Overclock.net › Forums › Intel › Intel CPUs › [MOD] LGA775 Support For LGA771 Xeon CPUs
New Posts  All Forums:Forum Nav:

[MOD] LGA775 Support For LGA771 Xeon CPUs - Page 1188

post #11871 of 12068
Quote:
Originally Posted by schuck6566 View Post

Can't honestly say, but sounds self defeating,because more voltage =more heat.So even if the ambient temp was lower,more voltage would increase the cpu temp & offset any gain from the lower ambient temps.@ least that's how I'm seeing the logic. Please feel free to let us know.thumb.gif

That's not exactly what i meant. The question i asked was a bit nonsense, i accept that. But let me explain.

At 4,4ghz i may need like 1,45ish to be stable in linx (or maybe it's impossible). Now, at that speed and voltage i get shut down PC. Temp issue. To some extent CPU speed gets to a wall when not even increasing the voltage will make it stable, even if you can keep it cold enough. That could be also the case.

In the actual conditions, room at 25ºC (and it will go far over that as we aproach summer) the temp goes so high that the Computer just shuts down for security reasons. in winter, knowing that, what i only need is a very cold night were ambient room would be like 5 or 6 degrees (obviously on purpose) , i could lower temp by almost 20ºC in reference to now, or even more giving more room the CPU to complete the stresstest at bareable temps so the Computer will not shut down, which would give me enough margin to try 4,4ghz at 1,45v.

At this time, i'm almost sure my CPU can pass a stresstest at 4,4ghz at 1,45v ish (not 4,5ghz since i would need like 1,50ish too much voltage and temps). The problem is that i'm not able right now to get it cold enough for it to pass it. i just need one night to try it. Then, knowing that not even a game would stress the CPU to that high, i can be at 4,4ghz just in the limit.

In my case, adding more voltage i don't think is the issue. The issue here is that it's overheating not voltage. i don't think i need more voltage. Just less ambient temp for it to pass it. There is not more voltage in the ecuation as i tried, so this is where your point may not be true.

to sum up:_

4,4ghz at 1,45v should be enough for the CPU to pass the test (according to what i've seen scaling). I cannot pass it now because temps are so high that the computer shuts down. If i get cold ambient room temps, there is a possibility that my AIO will handle that and cool the CPU enough for it to pass the test. No more voltage should be needed. Maybe 0, 650v but that should not change much the temps.

By now, i will stay at 4,3ghz 1,41250v linx stable. i don't think i should go for 4,5ghz at 1,45v windows stable only, since at any moment i can get a bsod, and that, we all know is likely to occur in the worst moment possible.
post #11872 of 12068
Quote:
Originally Posted by chris89 View Post

Whoops i slipped and this spilled out.. on accident haha wink.gif no more I promise... on $1 board ebay well below default specs haha


1$ board?? lucky boy, thumb.gif

Can you pass any stresstest with that?
post #11873 of 12068
Quote:
Originally Posted by Piskeante View Post

That's not exactly what i meant. The question i asked was a bit nonsense, i accept that. But let me explain.

At 4,4ghz i may need like 1,45ish to be stable in linx (or maybe it's impossible). Now, at that speed and voltage i get shut down PC. Temp issue. To some extent CPU speed gets to a wall when not even increasing the voltage will make it stable, even if you can keep it cold enough. That could be also the case.

In the actual conditions, room at 25ºC (and it will go far over that as we aproach summer) the temp goes so high that the Computer just shuts down for security reasons. in winter, knowing that, what i only need is a very cold night were ambient room would be like 5 or 6 degrees (obviously on purpose) , i could lower temp by almost 20ºC in reference to now, or even more giving more room the CPU to complete the stresstest at bareable temps so the Computer will not shut down, which would give me enough margin to try 4,4ghz at 1,45v.

At this time, i'm almost sure my CPU can pass a stresstest at 4,4ghz at 1,45v ish (not 4,5ghz since i would need like 1,50ish too much voltage and temps). The problem is that i'm not able right now to get it cold enough for it to pass it. i just need one night to try it. Then, knowing that not even a game would stress the CPU to that high, i can be at 4,4ghz just in the limit.

In my case, adding more voltage i don't think is the issue. The issue here is that it's overheating not voltage. i don't think i need more voltage. Just less ambient temp for it to pass it. There is not more voltage in the ecuation as i tried, so this is where your point may not be true.

to sum up:_

4,4ghz at 1,45v should be enough for the CPU to pass the test (according to what i've seen scaling). I cannot pass it now because temps are so high that the computer shuts down. If i get cold ambient room temps, there is a possibility that my AIO will handle that and cool the CPU enough for it to pass the test. No more voltage should be needed. Maybe 0, 650v but that should not change much the temps.

By now, i will stay at 4,3ghz 1,41250v linx stable. i don't think i should go for 4,5ghz at 1,45v windows stable only, since at any moment i can get a bsod, and that, we all know is likely to occur in the worst moment possible.
So basicly it's the board temps that are shutting you down @ the higher OC"s? Not the cpu temps? Reason I'm asking is part of that may be poor ventilation in case, and part may be able to be handled by turning off the temp warnings/monitoring in the bios if you're gonna handle it with the 3rd party software & keep a eye on it during the testing. Or maybe just throw an extra fan pointing at the chipset to move the hotter air away from the board itself?(with the AIO we loose the fan over the cpu moving air over the board/chips) Does it specify WHAT the temp issue IS? CPU,Northbridge,Southbrifge,or Tcase? TCASE=Case Temperature is the maximum temperature allowed at the processor Integrated Heat Spreader (IHS). <-- different horse from the tjmax. An extra fan blowing across the board has been known to lower the last 3 temps bt 1-3 degrees or more in a lot of cases. If you're only going over a degree or 2 and getting tripped up like that,zip tying a fan in your case so it blows on the board may be worth a try.wink.gif
post #11874 of 12068
@Piskeante It will pass CPU Queen and Cache & Memory Benchmark among Experience Index. Cpu & cache and memory benchmark hit 11,500MB/s copy @ 3.5ghz. Just need cpu voltage control among others. Running stock voltage there. That's why going newer-tech ddr3 chipset.

