bumpity bump, what stress test do I run to see if I can change the tRD from 13 to 10 and survive it? Memtest?
- Jan 4, 2004
- 21,303
- 4
- 81
3.95 GHz 8+ hours Blend stable
3.95 GHz 5 passes IBT stable
I gave up on 4 GHz, as it requires more vcore than 1.4v, which i am not willing to do.
Still working on lowering voltages, etc, & will work on RAM more later, but here are my present settings anyway.
AI Overclock tuner: Manual
CPU Ratio Setting: 8.5
FSB Frequency: 465
PCI-E Frequency: 101
FSB Strap to North Bridge: 465
DRAM Frequency: 930
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: Manual
1st Information:
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time: 60
Write Recovery Time: 6
Read to Precharge Time: 3
2nd Information :
READ to WRITE Delay (S/D): 9
Write to Read Delay (S): 4
WRITE to READ Delay (D): 6
READ to READ Delay (S): 5
READ to READ Delay (D): 7
WRITE to WRITE Delay (S): 5
WRITE to WRITE Delay (D): 7
3rd Information :
WRITE to PRE Delay: 15
READ to PRE Delay: 6
PRE to PRE Delay: 2
ALL PRE to ACT Delay: 7
ALL PRE to REF Delay: 7
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
AI Clock Twister: Light
AI Transaction Booster: Manual
Common Performance Level: 10
All Pull-Ins: Disabled
CPU Voltage: 1.4
CPU GTL Voltage Reference (0/2): 0.680
CPU GTL Voltage Reference (1/3): 0.670
CPU PLL Voltage: 1.58
FSB Termination Voltage: 1.36
DRAM Voltage: 1.96
NB Voltage: 1.38
NB GTL Reference: 0.67
SBridge Voltage: 1.2
PCIE SATA Voltage: 1.6
Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : 100
NB Clock Skew : 200
CPU Margin Enhancement: Optimized
3.95 GHz 5 passes IBT stable
I gave up on 4 GHz, as it requires more vcore than 1.4v, which i am not willing to do.
Still working on lowering voltages, etc, & will work on RAM more later, but here are my present settings anyway.
AI Overclock tuner: Manual
CPU Ratio Setting: 8.5
FSB Frequency: 465
PCI-E Frequency: 101
FSB Strap to North Bridge: 465
DRAM Frequency: 930
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: Manual
1st Information:
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time: 60
Write Recovery Time: 6
Read to Precharge Time: 3
2nd Information :
READ to WRITE Delay (S/D): 9
Write to Read Delay (S): 4
WRITE to READ Delay (D): 6
READ to READ Delay (S): 5
READ to READ Delay (D): 7
WRITE to WRITE Delay (S): 5
WRITE to WRITE Delay (D): 7
3rd Information :
WRITE to PRE Delay: 15
READ to PRE Delay: 6
PRE to PRE Delay: 2
ALL PRE to ACT Delay: 7
ALL PRE to REF Delay: 7
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
AI Clock Twister: Light
AI Transaction Booster: Manual
Common Performance Level: 10
All Pull-Ins: Disabled
CPU Voltage: 1.4
CPU GTL Voltage Reference (0/2): 0.680
CPU GTL Voltage Reference (1/3): 0.670
CPU PLL Voltage: 1.58
FSB Termination Voltage: 1.36
DRAM Voltage: 1.96
NB Voltage: 1.38
NB GTL Reference: 0.67
SBridge Voltage: 1.2
PCIE SATA Voltage: 1.6
Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : 100
NB Clock Skew : 200
CPU Margin Enhancement: Optimized
I just set mine to 400mhz fsb without any other changes and it is stable. I am happy with relatively cool, low(ish) power consumption and 3.4Ghz.
I tried your OC n7 and crashed 15 min into it. LoL, these damn chips are like dating sisters, just because there manufactured at the same place doesn't mean you get the same thing.
- Jan 4, 2004
- 21,303
- 4
- 81
I've looked on various forums & tried the settings a bunch of different people have used & none of theirs have worked for me, so that doesn't surprise me.
