<div class="FTQUOTE"><begin quote>Originally posted by: Shampoo
My guess is that the Asus techs probably relaxed the memory subtimings, since this board has been reported to use very tight memory subtimings by various review sites.
If that's the case, then things should be good with 0608 since memory subtimings don't have that much of an effect on performance, but more on stability in my experiences.
__________________________________________________________________________
You know, you'd be surprised about the subtimings. With my 800Mhz, 4-4-4-12-2, Everest reads:
Enhanced Performance Profile:
Profile Name High Performance
Optimal Performance Profile Yes
Memory Speed DDR2-800 (400 MHz)
Voltage 2.1 V
Memory Timings 4-4-4-12 (CL-RCD-RP-RAS)
Row Cycle Time (tRC) 24T
Command Rate (CR) 2T
Write Recovery Time (tWR) 6T
Memory Module Manufacturer:
Company Name GSkill International Enterprise
Product Information http://www.gskill.com/indexen.html
Thats the EPP. Now the current subtimings set to auto, as reported by Everest:
Memory Timings:..........................CURRENT....G.Skill Recommended
CAS Latency (CL)............................4T.................................4T
RAS To CAS Delay (tRCD)................4T................................4T
RAS Precharge (tRP)........................4T................................4T
RAS Active Time (tRAS)...................12T..............................12T
Row Cycle Time (tRC)......................18T...............................2T (from 18! EPP=24)
Row Refresh Cycle Time (tRFC)........40T..............................27T
Command Rate (CR)........................2T................................2T
RAS To RAS Delay (tRRD).................3T...............................1T
Write Recovery Time (tWR)...............5T................................2T
Read To Write Delay (tRTW).............5T...............................6T
Write To Read Delay (tWTR).............8T...............................6T
Needless to say, with subtimings tightened as recommended by G.Skill, my SANDRA would go beyond 6600MB/s from current ~6500MB/s, but my computer would... hardlock, or refuse to boot (on reboot - had to hit the reset button, go back into BIOS, and return the subtimings to all AUTO). Rock solid since, with 4 sticks.
I thing the board can determine the subtimings much better on its own, and you should "tinker" with them only if you have problems.
My guess is that the Asus techs probably relaxed the memory subtimings, since this board has been reported to use very tight memory subtimings by various review sites.
If that's the case, then things should be good with 0608 since memory subtimings don't have that much of an effect on performance, but more on stability in my experiences.
__________________________________________________________________________
You know, you'd be surprised about the subtimings. With my 800Mhz, 4-4-4-12-2, Everest reads:
Enhanced Performance Profile:
Profile Name High Performance
Optimal Performance Profile Yes
Memory Speed DDR2-800 (400 MHz)
Voltage 2.1 V
Memory Timings 4-4-4-12 (CL-RCD-RP-RAS)
Row Cycle Time (tRC) 24T
Command Rate (CR) 2T
Write Recovery Time (tWR) 6T
Memory Module Manufacturer:
Company Name GSkill International Enterprise
Product Information http://www.gskill.com/indexen.html
Thats the EPP. Now the current subtimings set to auto, as reported by Everest:
Memory Timings:..........................CURRENT....G.Skill Recommended
CAS Latency (CL)............................4T.................................4T
RAS To CAS Delay (tRCD)................4T................................4T
RAS Precharge (tRP)........................4T................................4T
RAS Active Time (tRAS)...................12T..............................12T
Row Cycle Time (tRC)......................18T...............................2T (from 18! EPP=24)
Row Refresh Cycle Time (tRFC)........40T..............................27T
Command Rate (CR)........................2T................................2T
RAS To RAS Delay (tRRD).................3T...............................1T
Write Recovery Time (tWR)...............5T................................2T
Read To Write Delay (tRTW).............5T...............................6T
Write To Read Delay (tWTR).............8T...............................6T
Needless to say, with subtimings tightened as recommended by G.Skill, my SANDRA would go beyond 6600MB/s from current ~6500MB/s, but my computer would... hardlock, or refuse to boot (on reboot - had to hit the reset button, go back into BIOS, and return the subtimings to all AUTO). Rock solid since, with 4 sticks.
I thing the board can determine the subtimings much better on its own, and you should "tinker" with them only if you have problems.
