info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Overclock 6700K Max Temps Prime95 Small FFT
12 views
Skip to first unread message
B00ze
Jan 23, 2017, 8:55:49 PM1/23/17
to
Good day.
So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
your overclocked CPU? Because on mine that particular test can make the
temps climb all the way to 93C. The heat-sink is a really good one
(NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
warmer than room temp, and I used good Arctic Silver TIM. While I may
have placed a tiny tiny bit too much I was real careful with the stuff
(maybe I didn't put enough?) Under normal testing I have good temps, but
under that extreme Small FFT Prime95 test, my CPU really cooks - do you
all get the same temps on air?
Thank you.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
oO-( )-Oo Virus detected, delete Windoze? (Y/n).
So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
your overclocked CPU? Because on mine that particular test can make the
temps climb all the way to 93C. The heat-sink is a really good one
(NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
warmer than room temp, and I used good Arctic Silver TIM. While I may
have placed a tiny tiny bit too much I was real careful with the stuff
(maybe I didn't put enough?) Under normal testing I have good temps, but
under that extreme Small FFT Prime95 test, my CPU really cooks - do you
all get the same temps on air?
Thank you.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
oO-( )-Oo Virus detected, delete Windoze? (Y/n).
Paul
Jan 24, 2017, 4:21:42 AM1/24/17
to
B00ze wrote:
> Good day.
>
> So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
> LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
> wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
> are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
> your overclocked CPU? Because on mine that particular test can make the
> temps climb all the way to 93C. The heat-sink is a really good one
> (NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
> warmer than room temp, and I used good Arctic Silver TIM. While I may
> have placed a tiny tiny bit too much I was real careful with the stuff
> (maybe I didn't put enough?) Under normal testing I have good temps, but
> under that extreme Small FFT Prime95 test, my CPU really cooks - do you
> all get the same temps on air?
>
> Thank you.
> Best Regards,
>
That heatsink seems to give pretty good performance here,
> Good day.
>
> So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
> LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
> wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
> are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
> your overclocked CPU? Because on mine that particular test can make the
> temps climb all the way to 93C. The heat-sink is a really good one
> (NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
> warmer than room temp, and I used good Arctic Silver TIM. While I may
> have placed a tiny tiny bit too much I was real careful with the stuff
> (maybe I didn't put enough?) Under normal testing I have good temps, but
> under that extreme Small FFT Prime95 test, my CPU really cooks - do you
> all get the same temps on air?
>
> Thank you.
> Best Regards,
>
in terms of the degreesC/W.
http://www.frostytech.com/articleview.cfm?articleid=2749&page=5
If the internal case air is at 35C, the Noctua added 20C (to get to Tcase),
you added another 25C case to junction (unknown, guess), the
Tjunction should be around 80C. Check and make sure the
computer case has a big enough fan on the back (to keep it at 35C).
If you notice that the CPU temp *drops* when you take
the side off the PC, that means your case cooling
sucks :-)
My case cooling isn't the greatest (since I fitted a
quieter fan on the back). I'm using the GPU as a proxy
for case air temp here. The hard drive is in the front of
the computer and hasn't budged. You can see the GPU temp
coming up a tiny bit, so I'd have to leave it running
for a bit longer, to get a true final temp. You can see
I could push this thing harder, but I'd have to fix
VCore somehow (too hot).
https://s28.postimg.org/ve04pgfl9/prime95_test.gif
That's cooled by a Noctua NH-D15 (with the center fan fitted).
I de-tuned the PC a bit when I got it, as the default
Asus settings were cooking the Vcore regulator. I turned
off the Turbo (which wouldn't make any difference in this
case, as it cannot Turbo if all cores are 100%). The measured
power, using a clamp-on ammeter, was 156W, before I turned
down the setting. That's the power feeding into VCore (ATX12V).
I have no easy way to measure the current flowing on VCore
itself. My clamp-on ammeter cannot fit around the copper
plane :-)
So my conditions aren't even close to a match for yours.
My Vcore heatsink was running 65C with the original settings
(with Prime95 as the test), so I couldn't leave the
thing that way. It probably won't go into thermal
runaway, but I'm not taking a chance. If only I'd
remembered to check the size of the VCore heatsink
before I bought the motherboard :-( Mistake.
Take a picture and show the dynamics of your setup,
from "idle" to running the test for 7-8 minutes. Just
to give some idea what hops up first.
https://postimage.io/index.php?um=flash
Things that matter:
1) Bent CPU.
2) Condition of the Intel TIM, between the silicon die
and the lid. Intel used low-temperatore solder between
the CPU and lid in the LGA775 days. To avoid "conflict
minerals", they went back to conventional dough-like material
on the newer processors.
3) It would take a generous, gushing layer of AS5 to insulate.
On the "too lean" side, you should use an inspection mirror from
the side, and look for the color of the AS5 wetting the junction.
Depending on the size of the heatsink, it might not be possible
to eyeball it. I apply half a rice grain bit of AS, compress, verify
the spread diameter, then adjust the dose for the second and final
installation. Then, hold the inspection mirror around the side,
and look for a wetted junction. Mine don't generally gush all
over as a result. I've under-dosed at least once, and then
it was third-time-lucky.
I stopped spreading it with a credit card, a long time ago :-)
That's just a waste.
Paul
hanky liu
Jan 24, 2017, 7:35:38 PM1/24/17
to
I have my 6700k at 4.5, I just use the xtu and tweak the memory timing and multiplier. When I run adobe media encoder or matlab sim, both keep the CPU at 95-100%, for couple hours at the time, my CPU temp doesn't go over 70.. and voltage is around 1.21-1.29 ish.... I have an Itx case with 980ti which the gpu generates more heat than anything other components..
