Wait a minute, you put CL Ultra on your bare die and then mounted your block directly?
Be prepared to find the IB die stuck to your block when you try and pry that PCB off of it. My indigo extreme was like glue, I had to sand the IHS and my H100 to get it off.
I worry that your IB CPU won't stay bonded to the PCB, but rather that it will prefer to stay bonded to your block should you attempt to separate the two at a later date.
(and if you do attempt it, please let us know if it went well)
Would be great to get confirmation. I know it can be readily pulled apart if done reasonably quickly after application (within days), but I was under the impression that it sets in, diffuses into the block and the die, and essentially solders them together like a cold-weld.
What other forums are you looking at? I'd like to check their findings.
I am using CL Ultra with results comparable, if a bit lower, than CL PRO. I have ordered some CL PRO to try next. Many at OCN still like PRO best, although more are trying Ultra all the time with good results.
Also, it has been a couple of weeks already with Ultra between my die and IHS. I will let all know what state it is in when I remove it and if it was still liquid or not. And if PRO lowers my temps or not too.
Thanks - the one labeled: "3770k IHS Removals - CPU temp dropped from 79C to 71C", or another one? I have already read through that thread. I hope there is another good one to read through that I missed!
Yuriman
Diamond Member
- Jun 25, 2004
- 5,530
- 141
- 106
Decided to repost my results both here and on XS. I have some updated temps now that CL Ultra has had some time to "settle in". Here is approximately what I posted on XS:
To begin with, my system is as follows:
3570K
ASRock Z77 Extreme4
2xHD4870 Crossfire
WC loop - Apogee GT on my CPU + MCW60 on GPU1 + Black Ice GT 360 (3x120mm) rad + Laing DDC w/modded top (1/2" intake)
2x 4GB G.Skill Sniper DDR3 1600 1.25v
Intel X-25M G2 80GB
BFG LS 680w
+ some other stuff that isn't terribly relevant to this post.
Throughout most of my testing, room temperature has had little variation. When I say "Prime" I am generally referring to Prime small FFT testing done for at least 5-10 minutes to get my water temperature up to peak. Voltages are all the lowest drooped voltages (ie under Prime or Linpack) as reported via CPU-Z. Temperatures are all reported via CoreTemp or RealTemp and are taken after the paste has had at least 24 hours to cure.
To begin with, my chip seemed to be one of the worse Ivy Bridge i5's. It needed slightly above average voltage to hit a given clock and had well above average temperatures. After reading Idontcare's post on Anandtech I decided to delid my chip and see if I could improve my temperatures and/or lower my operating voltage.
Before delidding, I had been using Phobya HeGrease between the IHS and my waterblock, a compound that reviews generally agree is among the best non-metal pastes. I had the following temperatures loaded with Prime Small FFT at a room temperature of approximately 25c:
4.4ghz - 78c (1.192v)
4.6ghz - 89c (1.304v)
Post-pop Prime Small FFT temps, with the IHS remounted over the die (secured in place with the stock processor retention bracket) and having replaced Intel's stock paste with Phobya HeGrease, I had the following temperatures:
4.4ghz - 79c
4.6ghz - 94c
^
Temperatures actually got worse after replacing Intel's TIM with HeGrease. Seemingly, my temperature woes were not a result of either excessive gap or poor thermal compound used between the IHS and die, which were the two leading hypotheses on Anandtech for high temperatures on Ivy Bridge chips.
I decided to check the flatness of my waterblock's base and the surface of the IHS, and discovered that the IHS was more than a little concave. I thought about lapping the IHS, but lacking sandpaper I decided just to mount my waterblock directly to the die.
^ Initial temperatures right after delidding, before TIM had a chance to settle in. Due to increased temperatures I had to increase voltage to get stable.
24 hours of cure time later (after my bare die mount), my Prime temperatues dropped to around 4c below where I had been before delidding (78c -> 74c). This still seemed very high for a bare die mount, so picked up some sandpaper and went at my waterblock.