I'm officially stable. Ran memtest with my 4-4-4-12, 2T timings at 875MHz without errors.
Ran a Blend stress test for 24 hours and a bit error free.
I'm liking my temps too, but I may reseat my heatsink once again. ugh.
Sandra gave me the following: Int=7667, Float=7677
I am happy.
The ram timings at that speed give me greater bandwidth than my previous unstable ram speeds of over 1000MHz.
Now it's time to start playing my games again instead of worrying about stability.
Cheers,
Mike
Ran a Blend stress test for 24 hours and a bit error free.
I'm liking my temps too, but I may reseat my heatsink once again. ugh.
Sandra gave me the following: Int=7667, Float=7677
I am happy.
The ram timings at that speed give me greater bandwidth than my previous unstable ram speeds of over 1000MHz.
Now it's time to start playing my games again instead of worrying about stability.
Cheers,
Mike
Hello,
I followed the guide from Core 2 and succefully overclocked my system to 3.1 ghz on a stock fan and 1:1 ratio (lowered the ram to 700 for 1:1). This was my first time in overclocking. Thanks Core 2 for the awesome guide!
Now my question is, how do I make my system into a single core? I've read people have better experiences in gamming on a single core. I have no idea in how do that and appreciated if you guys can help me out, thanks.
E6600
P5N-E SLI
I followed the guide from Core 2 and succefully overclocked my system to 3.1 ghz on a stock fan and 1:1 ratio (lowered the ram to 700 for 1:1). This was my first time in overclocking. Thanks Core 2 for the awesome guide!
Now my question is, how do I make my system into a single core? I've read people have better experiences in gamming on a single core. I have no idea in how do that and appreciated if you guys can help me out, thanks.
E6600
P5N-E SLI
If you are running XP, use the following patch from Microsoft:
http://support.microsoft.com/kb/896256
Off topic, but I was just wondering about the dual-core patch.
From what I've read it seems to be a patch for AMD X2 processors.
I haven't read anywhere else where an Intel Core2Duo user has had problems like those of the X2.
Other forums point to a discussion on the forums here at anandtech, but it seems to have disappeared.
Any comments?
From what I've read it seems to be a patch for AMD X2 processors.
I haven't read anywhere else where an Intel Core2Duo user has had problems like those of the X2.
Other forums point to a discussion on the forums here at anandtech, but it seems to have disappeared.
Any comments?
It is for all x86 platform processors, and it is even recommended for the Pentium 4 with HT.
The way I understand it, the workload is being processed by 1 core, and then the XP "realizes" that the other core has more free resources, and jumps to the other core, switching uncontrollably between the cores, or threads in CPU's with HT.
It seems to have helped me, making the games playing more smoothly.
But... read this:
http://www.overclock.net/windo...h-dual-core-procs.html
It seems like I did not enable it in the registry... Let me see what that does...
I just remembered some patches for XP to fix hibernation - read about these, pehaps Vista has similar problems:
http://support.microsoft.com/kb/306676
http://support.microsoft.com/?kbid=909095
Hope this helps.
Well, tried disabling the patch (Registry entry 0) and enabling it (DWord Value 1), and I really don't see much of a difference. I does seem little smoother with the patch enabled, so I left it on.
I believe that the processing power of Core 2 Duo processors is so great that even switching cores don't affect performance too much. I is probably more prominent on slower Athlons, like 3800+ X2, where every ounce of performance counts...
I believe that the processing power of Core 2 Duo processors is so great that even switching cores don't affect performance too much. I is probably more prominent on slower Athlons, like 3800+ X2, where every ounce of performance counts...
I actually noticed a difference as well.
Not a positive one as of yet, but I will see how my system behaves with this patch over the next couple of days to get a subjective view on things.
So far just messing with the task manager's Performance tab I see that one core is being use more than the other when grabbing an open window and dragging it like nuts all across the desktop.
This is with the patch enabled.
Before I installed or edited anything I did the same thing and the cpu usage graphs looked a lot more in sync.
We will see as only time will tell.
I'm going to reinstall windows soon as well, since I have it stable now, I'd like to install windows with a stable set of ram.