It runs at the E5450's stock idle temperature at load, 40C all cores. Which tells me with voltage, 5Ghz no problem. 4Ghz is a breeze no sweat 1 run in the BIOS and 4Ghz 100% stable.
Edited by chris89 - 4/14/17 at 8:15am
post #11875 of 12068
Quote:
Originally Posted by schuck6566 View Post

So basicly it's the board temps that are shutting you down @ the higher OC"s? Not the cpu temps? Reason I'm asking is part of that may be poor ventilation in case, and part may be able to be handled by turning off the temp warnings/monitoring in the bios if you're gonna handle it with the 3rd party software & keep a eye on it during the testing. Or maybe just throw an extra fan pointing at the chipset to move the hotter air away from the board itself?(with the AIO we loose the fan over the cpu moving air over the board/chips) Does it specify WHAT the temp issue IS? CPU,Northbridge,Southbrifge,or Tcase? TCASE=Case Temperature is the maximum temperature allowed at the processor Integrated Heat Spreader (IHS). <-- different horse from the tjmax. An extra fan blowing across the board has been known to lower the last 3 temps bt 1-3 degrees or more in a lot of cases. If you're only going over a degree or 2 and getting tripped up like that,zip tying a fan in your case so it blows on the board may be worth a try.wink.gif

nope, 1,3v on my board is enough to go 4,4ghz. That's not the issue.

It shuts down because when the CPU , not even thermal throttling can go below max temp (which should be at around 100ºC), it just gives up and the board says bye bye.
I need more "horsepower" to cool the CPU, and that's not at my hand right now. The only thing i can do is get as low as possible in the ambient temps, so that very cold air goes through the AIO to keep it below 90ºC.
post #11876 of 12068
Quote:
Originally Posted by chris89 View Post

@Piskeante It will pass CPU Queen and Cache & Memory Benchmark among Experience Index. Cpu & cache and memory benchmark hit 11,500MB/s copy @ 3.5ghz. Just need cpu voltage control among others. Running stock voltage there. That's why going newer-tech ddr3 chipset.

those are benchmarks, not stresstest. According to what you say, you are just stable with low requisits bench, but would fail IBT, Prime95 or Linx, am i correct??

afaik the best of your config is running newer boards with ddr3. no argument possible there. Your platform is better. Get a good board and show us what that Opteron can do. Cinebench R15 would be very nice, since it's a good method of testing the performance of a CPU.
post #11877 of 12068
I'm waiting on RAM at the moment. Ordered over a month ago. Takes forever. Will LYK. thumb.gif
post #11878 of 12068
Quote:
Originally Posted by chris89 View Post

I'm waiting on RAM at the moment. Ordered over a month ago. Takes forever. Will LYK. thumb.gif

500 points in Cinebench R15 to beat me. a bit more to beat Tnlgg. For that, it doesn't matter if you are not stable. It's just for fun (but remember to say you are not stable), just for others to know it.
post #11879 of 12068
Quote:
Originally Posted by Piskeante View Post

500 points in Cinebench R15 to beat me. a bit more to beat Tnlgg. For that, it doesn't matter if you are not stable. It's just for fun (but remember to say you are not stable), just for others to know it.

Cant beat me,i can run Cinebench R15 @ 4.8GHz if i want to,just for fun.
CryMods [ON] OFF
(12 items)
 
  
CPUMotherboardGraphicsRAM
[X5470 E0 @4.7GHz] [Gigabyte P35 DQ6 / Bios F9] [Sapphire Nitro+ RX 480 4GB] [Corsair 8GB DDR2 1066MHz @2.1V] 
Hard DriveOSMonitorKeyboard
[240GB SSD + WD Blue 1TB SATA] [Windows 10] [1x AOC G2460PF & 2x AOC G2460VQ6] [Logitech G15] 
PowerCaseMouse
[EVGA 750 GQ] [Lian Li PC-P50] [Logitech RX 1000] 
  hide details  
Reply
CryMods [ON] OFF
(12 items)
 
  
CPUMotherboardGraphicsRAM
[X5470 E0 @4.7GHz] [Gigabyte P35 DQ6 / Bios F9] [Sapphire Nitro+ RX 480 4GB] [Corsair 8GB DDR2 1066MHz @2.1V] 
Hard DriveOSMonitorKeyboard
[240GB SSD + WD Blue 1TB SATA] [Windows 10] [1x AOC G2460PF & 2x AOC G2460VQ6] [Logitech G15] 
PowerCaseMouse
[EVGA 750 GQ] [Lian Li PC-P50] [Logitech RX 1000] 
  hide details  
Reply
post #11880 of 12068
Quote:
Originally Posted by Tnlgg View Post

Cant beat me,i can run Cinebench R15 @ 4.8GHz if i want to,just for fun.

jejeje, i should have imagined that. I can also do 4,6ghz upping the voltage over "safe"value 1,45v in case it was necessary., but i don't think it will be, thumb.gifthumb.gif

Which was your score at 4,7ghz??
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › [MOD] LGA775 Support For LGA771 Xeon CPUs