complicated than that.

First off I noticed that older versions of prime could complete just fine, whereas the latest version (at the time) would fail within seconds or minutes.

I had the luxury of using something like 7 or 8 different chips, and so I could compare. SOME batches did Prime better than others. The batches which were better at prime were not necessarily the best overclocking chips however. It all seemed fairly random.

But to give you an extreme example : I came across a few chips which would do absolutely EVERYTHING I threw at them @ 2800 MHz (phase change cooled of course) but would fail prime within seconds @ 2350 MHz same cooling. At the time I also talked to AMD about this issue, but they didnt really have anything meaningful to say about it. I didnt expect that either because if there was an explanation it would bound to be a highly complicated technical one.

Anyway I wasnt very confident in chips failing prime at such a low speed running much higher speeds so I did weeks of testing with any program I could find, and the result was that only prime95 kept messing up. I concluded that if an A64 fails prime95 at a given speed it doesnt necessarily mean its not completely stable. It means nothing more or nothing less than that it simply doesnt want to run prime95 at that speed. Since the problem CAN be eliminated by underclocking the CPU, it is definately an issue related to the CPU itself. It has absolutely nothing to do with FSB overclocking as I could play with the FX' multiplier.

So, do Athlon 64 chips have issues with prime95 : in my opinion that's a big YES. But somehow I'm not worried about it, I never managed to catch the chips making errors in real usable programs.

If your chip is failing Super Pi however that would be a serious concern. But prime95 alone will never dictate what is stable, for me. I also wouldnt blame Winchester too quickly, as I noticed huge differences in prime95 between different weeks and batches. One could run it at nearly every speed others would fail quickly.

But I can definately tell you that every member of the K8 family has issues with Prime. Some batches are better and some are worse but if you look hard enough you always find it. The "better" batches tended to be voltage hungry chips by the way. The ones that didnt stand voltage very well (though running cooler than the others) had a tendency to do worse at prime, but would most likely pass 3Dmark/Pi at the same speeds as the others

never did I meet any K8 though which would not run prime95 at default speed & voltage with the stock cooler.

that's my experience.

---

since you worked at that huge company and had the pleasure of doing all those stablility tests, i ask you now. what do YOU use for stability testing?

im so used to prime, blend, prioitiy 10, 72 hours. I hold true to this meathod till today.

i want to know what you think me, and the rest of the forum members should do to test our system are 100% like a rock

So, did anyone with Prime failing at stock tried to LOWER the multiplier ? Did it help ?

:stick: :stick: :stick: :stick: :stick:

Ok, maybe Athlon 64's in general have a problem running prime. However this does not mean that prime95 is not a good measure of stability, imo what it means is there is a hardware flaw somewhere in the A64 architecture that prevents it from running prime 95 in w/e algorithim that it uses. Obviously prime95 is not buggy, I have never had any problems with it, it is as stated a hardware problem, because it will pass no matter what at lower speeds (unless there is something wrong with the cpu of course). So there is obviously I think, a design flaw in the A64, and if this is the case, I don't think it is likely to be cleared up anytime soon since they may have to completely redesign the A64. We just don't notice it since, obviously, how many things we do calculate prime numbers in the same way prime95 does? It's not much of a problem but it's not cool that All Athlon 64's at some point, are broken in an aspect.

Quote:

---

Originally Posted by kryptobs2000

Ok, maybe Athlon 64's in general have a problem running prime. However this does not mean that prime95 is not a good measure of stability, imo what it means is there is a hardware flaw somewhere in the A64 architecture that prevents it from running prime 95 in w/e algorithim that it uses. Obviously prime95 is not buggy, I have never had any problems with it, it is as stated a hardware problem, because it will pass no matter what at lower speeds (unless there is something wrong with the cpu of course). So there is obviously I think, a design flaw in the A64, and if this is the case, I don't think it is likely to be cleared up anytime soon since they may have to completely redesign the A64. We just don't notice it since, obviously, how many things we do calculate prime numbers in the same way prime95 does? It's not much of a problem but it's not cool that All Athlon 64's at some point, are broken in an aspect.

---

Or Prime is broken or Windows is broken. I think you're being extremely short-sighted in your conclusions. What technical proof do you offer? There are all sorts of tools a programmer may use to provide the proof one needs. I'm betting on AMD being right on this one.