This windows installation is of the unstable ram kind. This install is from when I had my ram at 1020MHz thinking it was stable.
We'll see how things unravel then.
Edit:: Coming to you from my notebook (Asus Z70VA) while testing my desktop.
Not a positive one as of yet, but I will see how my system behaves with this patch over the next couple of days to get a subjective view on things.
So far just messing with the task manager's Performance tab I see that one core is being use more than the other when grabbing an open window and dragging it like nuts all across the desktop.
This is with the patch enabled.
Before I installed or edited anything I did the same thing and the cpu usage graphs looked a lot more in sync.
We will see as only time will tell.
I'm going to reinstall windows soon as well, since I have it stable now, I'd like to install windows with a stable set of ram.
This windows installation is of the unstable ram kind. This install is from when I had my ram at 1020MHz thinking it was stable.
We'll see how things unravel then.
Edit:: Coming to you from my notebook (Asus Z70VA) while testing my desktop.
What I've noticed so far.
When the patch is disabled (0), the two cores, as described in other forums, seem to be erratic and all over the place in terms of sharing the load together.
When the patch is enabled (1), the two cores seem to work together, although not always at the same load level, but they are better sync'ed.
This is good to see. I think I'll keep the patch.
Edit:: In no way was my testing scientific...just eye balling graphs in the Performance tab of the Task Manager after running Aquamark 3 and moving some windows around.
When the patch is disabled (0), the two cores, as described in other forums, seem to be erratic and all over the place in terms of sharing the load together.
When the patch is enabled (1), the two cores seem to work together, although not always at the same load level, but they are better sync'ed.
This is good to see. I think I'll keep the patch.
Edit:: In no way was my testing scientific...just eye balling graphs in the Performance tab of the Task Manager after running Aquamark 3 and moving some windows around.
Shampoo,
If you feel that some of the system files might have gotten corrupted, why don't you just do the "Repair Install"...?
You will have to update Windows with probably a gazillion patches, but at least you will get to keep all your applications.
Just a thought...
If you feel that some of the system files might have gotten corrupted, why don't you just do the "Repair Install"...?
You will have to update Windows with probably a gazillion patches, but at least you will get to keep all your applications.
Just a thought...
Well, I finally gave up. Trusting the experiences of Tuvoc and Shampoo, and feeling a little obsolete with 0202, I performed the dreaded BIOS update to 0608.
And I figured that 6 months would be enough for ASUS programmers to get things right for all the new software and hardware configurations.
I used the good old and trusted... ASUS update in Windows, as I have successfully done many times for my 775Dual-VSTA and V5305WM HP Compaq laptop (necessary before "upgrading" from XP Media Edition to Vista Home Premium). I just would like to stress again the need to terminate all the User tasks, especially Anti Virus and any type of System Settings protection (WinPatrol etc.), leaving only explorer.exe and taskmgr.exe. Much easier than any other method, IMHO, or maybe I am just... lazy LOL.
There is a place in the ASUS update program where you check (I think it is checked by default) to "Load Setup Defaults after update", but after the Reboot, I entered the BIOS, went straight to the end, loaded the defaults, "Save and Exit", and only then I re-entered the BIOS to change the settings Manually.
Now, my impressions. The C2D Microcode Update Revision has been changed from 44 to C6, as ASrock did in April in response to Intel bug reported earlier. Perhaps that microcode bug prevents some people from resuming from hibernation.
The temperatures (Everest 4.0) have changed dramatically, approximately 10C more than before! My max temps under Orthos load are now CPU ~55C, Core 1/2 70C. Well, the newer E6600 that I purchased after the price drop in April and installed on 775Dual-VSTA idles at ~35C, but Core 1/2 show ~30C. Intel must have changed the temp. sensor - I don't believe that the Core temps are lower that Tjunction (CPU temp). It's gotta be hotter inside the chip, right...? That's why I am not really concerned about the higher temps - I know that the chip is not running warmer than before, what's changed is the readings.
Orthos is running now as we speak, so far no errors. I don't usually run the tests longer than 1 hour. If I don't get errors/lockups/hangs in such intense games as Quake 4 or S.T.A.L.K.E.R., it means that the machine is fine IMHO.