Big Lar, i used your settings as an intial stepping stone, but i tried pretty much every GTL combination along with your voltages & then higher.
Basically i tried 0.61 up to 0.70 for the various voltage steps for vtt/nb, & it was very sensitive.
The 0.68 - 0.67 VTT & 0.67 for NB are set there because anything lower wouldn't pass.
Not sure why mine needs higher values...most people w/ the 45nm quads are able to use lower reference values, but not for mine.
I am working on lowering VTT/NB voltage now, & i'll trying lowering the GTL values later also, but i'm not sure how successful that will be.
Ok this doesn't make any sense. I decided I was just pushing it to far and backed it up a bit to a 3.8 or so OC. Stable. Got tRD set at 10. For shits and grins I upped the FSB to 475 with vcore at 1.4 and it went in.. it never went in with higher voltages or with this lower tRD. hoping for the best, question: when I run memtest how many copies should I run? Everest is showing me 2896MB of free memory atm so I was going to divide that by 4 and run 4 copies of it, does that work? Or should I divide by 2 and run 2, or just run one, and test the full 2896? gonna run IBT in the meantime to see if it will actually stay stable. Also I haven't touched GTL at all. I keep reading about how theres not enough accurate information about it so haven't bothered.
Try 4 instances of memtest with 500mb per instance, seeme to work well with a quad. Probably best to let each run to 300/400% to test well. GTL's, are just a reference for voltage. Sorta like a switch, line in/line out....however they key is to get the switch opening and closing at the Exact right times, that is my rendition of what gtl's are.
Larry
Getting close to being 24 hour stable with prime!!! Will run memtest overnight next if all is well and post back screens and results if all goes well. Stay tuned..
Oh, before I get to the memtest part, you said run 4 copes at 500mb per instance, from my understanding of what I have read Vista uses x amount of ram at any given time but the ram it is using is constantly changing, using different parts in other words, so is this why you suggest 500mb per instance and eventually if allowed to run long enough, such as overnight, it will cycle through all of it reguardless of the 500x4 not being what it lists as available, other threads I have read suggest shutting down everything that doesn't need to be running during the test, then opening whatever monitoring programs you plan on using while it is running, finding what is available and dividing this number by four to determine how much each test will run. If I understand how it works, then either method should do the same thing right? What is the benefit or con of only doing 500 as opposed to say 755 each test? Not trying to say one way or the other is right, but as the 2 old sayings go "Knowledge is power" and "Now you know, and knowing is half the battle"
Oh, before I get to the memtest part, you said run 4 copes at 500mb per instance, from my understanding of what I have read Vista uses x amount of ram at any given time but the ram it is using is constantly changing, using different parts in other words, so is this why you suggest 500mb per instance and eventually if allowed to run long enough, such as overnight, it will cycle through all of it reguardless of the 500x4 not being what it lists as available, other threads I have read suggest shutting down everything that doesn't need to be running during the test, then opening whatever monitoring programs you plan on using while it is running, finding what is available and dividing this number by four to determine how much each test will run. If I understand how it works, then either method should do the same thing right? What is the benefit or con of only doing 500 as opposed to say 755 each test? Not trying to say one way or the other is right, but as the 2 old sayings go "Knowledge is power" and "Now you know, and knowing is half the battle"
evolucion8
Platinum Member
- Jun 17, 2005
- 2,867
- 3
- 81
That's so true. I was dissapointed when I saw people overclocking the same CPU as mine to 3.0GHz and I barely got 2.70GHz, then I bought the same identical CPU, and it couldn't reach 2.65GHz, only 2.50GHz stable, loll
Originally posted by: jjrowley
Getting close to being 24 hour stable with prime!!! Will run memtest overnight next if all is well and post back screens and results if all goes well. Stay tuned..
