Memory timings at 4-4-4-12 for 11 hrs 32 mins zero errors. Now I may crank up the FSB a bit to 3.7Ghz tomorrow and let it cook all day.
I have been working with a Gigabyte guy still trying to resolve the memory overvoltage issure. They have been extremely difficult to work with. They seem to avoid my question all together. I got this piece of info for what it's worth. I didn't ask about this but this is what they gave me.
It is something interesting tho.
It will depend on the type of memory being used, if you are using default memory which are rated at 1.8v it will auto detect at that spec. If you are using memory which are rated at 2.2v bios will detect slightly higher due to the memory being rated this high. We have tested here with 1.8v-1.9v and 2.2v memory
1.8v-1.9v memory will detect at 1.8v and 2.2v detected close to 2v
It is something interesting tho.
It will depend on the type of memory being used, if you are using default memory which are rated at 1.8v it will auto detect at that spec. If you are using memory which are rated at 2.2v bios will detect slightly higher due to the memory being rated this high. We have tested here with 1.8v-1.9v and 2.2v memory
1.8v-1.9v memory will detect at 1.8v and 2.2v detected close to 2v
Maybe the board utilizes EPP's if they are available on the Ram to set these things? I don't know.
On a side note, I just had to rebuild the "family computer" and used the P35-ds3r this time, along with an E6750 and Corsair PC6400. Set the ratio to 1:1, upped the Dram voltage +0.2v, Ram to 400, 4-4-4-12, and left all other voltages alone. A swift and easy O/C to 3.2GHz, stable and cool. I love this CPU.
@mrfatboy - I don't know if you are still discussing bios issues with the Giga guys...but I noticed something interesting moving from one version of the Bios - F2 - to a higher, more recent version. That is, EIST (Enhanced Speedstep) is disabled in the F4 and F5 versions. The board will reduce multipliers on the P35C-DS3R (E6750 drops from x8 to x6) but the voltage doesn't drop. This is according to CPU-Z, which attempts to show real time vcore voltages. If I open Everest Ultimate Edition and take a look, it shows that EIST is supported, but disabled in the F4 and above bios versions. If I re-load my saved F2 bios, CPU-Z shows vcore voltages dropping to 0.963, and Everest confirms that EIST is supported AND enabled. Weird, don't you think?
Why am I concerned about this at all? Because if I go to Gigabyte's GA-P35C-DS3R motherboard page and look up the CPUs supported table, it says that the E6750 CPU @ 1333FSB is supported by the F4 bios and higher. Yet I'm overclocking just fine at a system bus of 1760 - 3.52GHz O/C. Is there really a need to load the F4 bios, in your opinion?
Regards,
I still have an active trouble ticket with Gigabyte but I have given up on them. They have been absolutely no use. I ask them a direct question and they totally answer something else if at all. Absolutely useless. I even asked the same question 5 different ways and they still don't answer it correctly. I'm not sure if they are avoiding the question, idiots, or don't understand english. I can understand the latter to a point but after 20 questions you would think they would get a guy that can understand english.
I am running F4g and EIST is still supported. My multiplier drops down to 6 when idle. Since my system is running fine at 3.6 I have been reluctant to upgrade the bios to the latter. File it under "if it ain't broke, don't fix it".
I am also under the impression that some of the bios functionality does not work if you are overclocking. Try setting your bios to stock settings and the turn on EIST/cia2 and others and see if voltage drops. There is even a temp readout in PC health for the Northbridge but that is only if you are at stock settings.
I am running F4g and EIST is still supported. My multiplier drops down to 6 when idle. Since my system is running fine at 3.6 I have been reluctant to upgrade the bios to the latter. File it under "if it ain't broke, don't fix it".
I am also under the impression that some of the bios functionality does not work if you are overclocking. Try setting your bios to stock settings and the turn on EIST/cia2 and others and see if voltage drops. There is even a temp readout in PC health for the Northbridge but that is only if you are at stock settings.
@mrfatboy - sorry to hear about your difficulties with their reps. As to F4g working (and maybe I'm making too much of this) with EIST, there are TWO power management systems at work here - C1E and EIST. The fact that your motherboard reduces the multiplier to x6 while implementing that bios version is almost certainly due to C1E functioning as it should. Everest Ultimate and CPU-Z confirms this on my rig too. What isn't happening is EIST - voltages are not being adjusted downward at idle - EIST is supported BUT is being disabled, no matter what you set your bios to do. Whereas in F2, it functions - meaning the board reduces the multiplier to x6 AND reduces vcore in a big way.
Maybe I'm missing something here and I hope someone can chime in, but what would be the sense of reducing the multiplier WITHOUT reducing vcore voltage? If the purpose of power management is to reduce energy consumption and reduce heat and stress on the CPU, why would a simple reduction in the FSB (thru the multiplier) without reducing the vcore voltage give any advantages at idle? You follow me? I'll bet you can confirm this yourself with the F4g bios: open up CPU-Z...you will see the multiplier reduction at idle...you will see that this multiplier is restored to x8 if you run, say, Orthos, but you will not see vcore voltages dropping at idle.
Regards,
Maybe I'm missing something here and I hope someone can chime in, but what would be the sense of reducing the multiplier WITHOUT reducing vcore voltage? If the purpose of power management is to reduce energy consumption and reduce heat and stress on the CPU, why would a simple reduction in the FSB (thru the multiplier) without reducing the vcore voltage give any advantages at idle? You follow me? I'll bet you can confirm this yourself with the F4g bios: open up CPU-Z...you will see the multiplier reduction at idle...you will see that this multiplier is restored to x8 if you run, say, Orthos, but you will not see vcore voltages dropping at idle.
Regards,
Lower the multipler does relieve some stress on the cpu. how much, i'm really not sure. It would be nice if the voltage was dropped also but I thought that function was disabled when you OC'd. I didn't know it worked in F2.
You are correct that the muliplier is only dropped at Idle. Which it should do. It would be nice if voltage dropped also When I fire up an app, i'm back up to 8 x 450
I guess it is something we have to live with with this board. Voltage regulation is not its strong suit.
You are correct that the muliplier is only dropped at Idle. Which it should do. It would be nice if voltage dropped also
When I fire up an app, i'm back up to 8 x 450 I guess it is something we have to live with with this board. Voltage regulation is not its strong suit.
Originally posted by: mrfatboy
Using F5, is the board still overvolting? if it is, they can't blame Everest anymore
Check my screenshot above. That's at +0.2V in BIOS.
Too bad we can't get any spillover to the CPU
lol, if i sent that screenshot to GGTS they still would not get it.
1.8v +.2v = 2.14v
me like math
1.8v +.2v = 2.14v
me like math
I have a 6750 and a DSR3 motherboard, with Corsair 4-4-4-12 DDR2 memory. I have a question: I can't find any way to fiddle with the timings in my BIOS, only that Memory Speed Multiplier. How do you set your timings? I'm only going for a moderate OC (atm 3.0 GHz), so can I live fine with 4-4-4-12?
@Prodigy - when you enter Bios setup (Delete key), and you're presented with the Bios menu, press ctrl-F1.
And yes, Everest has always reported the overvoltage. Either the board does not default to 1.8v on the DRAM, or the board overvolts when +0.2v is applied. There isn't any question about this. I don't see the importance, frankly, of getting TS to admit to it. It's a known parameter - you simply make the adjustment. I'd rather run the F2 bios and take advantage of EIST/C1E's power management undervolting the vcore at idle, than having the F5 bios tell me something I already know. Are there any other advantages to moving to F5? Stability? New features? Capabilities? I haven't found any of that.
And yes, Everest has always reported the overvoltage. Either the board does not default to 1.8v on the DRAM, or the board overvolts when +0.2v is applied. There isn't any question about this. I don't see the importance, frankly, of getting TS to admit to it. It's a known parameter - you simply make the adjustment. I'd rather run the F2 bios and take advantage of EIST/C1E's power management undervolting the vcore at idle, than having the F5 bios tell me something I already know. Are there any other advantages to moving to F5? Stability? New features? Capabilities? I haven't found any of that.
The reason to get gigabyte to admit to it is to have 100% piece of mind when it comes to setting up our boards and ram correctly. The manual says the board defaults to 1.8v for the ram. Most people out there (who are not reading this thread) believe this to be true and are probably overvolting their ram. Thus possiblly resulting in more ram failures then normal. There might be even a connection with the "reboot cycle" problem.
Gigabyte was the one that told me that Everest was the problem. It was reading our boards incorrectly and locking up the voltage sensor thus making all voltage reading software incorrectly. Total BS.