Before:
After (1000grit max):
Lapping my waterblock brought my load temperatures down to 71c at 4.4ghz, and down to 84c (from 89c before delidding) - I was again stable at 4.6ghz. IBT load was down to 91c from 98c before delidding (~1.3v), but this still wasn't anywhere near the results I wanted.
I ordered some CoolLaboratory Liquid Ultra, and applied it:
Which brings me to where I am now.
Prime loaded -
4.4ghz:
78c unmolested
71c lapped WB + bare die mount
61c using CL Ultra instead of Phobya HeGrease
4.6ghz:
89c unmolested
84c lapped WB + bare die mount
70c using CL Ultra
4.7ghz:
?? unmolested
100c lapped WB + bare die mount
78c using CL Ultra
IBT loaded -
4.6ghz:
98c unmolested
91c lapped WB + bare die mount
77c using CL Ultra
______________
I hope this is useful!
To begin with, my system is as follows:
3570K
ASRock Z77 Extreme4
2xHD4870 Crossfire
WC loop - Apogee GT on my CPU + MCW60 on GPU1 + Black Ice GT 360 (3x120mm) rad + Laing DDC w/modded top (1/2" intake)
2x 4GB G.Skill Sniper DDR3 1600 1.25v
Intel X-25M G2 80GB
BFG LS 680w
+ some other stuff that isn't terribly relevant to this post.
Throughout most of my testing, room temperature has had little variation. When I say "Prime" I am generally referring to Prime small FFT testing done for at least 5-10 minutes to get my water temperature up to peak. Voltages are all the lowest drooped voltages (ie under Prime or Linpack) as reported via CPU-Z. Temperatures are all reported via CoreTemp or RealTemp and are taken after the paste has had at least 24 hours to cure.
To begin with, my chip seemed to be one of the worse Ivy Bridge i5's. It needed slightly above average voltage to hit a given clock and had well above average temperatures. After reading Idontcare's post on Anandtech I decided to delid my chip and see if I could improve my temperatures and/or lower my operating voltage.
Before delidding, I had been using Phobya HeGrease between the IHS and my waterblock, a compound that reviews generally agree is among the best non-metal pastes. I had the following temperatures loaded with Prime Small FFT at a room temperature of approximately 25c:
4.4ghz - 78c (1.192v)
4.6ghz - 89c (1.304v)
Post-pop Prime Small FFT temps, with the IHS remounted over the die (secured in place with the stock processor retention bracket) and having replaced Intel's stock paste with Phobya HeGrease, I had the following temperatures:
4.4ghz - 79c
4.6ghz - 94c
^
Temperatures actually got worse after replacing Intel's TIM with HeGrease. Seemingly, my temperature woes were not a result of either excessive gap or poor thermal compound used between the IHS and die, which were the two leading hypotheses on Anandtech for high temperatures on Ivy Bridge chips.
I decided to check the flatness of my waterblock's base and the surface of the IHS, and discovered that the IHS was more than a little concave. I thought about lapping the IHS, but lacking sandpaper I decided just to mount my waterblock directly to the die.
^ Initial temperatures right after delidding, before TIM had a chance to settle in. Due to increased temperatures I had to increase voltage to get stable.
24 hours of cure time later (after my bare die mount), my Prime temperatues dropped to around 4c below where I had been before delidding (78c -> 74c). This still seemed very high for a bare die mount, so picked up some sandpaper and went at my waterblock.
Before:
After (1000grit max):
Lapping my waterblock brought my load temperatures down to 71c at 4.4ghz, and down to 84c (from 89c before delidding) - I was again stable at 4.6ghz. IBT load was down to 91c from 98c before delidding (~1.3v), but this still wasn't anywhere near the results I wanted.
I ordered some CoolLaboratory Liquid Ultra, and applied it:
Which brings me to where I am now.