Oh, before I get to the memtest part, you said run 4 copes at 500mb per instance, from my understanding of what I have read Vista uses x amount of ram at any given time but the ram it is using is constantly changing, using different parts in other words, so is this why you suggest 500mb per instance and eventually if allowed to run long enough, such as overnight, it will cycle through all of it reguardless of the 500x4 not being what it lists as available, other threads I have read suggest shutting down everything that doesn't need to be running during the test, then opening whatever monitoring programs you plan on using while it is running, finding what is available and dividing this number by four to determine how much each test will run. If I understand how it works, then either method should do the same thing right? What is the benefit or con of only doing 500 as opposed to say 755 each test? Not trying to say one way or the other is right, but as the 2 old sayings go "Knowledge is power" and "Now you know, and knowing is half the battle"
500mb is just what works well for me, heck I am running XP still tho with 4gb of ram
I would say that whatever will run with a bit of cushion should work fine, no realneed to run it overnite as if there are any errors, Memtest will find them usually by 300%.Larry
well I don't know what happened with the prime 95 not staying running, but i suspect I didn't have it set right, because after 14 hours it shut itself down, log says everything passed, and I noticed when I open it up I have it set to run 4 torture test threads, so I assume it ran these four and stopped? blah, i'll run it another 24 hours, but for now heres where I'm at.
CPU-Z Validation page
http://valid.x86-secret.com/show_oc.php?id=442450
CPU-Z actually showing v-core
http://i83.photobucket.com/alb...esserowley2000/014.jpg
memtest
http://i83.photobucket.com/alb...esserowley2000/013.jpg
the beast
http://i83.photobucket.com/alb...mputer%20Build/008.jpg
CPU-Z Validation page
http://valid.x86-secret.com/show_oc.php?id=442450
CPU-Z actually showing v-core
http://i83.photobucket.com/alb...esserowley2000/014.jpg
memtest
http://i83.photobucket.com/alb...esserowley2000/013.jpg
the beast
http://i83.photobucket.com/alb...mputer%20Build/008.jpg
yeah I had to turn it on with the tablet tool pack, this thing is great!
anyways, I'm still puzzled as to why Prime stopped, I didn't get any error messages, when I came home my computer was off so I booted it up, didn't get any of those windows crashed messages, checked prime log and looks like it all passed, does vista have some auto feature that might be shutting down my computer on me when idle for so long? here is what the prime log shows me anyways:
(edited out the results log, was tired of scrolling)
anyways, I'm still puzzled as to why Prime stopped, I didn't get any error messages, when I came home my computer was off so I booted it up, didn't get any of those windows crashed messages, checked prime log and looks like it all passed, does vista have some auto feature that might be shutting down my computer on me when idle for so long? here is what the prime log shows me anyways:
(edited out the results log, was tired of scrolling)
I don't have an Intel chip, but mine would shut off from overheating. What were the temps when it shut off?
high 40's low 50's. I saved that OC profile in my bios and started a new one actually. I kept reading all kinds of stuff about the 45nm chips and voltage, so I'm going to see how I can do at 1.3625 the recommended safe cap on this thing. Right now at 3.867 with 455 FSB. gonna run some prelims and make sure its stable then I'll run an overnighter when I find the cap at this voltage. will post back with results.
It definitely doesn't sound like heat. The other thing I would guess is PSU, but you have a pretty beefy PSU. Although I will say one thing about the TX750W PSU from Corsair - it has some horrid voltage ripple at load. I would have never expected that from Corsair, but I guess these things happen sometimes.
Did your second run go better than the first? I plan on buying this processor within the next couple of weeks, so this interests me quite a bit too.
@1.3625 you can hit 3.825 no problem. Just going to take a bit of tweaking to find the boundaries. I'm going to save this OC in my profile on my asus board along with my 4.003 stable OC and switch back and forth as I feel froggy.
I don't have anything to compare it too, but I would recomend this board and chip to anyone looking for something moderate to play with. It has been fairly forgiving for my first time, and not too overwhelming.
edit: dude 25mv isn't going to effect anything! the pictures make it look significant, but its really not.
I don't have anything to compare it too, but I would recomend this board and chip to anyone looking for something moderate to play with. It has been fairly forgiving for my first time, and not too overwhelming.
edit: dude 25mv isn't going to effect anything! the pictures make it look significant, but its really not.
I've been messing around a bit with the cpu margin setting thingy and have found that I am doing a bit better with Compatible versus Optimal of which I was running. Not much diff yet but I picked up a coupla fsb without raising the cpu voltage so far. More testing in the works.
Larry
Larry
OK, So here's how I mess with the NB GTL
So you have; 1.34vtt and a cpu gtl of 0.680, and are running fairly well at 1.36NB. My formula if you will below pertaining to Your settings.
1.34 x .680 = 0.9112 divided by 1.36 = 0.67 So to break it down,=
vtt x cpu/gtl = TREF divided by NB Voltage = NB GTL So with my theory your NB GTL should be .670
It seems to work out pretty close, at least for my chip. Hope that helps.
Larry
AI Overclock tuner: Manual
CPU Ratio Setting: 8.5
FSB Frequency: 471
PCI-E Frequency: 100
FSB Strap to North Bridge: 400
DRAM Frequency: DDR2-942MHz
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: Manual
1st Information:
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time: 60
Write Recovery Time: 6
Read to Precharge Time: 3
2nd Information :
READ to WRITE Delay (S/D): 9
Write to Read Delay (S): 4
WRITE to READ Delay (D): 6
READ to READ Delay (S): 5
READ to READ Delay (D): 7
WRITE to WRITE Delay (S): 5
WRITE to WRITE Delay (D): 7
3rd Information :
WRITE to PRE Delay: 15
READ to PRE Delay: 6
PRE to PRE Delay: 2
ALL PRE to ACT Delay: 7
ALL PRE to REF Delay: 7
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
AI Clock Twister: Moderate
AI Transaction Booster: Manual
Common Performance Level: 10
All Pull-Ins: Disabled
CPU Voltage: 1.41250
CPU GTL Voltage Reference (0/2): 0.675
CPU GTL Voltage Reference (1/3): 0.675
CPU PLL Voltage: 1.60
FSB Termination Voltage: 1.36
DRAM Voltage: 2.20
NB Voltage: 1.36
NB GTL Reference: 0.675
SBridge Voltage: 1.2
PCIE SATA Voltage: Auto
Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Auto
CPU Clock Skew : 100
NB Clock Skew : 200
CPU Margin Enhancement: Compatible
I can't stay away from the higher numbers, but I worry that my numbers are getting a bit high. What do you guys think?
CPU Ratio Setting: 8.5
FSB Frequency: 471
PCI-E Frequency: 100
FSB Strap to North Bridge: 400
DRAM Frequency: DDR2-942MHz
DRAM CLK Skew on Channel A1: AUTO
DRAM CLK Skew on Channel A2: AUTO
DRAM CLK Skew on Channel B1: AUTO
DRAM CLK Skew on Channel B2: AUTO
DRAM Timing Control: Manual
1st Information:
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time: 60
Write Recovery Time: 6
Read to Precharge Time: 3
2nd Information :
READ to WRITE Delay (S/D): 9
Write to Read Delay (S): 4
WRITE to READ Delay (D): 6
READ to READ Delay (S): 5
READ to READ Delay (D): 7
WRITE to WRITE Delay (S): 5
WRITE to WRITE Delay (D): 7
3rd Information :
WRITE to PRE Delay: 15
READ to PRE Delay: 6
PRE to PRE Delay: 2
ALL PRE to ACT Delay: 7
ALL PRE to REF Delay: 7
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
AI Clock Twister: Moderate
AI Transaction Booster: Manual
Common Performance Level: 10
All Pull-Ins: Disabled
CPU Voltage: 1.41250
CPU GTL Voltage Reference (0/2): 0.675
CPU GTL Voltage Reference (1/3): 0.675
CPU PLL Voltage: 1.60
FSB Termination Voltage: 1.36
DRAM Voltage: 2.20
NB Voltage: 1.36
NB GTL Reference: 0.675
SBridge Voltage: 1.2
PCIE SATA Voltage: Auto
Load Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Auto
CPU Clock Skew : 100
NB Clock Skew : 200
CPU Margin Enhancement: Compatible
I can't stay away from the higher numbers, but I worry that my numbers are getting a bit high. What do you guys think?
TRENDING THREADS
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 10K
-
Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)- Started by Vattila
- Replies: 13K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
-
Facebook
Twitter