Hope it all continues though, as I have enjoyed the computer's perfect stability for 6 month with 0202.
And I figured that 6 months would be enough for ASUS programmers to get things right for all the new software and hardware configurations.
I used the good old and trusted... ASUS update in Windows, as I have successfully done many times for my 775Dual-VSTA and V5305WM HP Compaq laptop (necessary before "upgrading" from XP Media Edition to Vista Home Premium). I just would like to stress again the need to terminate all the User tasks, especially Anti Virus and any type of System Settings protection (WinPatrol etc.), leaving only explorer.exe and taskmgr.exe. Much easier than any other method, IMHO, or maybe I am just... lazy LOL.
There is a place in the ASUS update program where you check (I think it is checked by default) to "Load Setup Defaults after update", but after the Reboot, I entered the BIOS, went straight to the end, loaded the defaults, "Save and Exit", and only then I re-entered the BIOS to change the settings Manually.
Now, my impressions. The C2D Microcode Update Revision has been changed from 44 to C6, as ASrock did in April in response to Intel bug reported earlier. Perhaps that microcode bug prevents some people from resuming from hibernation.
The temperatures (Everest 4.0) have changed dramatically, approximately 10C more than before! My max temps under Orthos load are now CPU ~55C, Core 1/2 70C. Well, the newer E6600 that I purchased after the price drop in April and installed on 775Dual-VSTA idles at ~35C, but Core 1/2 show ~30C. Intel must have changed the temp. sensor - I don't believe that the Core temps are lower that Tjunction (CPU temp). It's gotta be hotter inside the chip, right...? That's why I am not really concerned about the higher temps - I know that the chip is not running warmer than before, what's changed is the readings.
Orthos is running now as we speak, so far no errors. I don't usually run the tests longer than 1 hour. If I don't get errors/lockups/hangs in such intense games as Quake 4 or S.T.A.L.K.E.R., it means that the machine is fine IMHO.
Hope it all continues though, as I have enjoyed the computer's perfect stability for 6 month with 0202.
Glad the update went well
Temperature calibration is often changed in BIOS updates, it has been like that for years with many different boards. But I do find it odd that something so simple can't be gotten right first time... Anyway, yes of course your chip should be running the same temps as before despite the readings. Actually 70C coretemp at 2.9Ghz with the stock heatsink sounds much more accurate
Temperature calibration is often changed in BIOS updates, it has been like that for years with many different boards. But I do find it odd that something so simple can't be gotten right first time... Anyway, yes of course your chip should be running the same temps as before despite the readings. Actually 70C coretemp at 2.9Ghz with the stock heatsink sounds much more accurate
Thanks, Tuvoc. You're right, it is probably more accurate now, but since I have not had problems before, it's most likely OK.
Now... the new BIOS. So far, it's exceeded my expectations! I kinda had an impression that it is time to update after 6 months and 5 or 6 BIOS versions, but I think that ASUS finally got things right this time.
The NHL07, the game that would not put too much load on the GPU, and is very much CPU dependent, plays much smoother (well, the Leafs lost 1:5 to the Devils, but that's another story...).
Memory bandwidth is higher, with all the settings same as before, including the Memory subtimings on Auto. Everest reads all the values same as I have mentioned on the previous page. Other SANDRA results also seem same as before.
Now... PCMark05 score is 7924, higher than the highest of 7830 before with 3GB of RAM.
3DMark06 score of 5936 is also highest than ever!
Looks to me that the codes for all the componentry have been changed for better, too!
I felt so ecstatic (LOL) that I decided to change my avatar - looks more like me anyway...
Now... the new BIOS. So far, it's exceeded my expectations! I kinda had an impression that it is time to update after 6 months and 5 or 6 BIOS versions, but I think that ASUS finally got things right this time.
The NHL07, the game that would not put too much load on the GPU, and is very much CPU dependent, plays much smoother (well, the Leafs lost 1:5 to the Devils, but that's another story...).
Memory bandwidth is higher, with all the settings same as before, including the Memory subtimings on Auto. Everest reads all the values same as I have mentioned on the previous page. Other SANDRA results also seem same as before.
Now... PCMark05 score is 7924, higher than the highest of 7830 before with 3GB of RAM.
3DMark06 score of 5936 is also highest than ever!