Prime loaded -
4.4ghz:
78c unmolested
71c lapped WB + bare die mount
61c using CL Ultra instead of Phobya HeGrease
4.6ghz:
89c unmolested
84c lapped WB + bare die mount
70c using CL Ultra
4.7ghz:
?? unmolested
100c lapped WB + bare die mount
78c using CL Ultra
IBT loaded -
4.6ghz:
98c unmolested
91c lapped WB + bare die mount
77c using CL Ultra
______________
I hope this is useful!
Great thread everyone! Read all of it over the last few days.
In conclusion, it appears circumcision of the cpu (removal of the IHS) and direct block-die contact results in greatest thermal conductivity (as expected).
I do have a few questions.
1) Once the cpu hold down mechanism is removed, it's strictly the pressure of the cooling block to the board that permits all cpu pads to engage their respective pins in the socket?
2) With respect to #1 then, how to ensure correct pressure? Obviously it is important to tighten each nut/wheel identical torque to keep the cpu seated level. Perhaps using an in/lb based torque wrench would be proper. I recall reading about this earlier in the thread, but wasn't entirely clear on the practical process? That is, how do you determine how tight is tight?
3) @Yuri, have you tried to remove the block from your cpu yet to se what state the CL Ultra is in now?
I had some other questions, but they have escaped me at the moment.
My intended build will be a i7 3770K in an Asus P8Z77-V Deluxe.
For cooling, I plan on going with the XSPC Raystorm EX240 Universal CPU Water Cooling Kit (http://www.frozencpu.com/products/1...ariant_Pump_Included_and_Free_Dead-Water.html)
This will be my first venture into liquid cooling. The current setup is a q9550 on a p5kdelue @ 3.8ghz with a zalman 97xx hsf. This setup works well, but it would be nice to get into a current cpu and eliminate the noise.
Thanks
In conclusion, it appears circumcision of the cpu (removal of the IHS) and direct block-die contact results in greatest thermal conductivity (as expected).
I do have a few questions.
1) Once the cpu hold down mechanism is removed, it's strictly the pressure of the cooling block to the board that permits all cpu pads to engage their respective pins in the socket?
2) With respect to #1 then, how to ensure correct pressure? Obviously it is important to tighten each nut/wheel identical torque to keep the cpu seated level. Perhaps using an in/lb based torque wrench would be proper. I recall reading about this earlier in the thread, but wasn't entirely clear on the practical process? That is, how do you determine how tight is tight?
3) @Yuri, have you tried to remove the block from your cpu yet to se what state the CL Ultra is in now?
I had some other questions, but they have escaped me at the moment.
My intended build will be a i7 3770K in an Asus P8Z77-V Deluxe.
For cooling, I plan on going with the XSPC Raystorm EX240 Universal CPU Water Cooling Kit (http://www.frozencpu.com/products/1...ariant_Pump_Included_and_Free_Dead-Water.html)
This will be my first venture into liquid cooling. The current setup is a q9550 on a p5kdelue @ 3.8ghz with a zalman 97xx hsf. This setup works well, but it would be nice to get into a current cpu and eliminate the noise.
Thanks
Yuriman
Diamond Member
- Jun 25, 2004
- 5,530
- 141
- 106
1) Correct
2) I wouldn't even attempt to attach my waterblock to the bare die if I didn't have a spring-based mounting bracket that allowed me to easily apply even pressure.
I did play with the mounting pressure when I first installed it on the bare die, and discovered that the springs do need to be slightly compressed, but not very far, and backing off one of the springs typically resulted in my PC just shutting off. It appears it doesn't really take all that much pressure to ensure good contact between the CPU contacts and the pins in the socket.
3) If I have time I'll try tomorrow but I'm inclined to believe that if the TIM is even remotely warm I won't have trouble separating the two. I used to pull Athlon64's out of their sockets all the time when removing heatsinks due to the TIM between the heatspreader and base of the heatsink, but that doesn't make arctic silver potentially dangerous to a bare die, since it has so much less area.