I'm 99.9% sure that this board is overvolting the ram. Gigabyte has flat out says it does not. Sort of bizarre don't you think? Why wouldn't they just say, ooops, you are right, just make the corrections. One of us is wrong I think there is enough evidence it isn't us
Gigabyte was the one that told me that Everest was the problem. It was reading our boards incorrectly and locking up the voltage sensor thus making all voltage reading software incorrectly. Total BS.
I'm 99.9% sure that this board is overvolting the ram. Gigabyte has flat out says it does not. Sort of bizarre don't you think? Why wouldn't they just say, ooops, you are right, just make the corrections. One of us is wrong
I think there is enough evidence it isn't us @mrfatboy - of course I agree with you. And it IS a shame if this "omission" on their part is leading to RMA's on the ram side. These are reasonably popular boards. I wonder why the memory community hasn't gone to Gigabyte with some strong objections?
But I love their boards - I really do.
Completely off topic and wondering aloud - the x38 boards are out, and to me, the only thing to get excited about is the PCIE 2.0 slots. I'm wondering if we'll see vid cards in the near term that can take advantage of the bandwidth potential, and two, I'm wondering if the Gigabyte boards that are labeled ver 2.0 have these slots...
See, my understanding is the P35 chipset does support PCIE 2.0 - it has to be implemented by the mobo manufacturer. And that's a reallt weird naming convention Gigabyte has, don't you think? Boards that are labeled ver 1.1 and then labeled ver 2.0 without any apparent physical change? It would be kind of sweet if the GA-P35C-DS3R ver 2.0 became the ultimate upgrade platform - Penryn, DDR3, AND PCIE2.0 capable....hmmm...
But I love their boards - I really do.
Completely off topic and wondering aloud - the x38 boards are out, and to me, the only thing to get excited about is the PCIE 2.0 slots. I'm wondering if we'll see vid cards in the near term that can take advantage of the bandwidth potential, and two, I'm wondering if the Gigabyte boards that are labeled ver 2.0 have these slots...
See, my understanding is the P35 chipset does support PCIE 2.0 - it has to be implemented by the mobo manufacturer. And that's a reallt weird naming convention Gigabyte has, don't you think? Boards that are labeled ver 1.1 and then labeled ver 2.0 without any apparent physical change? It would be kind of sweet if the GA-P35C-DS3R ver 2.0 became the ultimate upgrade platform - Penryn, DDR3, AND PCIE2.0 capable....hmmm...
I still like my board. It can act a little funky on reboots. And that dreaded "cycle reboot" problem is always hanging over my head. But at 3.6 24/7 OC I'm happy. BTW, the ver 2.0 of this board actually was a physical change. They got rid of the parallel and serial ports and added more USB.
@mrfatboy - I think we're past worrying about the cycle reboot problem, don't you think? I've been running stable at 3.52GHz 24/7 for about 90 days now (with C1E and EIST employed, natch) and I've popped this bios and that into it, and never had that issue, not once. Not sure you have either? What would make you think it would ever rear it's ugly head after all this time?
Regards,
Regards,
Nobody knows what caused or is causing the "cycle reboot" problem. It was never tied to any one bios either. People with bios versions f2 thru f4 had the problem. My theories include:
1) Bad batch of boards (my top theory)
or
2) A problem with this board in general that is suseptable to flakey/bad power supplys, overvolting, and overall voltage regulation that causes the bios to get corrupted.
Or a combination of both.
Since we don't have 100% proof what caused it I not sure that is had "passed".
Again, when I (and many others) confronted Gigabyte about this problem they just ignored us.
Who knows. I have not seen anymore posts about in awhile.
1) Bad batch of boards (my top theory)
or
2) A problem with this board in general that is suseptable to flakey/bad power supplys, overvolting, and overall voltage regulation that causes the bios to get corrupted.
Or a combination of both.
Since we don't have 100% proof what caused it I not sure that is had "passed".
Again, when I (and many others) confronted Gigabyte about this problem they just ignored us.
Who knows. I have not seen anymore posts about in awhile.
is anyone using the P35C-DS3R with 4 sticks of 1gig mem? ive got 2 sticks of corsair XMS2 Cas 4 and i want to add another 2 sticks to make 4 gig in total but im worried it might f8ck up my Overclock.
Does anyone know of problems using 4x1gig ?
Does anyone know of problems using 4x1gig ?