Ok then either the A64 is broken working with windows and prime95 or just prime95. Pentiums don't have this problem, nor do Athlon XP's. I don't see how you can deny the fact that there is obviously something wrong with the Athlon 64's and prime95. We know it's not a software problem (on primes fault anyways, I don't see why windows was brought into this, altho everyone knows it's a buggy pos). So the only other option would be hardware = A64.

Decided Problem

Memory 1T >>> 2T Prime Stable

Mickey Mouse :D :D :D

Quote:

---

Originally Posted by MouseSeries

Decided Problem

Memory 1T >>> 2T Prime Stable

Mickey Mouse :D :D :D

---

What date code is your CPU? 1t changed to 2t would indicate that your memory was the problem. 2t on the failing Winchesters does not help.

Quote:

---

Originally Posted by kryptobs2000

Ok then either the A64 is broken working with windows and prime95 or just prime95. Pentiums don't have this problem, nor do Athlon XP's. I don't see how you can deny the fact that there is obviously something wrong with the Athlon 64's and prime95. We know it's not a software problem (on primes fault anyways, I don't see why windows was brought into this, altho everyone knows it's a buggy pos). So the only other option would be hardware = A64.

---

I just Primed 14 straight hours with mine... 3200+@2500
Nothing wrong with A64 and Prime95... You need to configure the RAM better... Its that simple..

Quote:

---

Originally Posted by Badge56

I just Primed 14 straight hours with mine... 3200+@2500
Nothing wrong with A64 and Prime95... You need to configure the RAM better... Its that simple..

---

Have you even read all of this thread? Congrats on your chip working, but I don't think yours working means there is no problem with prime and A64. If you read the thead you will find that there are a few ppl with working chips, but also a lot that aren't.

Quote:

---

Originally Posted by Badge56

I just Primed 14 straight hours with mine... 3200+@2500
Nothing wrong with A64 and Prime95... You need to configure the RAM better... Its that simple..

---

lol, that made me laugh, do you work for amd?

Quote:

---

Originally Posted by Badge56

I just Primed 14 straight hours with mine... 3200+@2500
Nothing wrong with A64 and Prime95... You need to configure the RAM better... Its that simple..

---

Welcome to the discussion - don't be so lame next time you post. You have luck with yours, good for you.

Quote:

---

Originally Posted by Badge56

I just Primed 14 straight hours with mine... 3200+@2500
Nothing wrong with A64 and Prime95... You need to configure the RAM better... Its that simple..

---

WRONG.

What is your CPU date code? Some work fine. I don't know how else to put that. I work for a memory company. I help people properly configure memory for a living, Thank you.

Read all the test data and look at all screenies in this thread.

:)

From what I have seen.

pre week 41 don't like prime:(

After week 41 they love it :)

And it doesn't matter if you overclock or not, they fail at 200fsb stock voltage etc just as much as they do at 250fsb etc etc.

does that look right?

Quote:

---

Originally Posted by bigtoe

From what I have seen.

pre week 41 don't like prime:(

After week 41 they love it :)

And it doesn't matter if you overclock or not, they fail at 200fsb stock voltage etc just as much as they do at 250fsb etc etc.

does that look right?

---

Hey Bigtoe - I've yet to see an AMD64 LOVE Prime ;) ..but I know what u mean!

Quote:

---

Originally Posted by dnottis

Hey Bigtoe - I've yet to see an AMD64 LOVE Prime ;) ..but I know what u mean!

---

LOL. I know where your coming from Dean ;)

Quote:

---

Originally Posted by bigtoe

From what I have seen.

pre week 41 don't like prime:(

After week 41 they love it :)

And it doesn't matter if you overclock or not, they fail at 200fsb stock voltage etc just as much as they do at 250fsb etc etc.

does that look right?

---

I have a memory problem so I can't be to sure but at stock speed anyways my week 39 seems to be ok with prime blend. (I ran it overnight). Once I get some different ram that can run at low voltage I can comment more on my specific CPU.

I don't want to be offtopic but anyone of you red this article from Inquirer?

AMD says there's no problem with Winchester chips!

They added a link to this thread! Have you noticed the barbaraty that the guys on AMD said, like this: :stick: "An AMD representative said that the firm's techies say that because the Athlon 64 chips are locked and cannot be overclocked, an attempt to increase the mobo's system clock is causing all the components to respond, rather than just the microprocessor." :stick:

I really don't know if you guys noticed this, just wanted to keep you informed. If you don't want this kind of post to repeat, please let me know, I wont be offended! :toast:

Best wishes

Jorge Lopes

Yep, it has been posted many times.
I think that AMD is avoiding a straight answer because it doesn't have one. The reason could be that if something went wrong during the manufacturing processes of week 41, they missed it, and now it's difficult to find out what happened back then.

I had the toughest time with a 3000+ 939 (unknown date code). It could only pass prime in single channel. I tested this cpu in 4 different motherboards (ECS K2, MSI-NEO2, Abit-AV8, and Gigabye Ultra) I returned the cpu for a 3200+ 939 with a 32nd week code (0432 TPMW) and now can pass prime95 in dual channel at stock speeds.

andyOCZ
You tried running 2T on the Winchester that failed on ya? And also, if you still have the chip that fails, can you try running memory at a ratio? Seems like MANY people running memory with a ratio don't have prime failing.

Quote:

---

Originally Posted by IvanAndreevich

andyOCZ
You tried running 2T on the Winchester that failed on ya? And also, if you still have the chip that fails, can you try running memory at a ratio? Seems like MANY people running memory with a ratio don't have prime failing.

---

Thats not fixing the problem, that merely points out that the memory controller is bunk :stick:

Quote:

---

Originally Posted by dnottis

Hey Bigtoe - I've yet to see an AMD64 LOVE Prime ;) ..but I know what u mean!

---

Perhaps you mean 0.09 AMD64s. Both of the 754 Clawhammers I owned loved Prime. The Newcastle 939 I'm using now loves Prime. The Winchester 939 I owned along with another Winchester 939 I borrowed did not love Prime. ;)

There are other problems with A64, not just Prime95. Try for example Linux installation or boot floppies. To have that done you will need to use to lowest possible memory settings.

This has never ever happened to me with any of the Intel chips, or with Athlon XP. AMD keeps on denying the problems but memory controller on A64 Winchester is the main suspect. That memory controller is something AMD cannot get done right despite all the denials.

It's all in the CPU guys -
http://www.3dxtreme.net/other/3000%20wk41/293x9-2T.jpg


The whole thread is here -
http://xtremesystems.org/forums/showthread.php?t=48127

hey, I'm just wondering. What is the benefit of even having the memory controller on die? I can see the obvious benefits of less latency or something (I don't know how it all works, but just guessing). But other than that.. I just don't see any benefits. I mean, on crappy low end boards, sure they would have equaly crappy memory controllers, but who cares? On high end boards, say if the 754 dfi had it's own memory controller. I'm sure it'd have a really good memory controller easily capable of 300mhz+ 2-2-2-5 (dosn't mean the ram can do it I'm just saying). And you could have dual channel on the 754 cpu's. So whats the point? Is it just to cut costs or a promotional gimmick, or is there actually a real benefit?

Less latency, why do you think the AMD64 are killing the P4's in memory bandwidth even though both are dual channel?

Well I thought there was more to it than just that. I'm not by any means pretending to know what I'm talking about, but the A64's (and Athlon XP's) architecture are very different than the p4's (memory controller aside) and I thought there was something else as to why they were better. The Athlons have more pipelines or something and so they can process more at the same time than the P4's. Correct me if I'm wrong, but.. I just thought that was why they have better memory performance.

Theres a rather easy way to find out if the prime95 problem is a hardware issue but it requires you to have access to a c++ compiler. Try compiling a copy of the latest linux kernel. Software compilation is way more sensitive to hardware errors than most other software and should definitly throw a fit if the cpu or memory error at all. I get compile errors on my system at speeds lower than what prime95 or memtest say are 100% stable even after 48 hours when overclocking. An if my memory serves me correctly this was the same way the pentium 3 1.13 debacle was brought to light forcing intel to recall those chips.

Yup. Linux kernel compiling is a good thing to try as well.
I noticed the problem loading Raid drivers from floppies, it was impossible to do until I figured out - set memory 2T, on the lowest possible speed and down to 200 bus. Then it worked. Therefore it is the memory controller. The same memory I used in Pentium 4C Linux installation, there it wasn't any problem all the way.

Quote:

---

Originally Posted by andyOCZ

WRONG.

What is your CPU date code? Some work fine. I don't know how else to put that. I work for a memory company. I help people properly configure memory for a living, Thank you.

Read all the test data and look at all screenies in this thread.

:)

---

OK... you are probably correct. I only have ONE winnie to play with. However many people :banana::banana::banana::banana::banana: and complain and don't configure there RAM correctly. The winnies, mine in any case does NOT like it much when two sticks are used so using AUTO or SPD in BIOS is a guarantie that it will not post or be verry unstable. I have played with both Corsair and Mushkin and results are similar... :stick:

Quote:

---

Originally Posted by bigtoe

From what I have seen.

pre week 41 don't like prime:(

After week 41 they love it :)

And it doesn't matter if you overclock or not, they fail at 200fsb stock voltage etc just as much as they do at 250fsb etc etc.

does that look right?

---

as i posted before, my week 33 3200+ primes fine so im not sure if it's as simple as pre wk 41 vs post wk 41 (and it's not just wk 41 afaik).

Quote:

---

Originally Posted by drpino

as i posted before, my week 33 3200+ primes fine so im not sure if it's as simple as pre wk 41 vs post wk 41 (and it's not just wk 41 afaik).

---

Quote:

---

Originally Posted by drpino

AMD A64 DH8-D0 3200+@2.4ghz (240x10@1.475v)

---

You never took it over 250mhz tho, which is often where the problem shows up, usually between 250-260. Some showup at 200mhz but either way. See if you can get it past 250mhz if ur ram can take it.

SUMMARY so far:

Prime95 is ok. If you fail to understand this, read the whole thread again and concentrate on constancy, repeatability and accurate predictability of the the errors (failures here are not accurately predictable, hence problem is in hardware).

Most A64s seem to be ok. At least some are provenly ok. They pass lots of prime95 testing at stock and overclocked speeds.

Some 90nm Winchester A64s appear to be failing at stock speeds. It is still not known if this is 100% due to the A64s themselves, or an an incompatibility between some A64s, some motherboards and/or some memory modules.

Also, it is important to notice that:

This thread should concentrate on problems with certain weekd code of 90nm Winchester A64 failing at stock speed, regardless of memory modules (default latencies) and Mobo model (def bios settings) used. The defining words are: stock speed.

Don't confuse it by talking about stability at overclocked speeds or too tight latencies.

If something fails at beyond stock speed, it is more than likely the problem of the hardware stability (cpu, mem, controller or fluctuations in other pars of circuitry). Also, it does NOT help to debug this problem of failures at stock speed!

We are only concerned with STOCK speed here.

So, Prime95 is ok (in regards to this problem), but perhaps not all AMD A64s (certain week code) are not.

Just "perhaps" though, I at least haven't seen enough information to support a definitive conclusion.

People who fail their A64 90nm Winchesters at stock speed, would really do everybody a favour, if they could test the same cpu with various different motherboards and memory sticks at stock speed. Of course, this is a lot of work, so I can't demand that of anyone. Still, it would help a lot.

Only by doing futher testing is it possible to assess whether there are actually defective A64 chips in the circulation that are not fit for stable use even at stock speed.

Here 'stable use' means calculations of ANY type that are of legal type for the cpu in question and which should NOT result in a failure.

My own hunch is that the problems are due to component tolerancies and incompatibilities and there could be some defective A64s in the circulation as well, in the sense that they are _overly_ sensitive to the components they are paired with (mobos, psus, dimms, etc.)

regards,
halcyon

PS If you have trouble understanding why prime95 or some other piece of scientific calculations that uses self-testing iterative calculation can fail, while almost all other apps work ok, then I suggest you pick up a basic computer science book and look up algorithmic accuracy as implemented in modern processors...

Quote:

---

Originally Posted by Badge56

OK... you are probably correct. I only have ONE winnie to play with. However many people :banana::banana::banana::banana::banana: and complain and don't configure there RAM correctly. The winnies, mine in any case does NOT like it much when two sticks are used so using AUTO or SPD in BIOS is a guarantie that it will not post or be verry unstable. I have played with both Corsair and Mushkin and results are similar... :stick:

---

The whole point was that most of us are using good quality matched dual channel kits. All of my memory had been pre-tested P95 stable on many platforms. I also have more RAM available to me than most for testing. :)

So, setting proven memory up properly on a proven board (P95 24 hours with a 4000+) with loose memory timings, agressive chipset timings off and having P95 fail at 200fsb indicates a REAL issue.

So far it seems mostly the cores used in week 37 are the problem.

I think if we are gonna come anyway farher with this we need too get some more struckture on this tread, look at my previous post link
from reading trou this tread I belive we are talkin about 2-3 CPUs that are failing prime95 at stock speed, where:
- one has been RMA
- one was ok, when goop was applied under northbridge
- one unknown status.

If these are the only chips that fails on "stock speed" people are makin a too big case out of this, just think of how many DOA mobos u read about here...

also I just tried a full reset of BIOS here, and in the memory section everyting is auto, except that 2T is set.
So I guess that 2T should be used then, in the definition of stock speed.

btw: I tried the newest version of Prime95 with blend setting last night, 12 hours no errors.
240x10 166/200 divider (200mhz on ram) with 2-2-2-5-1T on my week 37 winchester 3500+

Prime95 v23.5 runs fine on all my pc's listed below. Prime95 v23.8 runs fine on pc #1 & 3, but errors (memory errors) out on pc #2 after a couple of hours. I think it's a bug in Prime95 v23.8 since the older version runs fine on all my pc's. My suggestion is to try Prime95 v23.5 if you have problems with v23.8. You can download v23.5 from here: ftp://mersenne.org/gimps/p95v235.exe

PC #1
Antec SLK3700-BQE Case (120mm Fan in front and rear).
ENERMAX EG465P-VE(FM) 460Watt (Dual Fan) Power Supply (+12v = 33Amps)
ASUS A8V Deluxe Motherboard (bios v1008)
AMD Athlon64 3500+ (Winchester)
EVGA 6800GT (66.93 Drivers)
1Gb of Crucial PC3200 DDR RAM (2x512Mb) Running in Dual Channel Mode
Two 74Gb Western Digital Raptors
SoundBlaster Audigy ZS
NEC 3500A DVD+/-R
TEAC 1.44Mb FDD
Dell UltraSharp 2001fp
Windows XP Pro with SP2

PC #2
Antec Performance II SX630II Case (80mm Fan in front and rear).
ENERMAX EG465P-VE(FM) 460Watt (Dual Fan) Power Supply (+12v = 33Amps)
ASUS P4C800-E Deluxe Motherboard (bios v1018)
Intel 3.2C (Hyperthreading Enabled)
Zalman CNPS7000A-Cu Pure Copper CPU Cooler
ATI Radeon 9800Pro 128Mb (Catalyst v4.8 Drivers)
1Gb of Crucial PC3200 DDR RAM (2x512Mb) Running in Dual Channel Mode
One 120Gb Seagate HDD (model ST3120026A)
NEC 2500A DVD+/-R
TEAC 1.44Mb FDD
Viewsonic E95 Monitor
Windows XP Pro with SP2

PC #3
Antec Performance II SX630II Case (80mm Fan in front and rear).
Antec SL-350S 350W SmartPower (Single fan) Power Supply
ASUS A7N8X-E Deluxe Motherboard (bios v1012)
AMD 2800+ AthlonXP (Barton)
1Gb of Crucial PC2700 DDR RAM (2x512Mb) Running in Dual Channel Mode
ATI Radeon 9500 Pro 128Mb Video Card (Catalyst v4.8 Drivers)
One 120Gb Western Digital HDD (model ST3120026A)
Pioneer DVR-A04 DVD-R
TEAC 1.44Mb FDD
Viewsonic E95 Monitor
Windows XP Pro with SP2

Hammertail,

if your #2 system does NOT fail every single time in exact same spot, then it's more than likely NOT a prime95 bug.

For how long and how many runs have you tried each versions?

Do remember that failures, when due to hardware problems, can be sporadic: sometimes test runs are ok for 8 hours and then fail on two consecutive runs of 30 minutes, then work for 4 hours, etc.

My System doesn't fail prime at stock and oc.
0441RPMW 3500+ with A-Data 1024MB ! (OC Speed: 2.75GHz 1.52V)
Ram wont go over 230 FSB 1:1 Dualchannel 1T. => Prime and SuperPI fails instantly.
Mainboard is an AV8 rev 1.1 Bios 1.7

I think Prime is working correctly and is still a must have prog. to test stability. For 3D i recommend 3dmark 2001.

From my point of view, the AMD 64 Memory controller is the problem. It seems that there are some chips around that do not work correctly even at stock speeds. For all of you that have problems at stock speeds: Lower Vcore, Vdimm and underclock to 1.8GHZ and test again. If you come to a point where prime is stable, go up with your settings again.

Let me just be ignorant. Let me just blame everything on Prime95!!! :(

My Winchester 3500+ week 37 failed Prime95 v238 at stock speed.
Went out and bought 3200+ week 43 from a local store. Failed at stock speed.

Memory which I use 2x 512MB OCZ 3200 Platinum Rev2. Tried with 3 different settings 2-2-2-5@1T, 2-2-2-5@2T and 2-2-2-10@1T. Memory has been tested hours with Memtest no trouble. I know the memory is good.

Motherboard MSI Neo2 Platinum BIOS 1.41.
CPU and Memory volt set to default (at startup)
OCZ PS 520W.

What other things that I can try before I become 100% ignorant and blame on Prime95??

Thanks!

Quote:

---

Originally Posted by p0k

Let me just be ignorant. Let me just blame everything on Prime95!!! :(

My Winchester 3500+ week 37 failed Prime95 v238 at stock speed.
Went out and bought 3200+ week 43 from a local store. Failed at stock speed.

Memory which I use 2x 512MB OCZ 3200 Platinum Rev2. Tried with 3 different settings 2-2-2-5@1T, 2-2-2-5@2T and 2-2-2-10@1T. Memory has been tested hours with Memtest no trouble. I know the memory is good.

Motherboard MSI Neo2 Platinum BIOS 1.41.
CPU and Memory volt set to default (at startup)
OCZ PS 520W.

What other things that I can try before I become 100% ignorant and blame on Prime95??

Thanks!

---

Enable hidden BIOS options (shift+F2, then alt+F3), and set Drive Strength=Weak (in the Core Cell section). This helps with my OCZ EB modules. Good luck :)

Quote:

---

Originally Posted by lsevald

Enable hidden BIOS options (shift+F2, then alt+F3), and set Drive Strength=Weak (in the Core Cell section). This helps with my OCZ EB modules. Good luck :)

---

Thanks for the suggestion. But it did not help. Prime95 v235 and v238 failed at 1 minute mark. @stock speed with Winchester 3500+ week 37 and 3200+ week 43. Memory timing 2-2-2-5@2T. :rolleyes:

Quote:

---

Originally Posted by p0k

Thanks for the suggestion. But it did not help. Prime95 v235 and v238 failed at 1 minute mark. @stock speed with Winchester 3500+ week 37 and 3200+ week 43. Memory timing 2-2-2-5@2T. :rolleyes:

---

I would try a bit more relaxed RAM timings (2-3-3-7) and slightly higher Vddr (2.75V) before trying 2T :)

OCZ specifies 2.75V for that RAM, if I found the correct page that is:
http://www.ocztechnology.com/product...platinum_rev_2

I have no way to measure temps accurately. The motherboard shows an idle temp of 25C so I know its way off. However I am running at stock Vcore with an Thermalright XP120 so I don't believe temps to be the challenge.

I haven't put the multimeter to the board but I have a Fortron FSP530 that was rock solid in my AXP system.

The memory is Patriot XLB series TCCD chips that are memtest stable at 2.5 3 3 8 timings and it locks up at memory speeds as low as 230.

The system in question is a Gigabyte ultra 939 with a 3000+ winchester week 39.

It will lock up in Prime95 only. [EDIT] More than Prime is failing. Gonna try to get a different CPU I guess

I'm thinking - memory controller sucks?

:confused: :confused: :confused:

Quote:

Quote:

---

Originally Posted by p0k
Let me just be ignorant. Let me just blame everything on Prime95!!!

My Winchester 3500+ week 37 failed Prime95 v238 at stock speed.
Went out and bought 3200+ week 43 from a local store. Failed at stock speed.

---

the chances that u got 2 defect CPUs are rather slim.
have u checked under ur northbridge too see if there is enuf goop?
have u tried different bioses?

there does seems too be a thing in common on most of the reports with failing prime at stock, most pll with trouble seems to have this setup:
MSI K8N Neo2 Platinium
2x 512MB OCZ 3200 Platinum Rev2
OCZ Powersteam 520W

do winchesters have problems with memtest too?

kryptobs2000
No. I get the impression you never really read this thread carefully.

Oh, no I thought someone said they ran memtest and it failed and they were asking if it was because of the memory controller. I can't find it now, I thought it was p0k's post, but I think he just never mentioned it and Ijust misread memtest or something. Sorry made myself look kinda stupid there :p: please overlook that

It sounds stupid but lower the Vdimm: Set your RAM to 3-4-4-8 2T 200Mhz and to 2.6Vdimm. Give it a try, I am sure with this settings your system will pass Prime without a glitch.

My system doesnt run stable if i put more than 2.65Vdimm on it. !? :confused:

Thats probably because your ram dosn't like voltage, although 2.65 isn't much at all. And it still dosn't fix the problem.

My Ram likes voltage, but not in my A64 sys.

When i put my A-DATAs in my NF7S it runs stable with 2.9 V but not my ABIT AV8.

n0w4i4: Well bugger me! I just tried that and it completed 32M SuperPI! That's the first time this rig has ever passed it, I've tried slack timings but not low mem volts :lol:. I'll try stock timings of 2-2-2-5 1T but at 2.6Vdimm and see if it still passes, see my sig for my RAM etc.

Bleh, 2-2-2-5 1T fails straight away. Gonna go back to the slack timings and see if it's Prime95 stable. Defo a bad chip, waiting for my 3500 Newcastle as I'm not risking getting another bad one.

@Crunchy

You have tested with 3-4-4-8 and which Vdimm ? 2.6 or what ?

I tried 3-4-4-8 2T at 2.6Vdimm and it passed SuperPI 32M, 2-2-2-5 1T at 2.6Vdimm however does not pass :mad:.

I tried slack timings before but at 2.7Vdimm and they failed SuperPI 32M every time.

You see, thats a really strange thing with Vdimm and prime. I am just wondering how this is possible ?

n0w4i4
What kind of test fails with high VDIMM for ya? Large FFTs?

More and more I am leaning toward recommending newcastle over winchester if you plan to do 3500+ or higher.
I just got another newcastle and it did 2550 priming @ stock voltage on air. Don't remember any winchester I had doing this feat on stock voltage. Will try to get it up a little at a time and see where air ceiling of this will be with stock cooling. This might be the perfect replacement for my deceased newcastle that benched at 3.15ghz on mach 2. I always thought that chip should do air around 2.7ghz or so and if this one gets there I might be in luck again :)

Prepare urself for an unedjucated guess at whats the problem.

Maybe since it's 90nm there is of course more of a chance for crosstalk. At higher vdimm the voltage goes through the cpu as well, it is more prone to leak than the 130nm. So maybe thats why higher vdimm causes problems. The same thing can be causing problems since the same thing happens, to my knowledge with increased clock frequencies (and I don't know much about this so I could be competely wrong). So the whole problem might just be because it is 90nm and it is more prone to crosstalk than 130nm. And that is what happens after you get to 250fsb (and in some cases 200fsb which could be caused with 2.9 or more vdimm).

I'm not sure if the terminology I used is right, so sorry if it's hard to understand. For people who know what I'm talking about, does this sound like a logical cause for the problems?

edit: haha... I just noticed I spelled educated wrong when andy made fun of me there :p: ... I'll leave it cause it just makes me sound like I know what I'm talking about.

Quote:

---

Originally Posted by kryptobs2000

Prepare urself for an unedjucated guess at whats the problem.

Maybe since it's 90nm there is of course more of a chance for crosstalk. At higher vdimm the voltage goes through the cpu as well, it is more prone to leak than the 130nm. So maybe thats why higher vdimm causes problems. The same thing can be causing problems since the same thing happens, to my knowledge with increased clock frequencies (and I don't know much about this so I could be competely wrong). So the whole problem might just be because it is 90nm and it is more prone to crosstalk than 130nm. And that is what happens after you get to 250fsb (and in some cases 200fsb which could be caused with 2.9 or more vdimm).

I'm not sure if the terminology I used is right, so sorry if it's hard to understand. For people who know what I'm talking about, does this sound like a logical cause for the problems?

---

Yor unedumacated guess is a fairly sound theory. :) This is more than likey exactly what is going on. Some of the CPU's just leak or crosstalk more than others.

Quote:

---

Originally Posted by kryptobs2000

hey, I'm just wondering. What is the benefit of even having the memory controller on die? I can see the obvious benefits of less latency or something (I don't know how it all works, but just guessing). But other than that.. I just don't see any benefits. I mean, on crappy low end boards, sure they would have equaly crappy memory controllers, but who cares? On high end boards, say if the 754 dfi had it's own memory controller. I'm sure it'd have a really good memory controller easily capable of 300mhz+ 2-2-2-5 (dosn't mean the ram can do it I'm just saying). And you could have dual channel on the 754 cpu's. So whats the point? Is it just to cut costs or a promotional gimmick, or is there actually a real benefit?

---

A K8 without an ondie memory controller is called an Athlon XP. (well technically the K8 in 32 bit mode also has SSE2, slightly tweaked cache and TLBs, and a little bit longer pipeline, but none of this massively affects per clock scaler performance, the difference is the memory controller)

Try one out sometime if you want to see what difference it makes.

You mean try an Athlon XP out? I had a 2500+ and I had better memory bandwidth than I do now, but I think I was running it in dual channel too I'm not sure. I know I was running it in dc, but not sure if I every benched it in dc because if I did, my score only went up 200mbps or so which dosn't seem right.

>>I know I was running it in dc, but not sure if I every benched it in dc because if I did, my score only went up 200mbps or so which dosn't seem right.

That's because nForce2 is not a true dual channel chipset and the Athlon XP is limited by the FSB. Which is offtopic and now history.

Hello all,

I have a CBBHD 0444 Winnie, K8N Neo2 platinium, and went cheap on the RAM side cause I wanted a lot of it (2Gb) and couldn't afford the Crucial Ballistix 1Gb PC4000 sticks ... ended up with 4 x 512Mb MarkVision ValueRAM (88 Cdn$ a stick here in Ottawa)

Didn't run MemTest86 at first, directly installed Windows XP, started a quick overclocking to see what this puppy was able to, reached 2.5Ghz with not much trouble (DRAM max = 133), then started Prime95 and ... FATAL ERROR after 3 minutes.

Up-ed voltage 5% then 8.3% ... same pb.

Went back to 2.4GHz, ran 3h @+8.3% then same pb

Back to 1.8HHz, all defaults ... same pb !!!

Then I used MemTest86 at stock defaults, and found 1 memory location giving 5 to 10% memory errors - tested all 4 sticks separatly, only 1 exhibits this 1 memory location errors at 5 to 10% rate (even clocked down to DRAM max = 100MHz)

This is probably my Prime95 problem at stock defaults - will return this stick tomorrow, and update you if I can Prime stable afterwards - at stock speeds to start with.

Cheers
Dom

So you were running 2Gb(4x512) of ram?

dom993
Sounds like you have a screwed up stick?

The Small FFTs don´t fail, but Blend and, large fails, which sounds logical.
Superpi 32M also fails after 15min.

The best is Memtest: It doesnt give me any errors at all, while D00m3 is crashing after 10 min. ahha this prog realy sucks, i will never use it again on A64 Systems.

n0w4i4
Which test in memtest are you using?

The default, dont now exactly, but it ran about 5-6 hours over night.

I am not a memtest user, but people have been saying something about using test 5. Try it! Apparently it is much better.

I'll try it when i am back from work ! :upset:

To summarize my investigation so far:

At stock speeds and voltage both SuperPI and Prime95 fail but memtest86+ extended test (7.5hrs per pass) passes OK.

At 2-2-2-5 1T (stock) and 2.6Vdimm both SuperPI and Prime95 fail.

At 3-4-4-8 2T and 2.6Vdimm both SuperPI 32M and Prime95 (14hrs) pass!

I might try leaving my settings as they are and bumping the voltage to 2.7Vdimm (stock) to see what happens. I've also got a 3500 Newcastle coming so I'll be able to check that it is the chip that is the problem.

If you have tested your NC please post your results. I can't wait to know why higher Vdimm causes to crash Prime and superpi but not Memtest !?

My guess: 10% AV8 (Bios) related
89% Winchster related
1% PSU

Just tried 2.7Vdimm, that's all I changed and SuperPI 32M passed fine. Dunno if that disproves your idea. Later I'll try changing one setting at a time and see where it'll fail, I'll also Prime95 to check it's still stable. Not sure when my Newcastle is coming, it's coming by courier but I haven't had an email to confirm it's been dispatched.

Edit: just to confirm - this is 3-4-4-8 2T but at 2.7Vdimm (stock).

Mhm ok .. i think things will change as soon you try to oc HTT ~ +20 Mhz !

i just got mine to pass 8.5 hours in blend mode with the settings in my sig... i'm a happy camper!

btw its a week 37 cpu... :P

Quote:

---

Originally Posted by spacey

i just got mine to pass 8.5 hours in blend mode with the settings in my sig... i'm a happy camper!

btw its a week 37 cpu... :P

---

Nice going, is the reason you are using a devider for your ram is because they cant handle running 1:1?

week 41 90nm 3500
over 24 hours prime in blend mode
epox EP-9NDA3+ 1gb ocz plat rev 2 o/c to 2420 mhz and memory is at 1t 2-3-3-5
air cooled

This will be my test rig until my water stuff arrives so if you want me to try some settings LMK

BeerCan
Memory 1:1?

spacey
I also think that you should try running MAX 1:1 your memory can handle. See if it passes then ;)

Does anything other than Prime crash with these faulty winchesters?

Are they still ok for gaming????

as far as I know, mostly yeah. But they obviously have some problems hardware based so they are not 100% stable.

I just wanna know if it's still possible to play games like ut04 and doom 3 while running overclocked for a day without it crashing.

That's all i would need it for. :P