My New Ocz Pc-3700 Rev.2 Sucks!!!

I just ordered a 512mb x 2 dual channel enhanced latency gold edition kit for almost $300 off of newegg and the memory is not very good at all. It will boot up at 233fsb but it wont remain stable and that's the rated speed. I even upped the voltage to the max my mobo allows 2.85v and put the least aggresive timings on it and still no luck. The Corsair pc3500 that I am trying to replace with this stuff did better than that! I am definitely dissapointed with it. I was kind of hoping to be able to run it 1:1 with a 250fsb or higher but it can't even run at it's rated speed. I am really surprised after hearing so many good reviews of this exact same memory. I will have to RMA it and get something else. What would be the best choice for running 1:1 with my processor at a 300mhz fsb??? Currently I have had the corasir run 5:4 at a 275mhz fsb but I definitely want to have a 1:1 fsb if there is any memory capable of doing it.

Well some good PC4000+ Hynix chip sticks can't hurt. The timings will be high but you will get some really high speeds out of them. Just about all companies have some out, off the top of my head Buffalo is $65.00US per 256 stick for 4200 at newegg. I'm pretty sure there are 512 sticks listed also.

So as long as I get something with the newer Hynix chips I should probably be able to do 1:1 at 275mhz fsb or higher? I have never heard of Buffalo, that price seems kind of cheap. I need 2 512mb sticks that are qualified for dual channel.

Go talk to Bigtoe over at XS or abx about the OCZ.

with the newer Hynix chips I should probably be able to do 1:1 at 275mhz fsb or higher?

Click to expand...

No I don't think you can count on that.
It would require a degree of luck.

Corsair 4400 and OCZ 4400 are supposed to run that speed and
others might, but from what I've read in posts and reviews most PC4000 give up at about 260.

Since OCZ have been speed binning their Hynix for the 4400 I doubt if even recent 4200EL is likely to reach much above 270
without overvolting them.

Hitting 300(stable) with 2.85V would require getting some very lucky sticks, I can't remember seeing anyone reporting this kind of performance. 300 Usually requires 3V or over.

If you were aiming for 275 1:1, OCZ 3700 Gold would seem a somewhat strange choice. Though I have seen posts about it running 300 5:4 I don't think they will do much above 260 1:1.
I guess you are lucky they do not perform to spec so that you can RMA them without restocking fee

Good luck!
hepp

I'm in the same boat. Just got my 2x512 OCZ PC3700 Gold Rev 2 the other day. It won't run Prime95 or fold stable at it's rated speed and timings. Memtest86+ passes a couple of test loops w/o errors, but I guess I really need to run that longer. So far the highest stable FSB I've found using this memory is 225 - that sucks. My CPU is stable to FSB 239, so PC3700 seemed like the right choice for 1:1 operation. These sticks were to replace my 2x512 Mushkin PC3200 LvII, which were able to run 1:1 at the same FSB (225) at 2-3-2-6 and 2.85v. So I'm very unhappy with the new sticks. Others have great results with this memory, but yours and mine must have come from the bottom of the barrel.

If you find any answers about this memory, keep us posted.

bill gates said:

I just ordered a 512mb x 2 dual channel enhanced latency gold edition kit for almost $300 off of newegg and the memory is not very good at all. It will boot up at 233fsb but it wont remain stable and that's the rated speed. I even upped the voltage to the max my mobo allows 2.85v and put the least aggresive timings on it and still no luck. The Corsair pc3500 that I am trying to replace with this stuff did better than that! I am definitely dissapointed with it. I was kind of hoping to be able to run it 1:1 with a 250fsb or higher but it can't even run at it's rated speed. I am really surprised after hearing so many good reviews of this exact same memory. I will have to RMA it and get something else. What would be the best choice for running 1:1 with my processor at a 300mhz fsb??? Currently I have had the corasir run 5:4 at a 275mhz fsb but I definitely want to have a 1:1 fsb if there is any memory capable of doing it.

Click to expand...

If u want to run 300 you need to find some good pc3200bt- d43 chips and find a board that'll give u 3+ volts. another inexpensive way would be to get u some adata pc4000 which is inexpensive and crank up the voltage and cross your fingers. all the other stuff is so speed binned that unless u bought the ocz el pc4000 or equivalent when it first came out , u aint gonna have much of a chance of getting to 300 unless u pay thru the nose for some hand picked 4400.
for instance when the pc4000 first came out and it was the fastet ram, i bought some ram from komusa that had adata chips.
since they just checked for a minimum speed of 250mhz i had a good chance of having the chips go alot higher(which they in fact did 290+) now they offer pc4200 so their pc 4000 is unlikey to go real high cause those chips are saved for their 4200.

Great!
Mine will be here tomorrow. I am hoping it will run 245fsb.
Wonder why you two have the problems when I have read about 10 posts on different boards where they have gotten to at least that.

Try turning the Turbo off and trying it again.
One guy was able to reach 260fsb, just by doing this.

I don't want to have to turn off turbo but I will give it a try. I will be sending these back for sure but now I don't know what to get. The pc-4200 from ocz that is not the lowlatency is only like another $20 so maybe I will give it a shot. How can I get more voltage to the memory is it hard to do? I need a guide on how to volt mod the vdimm on an Asus P4P800 deluxe. If I could push 3.0v+ I would probably have much better luck. It just sucks to have a P4 that is stable at 285-300mhz fsb with no ram to go with it. And it sucks that the ram costs almost as much as the processor and motherboard combined! There has to be something out there that will work somewhere?????? As soon as DDR II comes out then were going to be screwed because they are going to stop advancing the limits of DDR and its going to be hard to get. What is the Hynix product code for the highest perfroming chips they make?

KrisMCool,
Your vcore seems a little low. Did you try to up it?

The WORST review I could find was this one:
http://www.overclockercafe.com/Reviews/memory/OCZ_PC3700_Gold-Rev2/pg_2.html

And it says:
"Kicking things up to the extreme last point of stability, we found a ceiling of 492MHz. I want to remind you that we hit 492 without changing any setting or adjusting the voltage. You can see that the bandwidth results at this level exceed the PC4000 reference results you saw in the previous screen shot. Impressive."

This was at 2.7v

After some searching in these forums I found a guide to mod the voltage to the ram and it doesn't seem to hard. Now I must go mod this thing and then I will post what my memory can do with 3.2v going to it! It better do more than 250mhz fsb or it's going back! $300 for memory that can't even run at these speeds is just a waste of money.

If memory passes Memtest86+ and fails at Prime95, it may not be the memory that is causing the instability.

Set timings to CAS Latency of 3, tRCD of 3, tRP of 3, and tRAS of 8. Have any PAT-settings off. Increase vDIMM to highest allowed by bios. Turn off any unused motherboard options, like Firewire and SATA (if not using). Start at ~DDR433 speeds (217MHz), and test memory stability by using Memtest86+ (1 run through - ~23 minutes; then loop Test #5 for a few runs - ~10 minutes). Use Prime95 to test general stability. Increase FSB in increments of 5MHz until you find out what it can do.

Then post back saying what it can, and can't do.

Thats why I said to try cranking the vcore.

what motherboards are you using ? for Asus remember to disable usb legacy or you will get errors in memtest

also with ocz pc3700 rev2 max will be around 250mhz 3-3-3-8 at 2.85v

Thanks Propilot and Speed_Mechanic for your replies and suggestions. I hope you can make something of my findings. First note that my previous memory (2x512 Mush PC3200 LvII) never showed these kinds of problems, even up to FSB 224 / DDR448 (1:1) @ 2-3-2-6, 2.85v, PAT enabled - not in Memtest, Prime nor folding. And I could run up to FSB 239 at 5:4, 2-2-2-5 w/o errors. Regardless of PAT, Turbo and voltage settings, with the OCZ @ 1:1, 2.5-3-3-7, Prime (blend torture test only, other tests pass) and folding (only w/ 2 instances running) produce errors above FSB 225 / DDR 450. I know I've got a crappy CPU, as it can't run stable at FSB 250 / 3.5 like so many others I've read about. I tried all voltage settings up to 1.625 while testing the CPU for max stable FSB to no avail - it crapped out when it crapped out, and increasing voltage seemed to produce errors faster. I know heat is not an issue - the highest temp I've ever recorded was 112°F by thermal probe attached to HSF where it meets IHS read by DigitalDoc5+, about 5° higher than AsusProbe and MBM report, and that was with 2 instances of Prime running. Normally it doesn't even hit 105°F. I've tried increasing AGP voltage to 1.6 w/ no change in results. I've heard others say that this also controls chipset voltage. Is this true, and if so I wonder why it isn't listed as such?

Setting the OCZ to 3-4-4-8 only got me to FSB 230 / DDR 460 stable. Because I can't get my system stable with the memory at it's rated speed and timings, I'm considering RMA. I just don't want to jump to conclusions yet. I'm also considering buying a MO stepping 2.8 to help rule out the CPU and most likely get better results anyway. Any advice would be appreciated. Thanks.

KrisMCool, you may need more vcore like 1.675v-1.725v on P4C800-E or you might have one crappy 2.8C - i'd try the memory with a 2.4C or 2.6C if you can still get them

1. make sure usb legacy is disabled
2. run memtest86+ v1.0 for a std test #1-7 loops and see where it errors if it's errors at test #5 it's probably memory speed, timings or vdimm, if it errors at test #1,2,3,4 it's probably cpu fsb, vcore issues

You can see my recent test results here: http://www.ocforums.com/showthread.php?s=&threadid=267030
I've had USB Legacy disabled since I read about it somewhere in these forums. I haven't tried a vcore setting above 1.625 yet - didn't want to make a crappy CPU even crappier by overvolting, but this point may be moot. Might give it a try - tried everything else I can think of so far. I also stop all unneccesary services for testing. Only thing running in the background is MBM, but not always.

I'm already shopping for a new CPU, so I'm open to suggestions.

bill gates

What did you end up doing with your set? Did you find a way to get better results? Or did you already RMA it?

I have decided to send them back to newegg. I played around with a lot of settings in the bios after talking with OCZ tech support and we just could not get them to run at their rated speed no matter what we did. The main thing you need to do is turn off turbo and memory accelration mode (P.A.T.). But that didn't work either so it is a bad batch of memory. And we tried to run the sticks 1 at a time and neither one of them would go to its rated speed. One thing I will say though is that OCZ Tecnology has the absolute best technical support I have ever had to deal with. Just for this fact alone I will re-order from their company again. They even offered to send me a hand-picked set of sticks but I decided that what I really want is some memory with the new Hynix chips so I can go to a 275mhz fsb 1:1 with my cpu, and I know these pc-3700 sticks can't go that high. I think the fact that they were offering to send me a hand-picked set says a lot for their great company. After seeing so many good reviews for this memory I would have to say these were just a bad set but its probably rare.