I believe somebody about 20 or so posts back (maybe more (about 3 weeks ago)) was getting the 4 sticks and they were supposed to report back on the OC situation. But they didn't as of yet. I just got my my 2 extra gigs of ballistix last week but have not put them in as fear of them screwing up my OC. Only $45 for 2gigs.
I could not resist!My buddy has the same rig as I do but with 4 gigs. He is at 3.2 @ stock volts on everything including ram. He has not tried anymore. He does'nt care.
I want to know also
Speak up if you got it!Hi guys
I´ve reading your posts for quite long time, and Im a neewbie on OC.
I could not overclock above 444 with my config.
I was afraid of increasing above 1.45 V and 0,3+.
Any help would be welcome.
I just want to know the best overclock to every day use in gaming.
Here, in Brazil, heat is an issue, when we easily have daylight temperatures of 40 C.
Abraços
From
Gaucho
12269 - 3dmark06
I´ve reading your posts for quite long time, and Im a neewbie on OC.
I could not overclock above 444 with my config.
I was afraid of increasing above 1.45 V and 0,3+.
Any help would be welcome.
I just want to know the best overclock to every day use in gaming.
Here, in Brazil, heat is an issue, when we easily have daylight temperatures of 40 C.
Abraços
From
Gaucho
12269 - 3dmark06
@sniperdaws
lol, I know I should but I have been too lazy and you are right I am a little nervous. I'm just waiting for this board to stop working I also been too busy cleaning up around my house. It almost burned down due to the California wildfires. I was going nuts without internet service for a week. I saved my rig tho
@Gaucho01
How did you get a 3dmark06 score higher than me? All my componets are clocked higher than yours. Were they clocked higher when you did the benchmark? What were the video care & ram settings then? Maybe it's time for me to move to the f5 bios If you needed to, I would not hesitate to put your vcore & (g)mch up 1 notch each to get 450 or higher.
Like you said your temps are your biggest problem. You just need to watch them and make sure they are in the safe range. What are your idle and stress temps? Refer to my OP for comparison temps.
Overall, your OC is good. You only need as much power as you need for gaming. BF2, COD4, Crysis, & Bioshock all run very smooth on my rig.
lol, I know I should but I have been too lazy and you are right I am a little nervous. I'm just waiting for this board to stop working
I also been too busy cleaning up around my house. It almost burned down due to the California wildfires. I was going nuts without internet service for a week. I saved my rig tho @Gaucho01
How did you get a 3dmark06 score higher than me?
All my componets are clocked higher than yours. Were they clocked higher when you did the benchmark? What were the video care & ram settings then? Maybe it's time for me to move to the f5 bios If you needed to, I would not hesitate to put your vcore & (g)mch up 1 notch each to get 450 or higher.Like you said your temps are your biggest problem. You just need to watch them and make sure they are in the safe range. What are your idle and stress temps? Refer to my OP for comparison temps.
Overall, your OC is good. You only need as much power as you need for gaming. BF2, COD4, Crysis, & Bioshock all run very smooth on my rig.
I just reran 3dmark06. I never redid the test after I updated the nvidia drivers. It made a difference but still not as good as Gaucho's. My vid card is even overclocked more than Gaucho's but his SM 2.0 & 3.0 scores still beat mine. Very bizarre. Anybody have any ideas? Maybe I'll post a comparasion in the vid card section. I also thought my 640 would be faster then the 320. All reviews state as such.
Mrfatboy's
3DMark Score 12231 3DMarks
SM 2.0 Score 5486 Marks
SM 3.0 Score 5270 Marks
CPU Score 3236 Marks
Guacho's
3DMark Score 12269 3DMarks
SM 2.0 Score 5524 Marks
SM 3.0 Score 5342 Marks
CPU Score 3170 Marks
EDIT:
My above test was with the Nvidia 163.71 drivers. I just updated to 163.75 and will redo the test.
Mrfatboy's
3DMark Score 12231 3DMarks
SM 2.0 Score 5486 Marks
SM 3.0 Score 5270 Marks
CPU Score 3236 Marks
Guacho's
3DMark Score 12269 3DMarks
SM 2.0 Score 5524 Marks
SM 3.0 Score 5342 Marks
CPU Score 3170 Marks
EDIT:
My above test was with the Nvidia 163.71 drivers. I just updated to 163.75 and will redo the test.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter