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

Memtest86+ successful but Machine Freezes

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

Barryng

Member
Joined
Nov 16, 2001
I just installed two Crucial DDR500 PC4000 3-4-4-8 Ballistix 1GB sticks (2 GB total).

At 250 MHz and 3-4-4-8 timing, the latest version of Memtest86+ runs flawlessly. However, Windows XP will eventually freeze. Sometimes it freezes immediately after it boots up, sometimes it takes hours of use.

For years Memtest86 has been very reliable in helping me determine a reliable speed for my memory. Why has it now failed to do this? Does the CPU : DRAM ratio effect something other than the memory?

This new memory is just a replacement for the memory I was previously running. I was running the old memory at a 5:4 ratio to keep it at 200 MHz as it was only PC-3500 (four 256 MB sticks PC3500 MUshkin Enhanced). The new memory (Two 1GB sticks PC4000) I was hoping to run at a 1:1 ratio and take advantage of the higher bus speed. The only thing that changed is the new memory. If I drop it back to a 5:4 ratio (200 MHz) it runs without a problem.

BTW, the mobo is an ABIT IC7-MAX3. The CPU is a 3.0 GHz Prescott at 3.75 GHz. This CPU has been running remarkably stable at this speed at very reasonable temperatures since last March. Max temperatures at 100% CPU load never exceed 140 F.
 
Have you increased your agp voltage? That also increases voltage to the northbridge/memory controller. I run mine at 1.65v @ 242 FSB (1.55 is default). What ram voltage are you running? Usually ddr500 ram wants around 2.8v.

I had a set of Corsair PC4400 that could memtest @ 275 MHz for days without error, but the system wasn't Windows stable at all. I gave up on the 14x multi with my 3.4E after that.
 
Mate, memtest stable does not imply Windows stability. Several people can tell you this from first hand experience. Yes, there is an overlap area where memtest stability gives you Windows stability, but it is not so when you hit the edge of memtest stablility. From personal experience, I would say the difference can be anywhere between 2MHz -10MHz. :)
 
I ran into the exact same problem with that exact same memory, 2gb crucial tracer ballistix DDR 500. I was so excited when I put it in (omg blue leds!!!11!!1) and it booted at 278 mhz. Sweet. Dual prime failed in like 2 seconds though. No problem, maybe that's a tad high. Backed off the OC a little - still unstable. Backed off more and more until I was at stock 250 mhz. Still unstable. I made sure timings were loosened all the way, experimented with voltages, 2T command rate but nothing could get it to run 250. Weird thing was memtest showed no errors, but it would fail prime within a minute or two. I had to go all the way down to 240 to get it prime stable.

I'm devastated now so I call Crucial tech support and explain my problem, looking for advice. The guy stops me in mid sentence and starts explaining how there is no such thing as DDR 500 ram(!!), that once the ram is over 200 mhz, it is "overclocked" and there is no guarantee of how high it will go. I tell him that the ram is RATED at 250 mhz at 3-4-4-8 timings so it should be able to run that! I consider those "stock timings" for DDR 500, but he won't conceed the point and gets dismissive with me and I give up.

I RMAed the ram. In hindsight, I guess putting better ram in wouldn't matter much since a Core 2 Duo would destroy my rig anyway. :cry:

Sorry for the rant.
 
RhoXS said:
I just installed two Crucial DDR500 PC4000 3-4-4-8 Ballistix 1GB sticks (2 GB total).

At 250 MHz and 3-4-4-8 timing, the latest version of Memtest86+ runs flawlessly. However, Windows XP will eventually freeze. Sometimes it freezes immediately after it boots up, sometimes it takes hours of use.

For years Memtest86 has been very reliable in helping me determine a reliable speed for my memory. Why has it now failed to do this? Does the CPU : DRAM ratio effect something other than the memory?

This new memory is just a replacement for the memory I was previously running. I was running the old memory at a 5:4 ratio to keep it at 200 MHz as it was only PC-3500 (four 256 MB sticks PC3500 MUshkin Enhanced). The new memory (Two 1GB sticks PC4000) I was hoping to run at a 1:1 ratio and take advantage of the higher bus speed. The only thing that changed is the new memory. If I drop it back to a 5:4 ratio (200 MHz) it runs without a problem.

BTW, the mobo is an ABIT IC7-MAX3. The CPU is a 3.0 GHz Prescott at 3.75 GHz. This CPU has been running remarkably stable at this speed at very reasonable temperatures since last March. Max temperatures at 100% CPU load never exceed 140 F.

Run Prime95. If it fails, then try reducing the speed of the memory until it's stable.
 
Just a followup:

Just after installing the new Crucial DDR500 memory a few weeks ago, the machine crashed a few times. The first few times the timings were set too fast (3-3-3-7 instead of 3-4-4-8). The last time the timings were set properly (at least I think they were). In any case, I have not observed a problem since. Both Memtest and Prime indicate no issues so I cannot now be sure if the timings were indeed set correctly the last time the machine froze.

There does not now seem to be a problem with this memory. It cooks away without issue at 250 MHz duing video transcoding tasks, etc. Maybe a little "burn in" was needed, maybe I inadvertently did not have the timings set properly when I thought they were, or maybe a little bit of both.
 
Back