As for the watercooler, from what I can see about it on FrozenCPU it looks like a fair kit, but the premium over buying the parts individually looks to be $50 to $100.
In total honesty, if I were to do it again, I wouldn't watercool. I can't justify spending several hundred dollars for the slightly improved cooling capacity, though that obviously may not apply to you. Mostly, I think of watercooling as allowing you to have a larger surface area for cooling your CPU, and a 2x120mm rad isn't really significantly bigger than some of the heatsinks you can buy now for a fifth the price. I'd aim for a triple rad or at least a high density double rad.
Last edited:
My experience mirrors those of Yuriman's. I initially did not use springs as counter-pressure to the torque applied from my H100 and what I observed was that if I over-tightened the bolts then the system would not post.
My approach was to loosely tighten the bolts, by no mean have them tight enough for good cooling. I just kept lightly tightening them and attempting to boot the computer. Eventually they were tight enough (I did 1/4 turns on all four bolts between boot attempts).
One the bolts were tight enough that the system booted into windows, I then popped open a temperature monitoring program and loaded up LinX.
Obviously the temps were high because the H100 block was loosely bolted down. Then I would slowly tighten the bolts down, 1/4 turns on all four in the criss-cross fashion. Keeping an eye on the temps and watching them fall as the mounting pressure improved.
Eventually the bolts would get too tight and the system would reboot. At that point I backed each bolt back by 1/4 of a turn and rebooted into the system and called it a day.
As I modified my H100 mounting some, I experienced this as well under high strain, so add a 3rd anecdotal report to that.
@Yuri & IDC, thanks for the replies. Kind of figured as much with regard to the block mounting torque (kind of sort of like doing plugs on an Al head in an area you can't get a torque wrench in).
As for the cooling solution, it is an investment, a big one at that. I'm bouncing between a H100, the solution above, and perhaps modding what I have now to work with a socket 1155. Seems with every upgrade (3-4 years), I end up buying another hsf.
OTOH, I keep reading about maintenance and what not, like you say, bigger hassle than its worth. What air cooler would you go with that would let you run the 3770K at 4.6gHz while still being in the 70-80C range while under full linpack load?
As for the cooling solution, it is an investment, a big one at that. I'm bouncing between a H100, the solution above, and perhaps modding what I have now to work with a socket 1155. Seems with every upgrade (3-4 years), I end up buying another hsf.
OTOH, I keep reading about maintenance and what not, like you say, bigger hassle than its worth. What air cooler would you go with that would let you run the 3770K at 4.6gHz while still being in the 70-80C range while under full linpack load?
I had a similar quandary when I bought my 2600k - couldn't decide between an H100 and a NH-D14.
In the end I went with both just to test them out and know where that end of the performance/buck spectrum lands.
My conclusion was that you really are better off ($-wise) going with the NH-D14.
I like my H100, and it outperforms my NH-D14, but only because I slapped on four $30 noiseblocker fans in push-pull config to keep the noise down to acceptable levels. (so $120 in fans is all it takes :|)
Now that I have my H100 I am happy to use it, but if I was faced with a question like you are and I had a case which could take the tower height of a NH-D14 then I'd go for the air cooler and save the money. Your temps will be neck and neck with an H80 and only slightly lower (2-3C) than an H100.
^^I'll read through that thread, thanks!
Btw, an idea occurred about the cutting device used for the circumcision. Instead of using a blade, why not use fishing line? It would probably require a helper, but should be just as easy to do without any damage to the processor, or scratching it up. On the auto sites, many use fishing line to remove factory emblems/lettering without scratching up the paint.
One helper would keep tension on the line, while the other manipulates the processor around it.
Btw, an idea occurred about the cutting device used for the circumcision. Instead of using a blade, why not use fishing line? It would probably require a helper, but should be just as easy to do without any damage to the processor, or scratching it up. On the auto sites, many use fishing line to remove factory emblems/lettering without scratching up the paint.
One helper would keep tension on the line, while the other manipulates the processor around it.
Circumcision with fishing line D: OK, that just effected some immediate shrinkage upon reading
I can't say whether it would work or not for most people, but I can say with some confidence that there is no way fishing line would have worked with my particular CPU. The adhesive was so solid, and the adhesion was so rock tight that I not only had to use a razor blade to get under the IHS but I also had to tap on the blade with a hammer just to coax it between the PCB and the IHS.
In my mind there is no question that the fishing line would have simply snapped while attempting to use it to saw through the adhesive. But going by other people's accounts of the ease with which they delidded their CPUs I suspect those folks could get away with the much simpler and CPU-friendly idea of yours. Someone should try it out :thumbsup:
Just wanted to share my experience with the delidding process and results.
Asus p8z77v deluxe
i7 3770K
g.skil ares 32GB (8gb x 4) http://www.newegg.com/Product/Produc...82E16820231561
xspc ex360 radiator w/raystorm block and d5 pump
Under stock settings, temperatures were ok (<55C) under load (Intel burn and prime95 small&large fft). It wasn't until overclocking that things started to really heat up, hitting temps of 85-95 under above loads at 4.3/4.4.
Figured screw it. Delidded, used CL Liquid Ultra, no IHS. Tried Gpz1100's idea of fishing line. What I had on hand just didn't have the strength (don't recall the rating, but it wasn't overly heavy duty). Went at it with a razor blade instead. That sucker was on there good. It finally yielded and was off.
Temps immediately dropped big time. At 4.8 @ 1.36 - 1.38, 60-62C was as warm as things got. With no thermal issues, I attempted to see what the processor was capable of. I could get the machine to boot, but was unstable at 4.9. Trying various voltages up to 1.40, no long term stability at either 4.8 or 4.7. The machine would BSOD running prime95 sooner or later. LLC set to high.
The goal of course was to achieve long term stability. At 4.6 with an offset of .050, things were pretty stable. Vc of 1.272 under load (vid 1.266V). Prime small&large, and intelburn resulted in no crashes/bsod after running each for an hr. I may be able to drop the offset some more, but really, .05V isn't going to make a big difference I think, especially since my thermal headroom is huge (~55C under above loads). LLC is still set to high. I could probably massage the settings to get it stable at 4.7 or 4.8, but really, 200 MHz isn't worth the time.
As for the cooling setup itself, it seems what I have high is way overkill. All of the above readings were with the fans turned down to about 1200rpm, and the pump at the #3 (half way) position. In fact, under light/noload the fans reduce ~900 rpm. I suppose it's justifiable in that this setup will be recycled on the next system build. Sort of like buy 4 zalman's all at once
All in all, interesting experience to say the least. I've been building machines for various venues for nearly 2 decades now. This is the first cpu I've intentionally caused direct voiding of warranty within a week of purchase While I likely could of left it stock and still had stability at the lower OC, the temps were not pleasing at all. Even notebook cpus don't run that hot. Leaving it stock would leave the water cooling system vastly underutilized.
It probably makes most sense to delid and use a decent(quite) air cooler with a small OC (4.2?), which I suspect would result in more reasonable temps (60-70C ? )
Asus p8z77v deluxe
i7 3770K
g.skil ares 32GB (8gb x 4) http://www.newegg.com/Product/Produc...82E16820231561
xspc ex360 radiator w/raystorm block and d5 pump
Under stock settings, temperatures were ok (<55C) under load (Intel burn and prime95 small&large fft). It wasn't until overclocking that things started to really heat up, hitting temps of 85-95 under above loads at 4.3/4.4.
Figured screw it. Delidded, used CL Liquid Ultra, no IHS. Tried Gpz1100's idea of fishing line. What I had on hand just didn't have the strength (don't recall the rating, but it wasn't overly heavy duty). Went at it with a razor blade instead. That sucker was on there good. It finally yielded and was off.
Temps immediately dropped big time. At 4.8 @ 1.36 - 1.38, 60-62C was as warm as things got. With no thermal issues, I attempted to see what the processor was capable of. I could get the machine to boot, but was unstable at 4.9. Trying various voltages up to 1.40, no long term stability at either 4.8 or 4.7. The machine would BSOD running prime95 sooner or later. LLC set to high.
The goal of course was to achieve long term stability. At 4.6 with an offset of .050, things were pretty stable. Vc of 1.272 under load (vid 1.266V). Prime small&large, and intelburn resulted in no crashes/bsod after running each for an hr. I may be able to drop the offset some more, but really, .05V isn't going to make a big difference I think, especially since my thermal headroom is huge (~55C under above loads). LLC is still set to high. I could probably massage the settings to get it stable at 4.7 or 4.8, but really, 200 MHz isn't worth the time.
As for the cooling setup itself, it seems what I have high is way overkill. All of the above readings were with the fans turned down to about 1200rpm, and the pump at the #3 (half way) position. In fact, under light/noload the fans reduce ~900 rpm. I suppose it's justifiable in that this setup will be recycled on the next system build. Sort of like buy 4 zalman's all at once
All in all, interesting experience to say the least. I've been building machines for various venues for nearly 2 decades now. This is the first cpu I've intentionally caused direct voiding of warranty within a week of purchase
While I likely could of left it stock and still had stability at the lower OC, the temps were not pleasing at all. Even notebook cpus don't run that hot. Leaving it stock would leave the water cooling system vastly underutilized.It probably makes most sense to delid and use a decent(quite) air cooler with a small OC (4.2?), which I suspect would result in more reasonable temps (60-70C ? )
Thanks for the info das, I just want to clarify: you had it at 4.8 at only 60C but it was still unstable, even with 1.4V?
Gauchotodd, correct. The instability is not temperature related as far as I can see.
We need to back track for a moment and define what is considered stable. For me, stability is defined by being able to run intelburn (or equiv) & prime95 (at both small/large/blend settings) for some period of time of at least 30-60 minutes without the program crashing or a bsod. I believe this puts the most load on the cpu, and if the test is successful, then I consider the system to be suitable for any task in daily use. In addition, if it passes these, then I'll run intelburn to utilize all ram (32gb will take a while for each pass). My testing methodology is to determine the cpu limits separate from the memory.
At 4.8ghz/1.4V intelburn was ok (using 4GB), but prime95 was bsod'ing on the small/large tests after a period of ~5-10 min. Bear in mind, the machine has 32GB of memory in it, but these tests are more focused on the cpu rather than memory. Although these tests use little memory, the rest of this ram is kept track of in some capacity. This probably has some effect also. The memory is being run at stock xmp settings, memory frequency was not increased from 1866mhz, nor were the timing tightened. I knew going in with 32GB timing would have to be on the loose side to retain stability.
I left the other voltages - VccPLL, VccIO, VccSA all on auto. Bumping them might correct these bsods, but is 200 MHz (from 4.6Ghz) is really worth the marginal gain. Right now the Vcc is somewhere around 1.26V and 1.29V while under load. Temps don't exceed 55C (mostly topping out at 52/53C), system board temp is under 30C (ambient is ~24C). System is rock stable with no bsod or crashed prime95/intelburn under any of the tests, esp prime95 small/large. The machine will never see such loads even under the heaviest daily usage.
I could probably have left it at 4.8 and still had a relatively reliable experience under typical use, but given my OCD nature (common with quite a few in here ) it would of bugged me.
Btw, I do want to make a few comments about newegg's video with JJ the asus marketing rep - http://youtu.be/4mkGQhE1o2w . While this video is informative and useful for someone completely new, like most other videos on youtube, it needs to be taken with a grain of salt (many grains). The video shows how easy it is to overclock. While this is true, he barely touches on how to confirm the validity of the overclock. Running the automated overclock function is insufficient to thoroughly assure system stability in my opinion. The tests are just too short to really load it up. Same with aida64 and some of the other programs he mentions.