Looks to me that the codes for all the componentry have been changed for better, too!
I felt so ecstatic (LOL) that I decided to change my avatar - looks more like me anyway...
Great news, well worth the wait then. I didn't check the memory bandwidth myself. must get around to doing that
Watch the CPU Voltage! I just experienced a failed boot, inability to enter BIOS (it said incompatibile VGA adaptor or something to this effect).
Checked Everest reports from before and after the update, and before the CPU would report the required voltage as 1.3V, after the update 1.4V.
So, on Auto settings, the CPU was "overvolted", hence the increased temps!
Lowered the voltage to 1.325 as specified for the C2D, still boot failure!
Increased the NB voltage to 1.58V - boot failure. Changed the NB to 1.38V - seems OK now.
Well, so much for early ecstasy LOL. Will keep you posted...
Check your CPU Core Voltage!
P.S. Orthos load temps went down, too, to about 47C CPU, 59C Core 1/2, in line with the temps before the update.
Checked Everest reports from before and after the update, and before the CPU would report the required voltage as 1.3V, after the update 1.4V.
So, on Auto settings, the CPU was "overvolted", hence the increased temps!
Lowered the voltage to 1.325 as specified for the C2D, still boot failure!
Increased the NB voltage to 1.58V - boot failure. Changed the NB to 1.38V - seems OK now.
Well, so much for early ecstasy LOL. Will keep you posted...
Check your CPU Core Voltage!
P.S. Orthos load temps went down, too, to about 47C CPU, 59C Core 1/2, in line with the temps before the update.
Maybe a CMOS clear might be a good idea after all ?
My voltages are normal. Normal for the quad core is that this board horribly *UNDERVOLTS* it on auto.
QX6700 default voltage 1.35v. At idle Everest reports 1.28v - 1.30v, immediately an undervolt of .05v, but under load that drops to 1.20 - 1.23 v !! VDroop on this board is horrendous with the quad core CPU. Disgraceful I would describe it... It passes 4 instances of Prime95 and I guess the undervolt keeps temps down a little, but if I wanted to overclock I'd almost certainly have to manually set it higher
My voltages are normal. Normal for the quad core is that this board horribly *UNDERVOLTS* it on auto.
QX6700 default voltage 1.35v. At idle Everest reports 1.28v - 1.30v, immediately an undervolt of .05v, but under load that drops to 1.20 - 1.23 v !! VDroop on this board is horrendous with the quad core CPU. Disgraceful I would describe it... It passes 4 instances of Prime95 and I guess the undervolt keeps temps down a little, but if I wanted to overclock I'd almost certainly have to manually set it higher
Will keep you posted... Now I gotta go and see Kid Rock live!
It's gonna be a great show, with the 4th of July fireworks in the end.
Maybe I should downgrade to 0202...?
Nah... gotta be progressive LOL.
P.S. Rebooted a few more times, played another game (Leafs 5 Capitals 3) - all seems fine.
Performance appears to be increased, opening windows and programs seems smoother. Well, can't just go back to 0202 without a fight, right...? Especially that the new BIOS' provide support for new hardware, and hopefully eliminate earlier mistakes in programming.
What a temperamental motherboard, huh...?
It's gonna be a great show, with the 4th of July fireworks in the end.
Maybe I should downgrade to 0202...?
Nah... gotta be progressive LOL.
P.S. Rebooted a few more times, played another game (Leafs 5 Capitals 3) - all seems fine.
Performance appears to be increased, opening windows and programs seems smoother. Well, can't just go back to 0202 without a fight, right...? Especially that the new BIOS' provide support for new hardware, and hopefully eliminate earlier mistakes in programming.
What a temperamental motherboard, huh...?
Kid Rock rules!
Oops, sorry - wrong forum!
OK, I have a little theory why our motherboards fail to boot, especially after the BIOS upgrades.
I think that Intel is attempting to protect the increased failure claims for the C2D processors. If the chip is not cooled properly, it will heat up very quickly during boot when it needs "full power" trying to identify components, and the difference in ambient and the Core 1/2 temperatures very quickly rises beyond their specifications.
I think that if the temp difference rises beyond ~15C, which is very easy on boot when all the components, including the CPU heatsink/TJunction are cold (ex. 27C ambient/components/TJunction, and >42C Core 1/2), the processor shuts down. I believe that they got kinda suprised by the "overclockability" of the Conroe design, and people taking a 1.86GHz chip easily to 3.5GHz and beyond, that they lowered the threshhold!
My E6600 would probably be OK at 2.4GHz, but overclocked must have been heating up pretty quickly. That's why it is harder to take the the CPU's with the Microcode C6 BIOS' to those higher overclock values, especially with the stock HSF.
That does not explain the higher default VCore voltage though... Well, these processors are relatively new, and mine was purchased in September 2006. There must have been some design/microcode flaws so early, thats why the newer chips (like the one that I got in April 2007) show the Core 1/2 temps lower than TJunction (CPU temp).
Two same chips, on the same motherboard, showing the Core 1/2 values ~15C apart! I remeber that before I got the P5N-E SLI, the temps on 775Dual-VSTA were ~35C CPU - 42C Core 1/2. Now, with the same type E6600 CPU, it shows ~32C CPU, ~28C Core 1/2. I doubt that the stock HSF's all of a sudden have gotten more efficient!
Anyway, all seems fine after I adjusted the CPU and the NB voltages. I know that my frequencies/timings are fine - the MB was perfectly stable before, so the problem must lay in the voltage(s).
Can you believe that the Conroe platform is already 1 year old and it still rocks...? Simply incredible...
UPDATE: Failed boot today, Thursday afternoon, "cold" computer in Air Conditioned environment.
Disabled CPU Internal Thermal Control, boot successfull - will see what happens...
Do you set it on "On" or "Off"...?
Oops, sorry - wrong forum!
OK, I have a little theory why our motherboards fail to boot, especially after the BIOS upgrades.
I think that Intel is attempting to protect the increased failure claims for the C2D processors. If the chip is not cooled properly, it will heat up very quickly during boot when it needs "full power" trying to identify components, and the difference in ambient and the Core 1/2 temperatures very quickly rises beyond their specifications.
I think that if the temp difference rises beyond ~15C, which is very easy on boot when all the components, including the CPU heatsink/TJunction are cold (ex. 27C ambient/components/TJunction, and >42C Core 1/2), the processor shuts down. I believe that they got kinda suprised by the "overclockability" of the Conroe design, and people taking a 1.86GHz chip easily to 3.5GHz and beyond, that they lowered the threshhold!
My E6600 would probably be OK at 2.4GHz, but overclocked must have been heating up pretty quickly. That's why it is harder to take the the CPU's with the Microcode C6 BIOS' to those higher overclock values, especially with the stock HSF.
That does not explain the higher default VCore voltage though... Well, these processors are relatively new, and mine was purchased in September 2006. There must have been some design/microcode flaws so early, thats why the newer chips (like the one that I got in April 2007) show the Core 1/2 temps lower than TJunction (CPU temp).
Two same chips, on the same motherboard, showing the Core 1/2 values ~15C apart! I remeber that before I got the P5N-E SLI, the temps on 775Dual-VSTA were ~35C CPU - 42C Core 1/2. Now, with the same type E6600 CPU, it shows ~32C CPU, ~28C Core 1/2. I doubt that the stock HSF's all of a sudden have gotten more efficient!
Anyway, all seems fine after I adjusted the CPU and the NB voltages. I know that my frequencies/timings are fine - the MB was perfectly stable before, so the problem must lay in the voltage(s).
Can you believe that the Conroe platform is already 1 year old and it still rocks...? Simply incredible...
UPDATE: Failed boot today, Thursday afternoon, "cold" computer in Air Conditioned environment.
Disabled CPU Internal Thermal Control, boot successfull - will see what happens...
Do you set it on "On" or "Off"...?
Finally 'Sleep' works for my Vista machine. My 8500GT came this morning and I replaced 7100GS with it. Nothing else has changed and the 'Sleep' is working now. I am using 158.45 Beta driver from Nvidia instead of using 158.24 for 7100GS. So this is either the VGA board or the driver.
Mine is set to enabled, as a safety thing - i.e. this turns throttling on to safeguard the CPU just in case it gets too hot
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: 24K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter