Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Well, that previous configuration with 3.60GHz at stock 1.25000V wasn't stable. Prime95 hit an error about 15 minutes into the blend test.

Oh well. I'll be running a test again at the 7.5x multiplier (3.0GHz) tonight and see how that goes. I'm really not that interested in overclocking the CPU, at least not in any way that would require a voltage increase.

[vacapp]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

At this point I could use any advice about further tweaking the advanced memory timings.

instead tCL/CAS 4 ??? maybe it does not need too much overclocked voltage ...

Also, is there any reason to worry about running the MCH core at 1.260V for 24/7 operation with the stock EP45-UD3R v1.1 chipset cooling?

I do not know how much heat will be produced if you increase a given amount of voltage, but I found in Intel's specification 1,21V as a given maximum of MCH core voltage. I would assume(hope?), Gigabyte makes stock cooling of a board to allow the usage of the maximum specifications of a given chipset .
I ve learned from the guides, the more heat and voltage, the less expectable lifetime. Something similar to this picture. But on the other hand if you do not make a huge OC (close to specified maximums I guess), it is possible that you change your rig much more earlier than you would realise the shortened lifecycle.

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

instead tCL/CAS 4 ??? maybe it does not need too much overclocked voltage ...

I couldn't get tCL/CAS stable at 4 (see this post).

I do not know how much heat will be produced if you increase a given amount of voltage, but I found in Intel's specification 1,21V as a given maximum of MCH core voltage. I would assume(hope?), Gigabyte makes stock cooling of a board to allow the usage of the maximum specifications of a given chipset .
I ve learned from the guides, the more heat and voltage, the less expectable lifetime. Something similar to this picture. But on the other hand if you do not make a huge OC (close to specified maximums I guess), it is possible that you change your rig much more earlier than you would realise the shortened lifecycle.

Yeah, I know greater heat and greater voltage mean shorter life in general, I was just looking to get an idea of whether running between 1.2V and 1.3V on the MCH consistently would be truly reckless, and it sounds like it's not.

Truthfully I'm starting to get discouraged. Even my seemingly "stable" configuration of 400MHz FSB, 7.5x CPU multiplier (3.0GHz), and 2.66C mem multiplier (1066 DDR) hard locked quite quickly when playing Left4Dead this evening. The only configuration that has ever been actually solid has been using auto/default settings. As soon as I start changing anything outside of that it gets shaky.

I guess I just may not have the patience to spend the weeks it will take to find good settings, changing just one thing at a time blindly and then stress testing over and over and over.

It would help if I had an actual confirmed understanding of the specific relationship between the components and the full effect of each BIOS setting. I may post a new thread with just my specific questions now that I've been working on this for a while and at least learned enough to perhaps ask intelligent ones.

[Lsdmeasap]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Ya, I think you may just be jumping around far to much in an attempt to learn to many things at once.

Decide what Speed you want to run, and work on that ONLY. And do take advantage of BIOS profiles if you are not already!! F11 on the MAIN page is save, and F12 is load

[KayossZero]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Ya, I think you may just be jumping around far to much in an attempt to learn to many things at once.

Decide what Speed you want to run, and work on that ONLY. And do take advantage of BIOS profiles if you are not already!! F11 on the MAIN page is save, and F12 is load

Agreed, F11 and F12 make the whole tweaking/OCing process slightly more tolerable.

[vacapp]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

I couldn't get tCL/CAS stable at 4 (see this post).

Yes, but what is the maximum specified voltage that your ram can use safely? It depends on - of course - which manufacturer has more OC capable moduls but see, I recommended it, because there are RAM sticks with a default of CAS 4 and the articleresults show also this CAS timing as an achievable oc : DDR2-1066, DDR2-1160!
Of course it could be, that your ram needs too much overvoltage -> heat... Was only a guess to get better performance.

Truthfully I'm starting to get discouraged. Even my seemingly "stable" configuration of 400MHz FSB, 7.5x CPU multiplier (3.0GHz), and 2.66C mem multiplier (1066 DDR) hard locked quite quickly when playing Left4Dead this evening. The only configuration that has ever been actually solid has been using auto/default settings. As soon as I start changing anything outside of that it gets shaky.

I guess I just may not have the patience to spend the weeks it will take to find good settings, changing just one thing at a time blindly and then stress testing over and over and over.

It would help if I had an actual confirmed understanding of the specific relationship between the components and the full effect of each BIOS setting. I may post a new thread with just my specific questions now that I've been working on this for a while and at least learned enough to perhaps ask intelligent ones.

All in all I have to notice you get to a similar conclusion like me.
For example you think you have a stable system and it doesnt work...

Do you know what differences the straps make besides determining the automatic settings for tRD? I'm having a lot of trouble, as the strap setting seems very much like a "black box", requiring brute force trial and error to get anything working.

If I had a better idea of the specific differences that were caused by using a particular strap it would help very much to narrow down what I need to adjust to achieve stability. Otherwise I'm sort of throwing adjustments at the system at random.

And I'm still really confused why 2.66D won't work at all when it's supposedly the correct strap for 400MHz FSB, but 2.66C works great at 400MHz when it's supposedly the 200MHz FSB strap!

I do not know whether straps have a connection to NBCC...

Let's take a look for a theoretical example (DDR2-1064 like you and maybe -FSB 399-400-401 nevermind- both straps work):
- default meaning 200MHz strap has better latency as the 400MHz, but less bandwidth
- but what is then with NBCC as you mentioned? Regardless the 200MHz strap, because of the calculated NBCC FSB (should be more relevant... ?) same bandwidth performance?
- does it mean you can have a much better tRD and same bandwidth, so the 200MHz strap gives a better overall system performance ?????

Can you use any strap if the multiplier fits to your DDR2 frequency?

according to Anandtech review

What happens if we can set the tRD independent of the strap in use? Of course, the strap selected would still dictate which memory dividers are available, but no longer are we forced to make this change solely for the purposes of modifying tRD.

Does it mean yes? But your system is still not stable... I am also confused.

What about updating your BIOS? (400MHz strap problem...)

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Yes, but what is the maximum specified voltage that your ram can use safely?

Look in my signature. It's Corsair, the exact model is Corsair Dominator TWIN2X4096-8500C5D. The specified voltage is 2.10V, and that's what I've been using. I've tried pushing up to 2.2, when I was trying to get CAS 4 stable at 500MHz (DDR 1000), but it didn't work. Well, who knows, the RAM might actually work at those, but some mysterious motherboard setting or interdependence I'm not aware of might still be screwing things up.

All in all I have to notice you get to a similar conclusion like me.
For example you think you have a stable system and it doesnt work...

Actually, a lot of the time it's not just thinking it's stable and being disappointed afterward. More often it's being compeltely mystified at why a particular configruation won't even boot, when everything I understand about the systems involved would indicate it should at least POST, if not run very stable since I was using very conservative settings.

Can you use any strap if the multiplier fits to your DDR2 frequency?

according to Anandtech review


Does it mean yes? But your system is still not stable... I am also confused.

Yes, yes, I read that article top to bottom. This is one of my key questions that I put in my new thread. Does the strap change anything but the available multipliers and the default setting for tRD? If not, then it does seem like I should be able to successfully use any multiplier regardless of the actual FSB, provided I adjust tRD correctly and my RAM is capable of the resulting frequency. However, my actual experiences seem to indicate that assumption is incorrect.

What about updating your BIOS? (400MHz strap problem...)

I tried updating to the new F7 version available here. It doesn't seem to have made any difference that I can tell.

[vacapp]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Look in my signature. It's Corsair, the exact model is Corsair Dominator TWIN2X4096-8500C5D. The specified voltage is 2.10V, and that's what I've been using. I've tried pushing up to 2.2, when I was trying to get CAS 4 stable at 500MHz (DDR 1000), but it didn't work. Well, who knows, the RAM might actually work at those, but some mysterious motherboard setting or interdependence I'm not aware of might still be screwing things up.

I ve checked, the problem will be that you have 2x2048MB. The GeiL rams I ve linked were only 2x1024MB. It seems thats the reason for such tight timings.

Actually, a lot of the time it's not just thinking it's stable and being disappointed afterward. More often it's being compeltely mystified at why a particular configruation won't even boot, when everything I understand about the systems involved would indicate it should at least POST, if not run very stable since I was using very conservative settings.

Yes, you are right of course, therefore I try also figure it out as much as possible.

Yes, yes, I read that article top to bottom. This is one of my key questions that I put in my new thread. Does the strap change anything but the available multipliers and the default setting for tRD? If not, then it does seem like I should be able to successfully use any multiplier regardless of the actual FSB, provided I adjust tRD correctly and my RAM is capable of the resulting frequency. However, my actual experiences seem to indicate that assumption is incorrect.

Yes, I would like to know also...

I tried updating to the new F7 version available here. It doesn't seem to have made any difference that I can tell.

Ok, was only a guess because of your trouble with the 400 strap...

[Outer Marker]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

I know this is a long shot, but I have a good feeling about this. Forget the specification of your RAM voltage. Push down, not up. Instead of setting 2.1V, set 2.000V in BIOS for DRAM voltage. This will adjust some termination and reference voltages on the RAM that I think (hope) will give you some stability.

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Yes, I would like to know also...

Sorry if I was snapping at you in my earlier post. I was just getting quite frustrated. My apologies.

In my new thread lsdmeasap has reported that choosing a given memory multiplier does indeed set the associated strap, just the same as specifying it directly.

I've now got a promising 400MHz FSB configuration in testing (also posted in the other thread), and hopefully will soon have things ironed out at my target 400 FSB / 533 RAM setup.

Cheers!

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

I know this is a long shot, but I have a good feeling about this. Forget the specification of your RAM voltage. Push down, not up. Instead of setting 2.1V, set 2.000V in BIOS for DRAM voltage. This will adjust some termination and reference voltages on the RAM that I think (hope) will give you some stability.

Interesting idea. It looks like I've got a stable 400MHz configuration going now, but I'll give undervolting the RAM a try to see if it can help running a lower tRD or possibly reach CAS 4.

In testing I have encountered situations where I had to undervolt the MCH to get a particular FSB & strap combination to run, so I can definitely see it's worth trying.

[Dalamar]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

It looks like I've dialed in a good configuration now that I'm satisfied with. It's posted in my new thread here. It's BIOS F7 with 400MHz FSB (x7.5 for 3.0GHz CPU), 533MHz RAM on the 2.66D multiplier, 5-5-5-15 timings and tRD 7, and all stock voltages. It's passed 24-hr Memtest86+ v2.11, and I'll be putting it through a 24-hr prime95 blend run soon, which I fully expect it will pass.

Thanks everyone who posted here for the help and support. I really appreciate it!

[Lsdmeasap]

Re: Help understanding GA-EP45-UD3R BIOS settings & Core2 FSB OC principles

Sounds like a good setup with solid timings and voltages, I will post in your other thread as well when I get that far. But, sounds good for sure!

Good job man!