In fact, he talks about intelburn/prime95 possibly causing damage because they overload certain functions of the cpu but don't test others. I think this too is in error because the new libraries of both programs use the avx instructions. I think the reason he mentions this is because while the cpu passes these other tests, under certain overclocks (such as in my case), it will fail prime95/intelburn tests. From a marketing perspective, this is won't help sell more asus boards. To me, a system is unstable if it fails any test period.
(Sorry about the long rant, just wanted to be detailed in my response).
We need to back track for a moment and define what is considered stable. For me, stability is defined by being able to run intelburn (or equiv) & prime95 (at both small/large/blend settings) for some period of time of at least 30-60 minutes without the program crashing or a bsod. I believe this puts the most load on the cpu, and if the test is successful, then I consider the system to be suitable for any task in daily use. In addition, if it passes these, then I'll run intelburn to utilize all ram (32gb will take a while for each pass). My testing methodology is to determine the cpu limits separate from the memory.
At 4.8ghz/1.4V intelburn was ok (using 4GB), but prime95 was bsod'ing on the small/large tests after a period of ~5-10 min. Bear in mind, the machine has 32GB of memory in it, but these tests are more focused on the cpu rather than memory. Although these tests use little memory, the rest of this ram is kept track of in some capacity. This probably has some effect also. The memory is being run at stock xmp settings, memory frequency was not increased from 1866mhz, nor were the timing tightened. I knew going in with 32GB timing would have to be on the loose side to retain stability.
I left the other voltages - VccPLL, VccIO, VccSA all on auto. Bumping them might correct these bsods, but is 200 MHz (from 4.6Ghz) is really worth the marginal gain. Right now the Vcc is somewhere around 1.26V and 1.29V while under load. Temps don't exceed 55C (mostly topping out at 52/53C), system board temp is under 30C (ambient is ~24C). System is rock stable with no bsod or crashed prime95/intelburn under any of the tests, esp prime95 small/large. The machine will never see such loads even under the heaviest daily usage.
I could probably have left it at 4.8 and still had a relatively reliable experience under typical use, but given my OCD nature (common with quite a few in here
) it would of bugged me.Btw, I do want to make a few comments about newegg's video with JJ the asus marketing rep - http://youtu.be/4mkGQhE1o2w . While this video is informative and useful for someone completely new, like most other videos on youtube, it needs to be taken with a grain of salt (many grains). The video shows how easy it is to overclock. While this is true, he barely touches on how to confirm the validity of the overclock. Running the automated overclock function is insufficient to thoroughly assure system stability in my opinion. The tests are just too short to really load it up. Same with aida64 and some of the other programs he mentions.
In fact, he talks about intelburn/prime95 possibly causing damage because they overload certain functions of the cpu but don't test others. I think this too is in error because the new libraries of both programs use the avx instructions. I think the reason he mentions this is because while the cpu passes these other tests, under certain overclocks (such as in my case), it will fail prime95/intelburn tests. From a marketing perspective, this is won't help sell more asus boards. To me, a system is unstable if it fails any test period.
(Sorry about the long rant, just wanted to be detailed in my response).
Even though you are wanting to focus on the CPU, if you restrict the memory size during the IBT test then the power-consumption, and thus CPU temperature, will be lower than it would otherwise be if you set the memory to "ALL".
Any results you see from me in these forums are from runs that have passed a minimum of 5-cycles with ~14.2GB of ram (I specifically use problem size 43122 for consistency across all test conditions). This generally takes about an hour to complete, more time if the CPU is clocked slower, less time if clocked fast of course.
@DAS,
I agree with IDC, use the small memory memory size for initial testing, but definitely run several passes using the full available memory for a final stress test.
I agree with IDC, use the small memory memory size for initial testing, but definitely run several passes using the full available memory for a final stress test.
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: 9K
-
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
-
Facebook
Twitter