B00ze
Jan 25, 2017, 12:15:34 AM1/25/17
to
On 2017-01-24 04:21, Paul <nos...@needed.invalid> wrote:
[snip]
> That heatsink seems to give pretty good performance here,
> in terms of the degreesC/W.
Yeah, I bought it even tho I had a Hyper EVO in stock, but the heat is
so concentrated over a small area that I wonder if it really makes a
difference if you buy a more expensive heatsink over the cheaper EVO.
> If the internal case air is at 35C, the Noctua added 20C (to get to Tcase),
> you added another 25C case to junction (unknown, guess), the
> Tjunction should be around 80C. Check and make sure the
> computer case has a big enough fan on the back (to keep it at 35C).
> If you notice that the CPU temp *drops* when you take
> the side off the PC, that means your case cooling
> sucks :-)
Case temp should be fine, it has 3 200mm fans + the back exhaust fan
(Thermaltake Urban T81 full-tower).
> My case cooling isn't the greatest (since I fitted a
> quieter fan on the back). I'm using the GPU as a proxy
That's the beauty of the case I got, since the fans are wide, they run
slower thus less noise.
> for case air temp here. The hard drive is in the front of
> the computer and hasn't budged. You can see the GPU temp
> coming up a tiny bit, so I'd have to leave it running
> for a bit longer, to get a true final temp. You can see
> I could push this thing harder, but I'd have to fix
> VCore somehow (too hot).
> https://s28.postimg.org/ve04pgfl9/prime95_test.gif
I don't understand SpeedFan's graph; that line is around the 45 mark but
I know it's not 45C. And 1.088v is really low (for my CPU anyway, dunno
for your Ivy Bridge 6-Core). Strange that your HD0 temp goes up along
with the CPU, you'd think it would stay cool being in front of the case
away from the back exhaust...
> That's cooled by a Noctua NH-D15 (with the center fan fitted).
> I de-tuned the PC a bit when I got it, as the default
> Asus settings were cooking the Vcore regulator. I turned
> off the Turbo (which wouldn't make any difference in this
> case, as it cannot Turbo if all cores are 100%). The measured
> power, using a clamp-on ammeter, was 156W, before I turned
> down the setting. That's the power feeding into VCore (ATX12V).
> I have no easy way to measure the current flowing on VCore
> itself. My clamp-on ammeter cannot fit around the copper
> plane :-)
There is some bug with HWMonitor on my board, it keeps saying that
"Package" power is between 1W-2W; I wish it would read correctly, my
Watts must be through the roof when it hits 1.360v with all cores
running AVX2 in Prime95.
> So my conditions aren't even close to a match for yours.
> My Vcore heatsink was running 65C with the original settings
> (with Prime95 as the test), so I couldn't leave the
> thing that way. It probably won't go into thermal
> runaway, but I'm not taking a chance. If only I'd
> remembered to check the size of the VCore heatsink
> before I bought the motherboard :-( Mistake.
Not OC'ed (i.e. at 1.200v) my CPU stays around 60C in Prime95 (and in
Intel Burn Test). I don't know why you "couldn't leave things that way"
- 65C is just fine, unless you plan on running Lin-pack 24/7 for the
next 12 years...
> Take a picture and show the dynamics of your setup,
> from "idle" to running the test for 7-8 minutes. Just
> to give some idea what hops up first.
Prime95 Small FFT:
https://www.dropbox.com/s/u69v3u4479k03es/Prime95-SmallFFT-1344v.jpg?dl=0
Prime95 128K FFT (this is as high as it goes, if I leave it running
longer it goes up a degree or two):
https://www.dropbox.com/s/odkzpw8yjz826ex/Prime95-128KFFT-1360v.jpg?dl=0
> Things that matter:
>
> 1) Bent CPU.
Yeah, I thought of that, maybe my HS put too much pressure on the CPU,
but I'd have to take the thing apart to check :-(
> 2) Condition of the Intel TIM, between the silicon die
> and the lid. Intel used low-temperatore solder between
> the CPU and lid in the LGA775 days. To avoid "conflict
> minerals", they went back to conventional dough-like
> material on the newer processors.
Not sure it's to avoid conflict minerals - I read recently that they
found the lid would crack after hot/cold cycles if they used solder
instead of the goo they now use.
> 3) It would take a generous, gushing layer of AS5 to insulate.
>
> On the "too lean" side, you should use an inspection mirror from
> the side, and look for the color of the AS5 wetting the junction.
> Depending on the size of the heatsink, it might not be possible
> to eyeball it. I apply half a rice grain bit of AS, compress, verify
> the spread diameter, then adjust the dose for the second and final
> installation. Then, hold the inspection mirror around the side,
> and look for a wetted junction. Mine don't generally gush all
> over as a result. I've under-dosed at least once, and then
> it was third-time-lucky.
I use the 2-Lines method from here:
http://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62&limit=1&limitstart=4
I used it once, then found out I'd put the heatsink in reverse (there
are little holes around the cooper posts, these I think should be facing
the fan, not the other way around,) so I removed it and did it again.
When I removed it, the paste was perfect, not too much, not too little,
spread nice and thin all over with very little run-off on the sides.
> I stopped spreading it with a credit card, a long time ago :-)
> That's just a waste.
Yeah.
I did some more googling, this time I added "Small FFT" to my terms, and
most people seem to agree that the latest Prime95 Small FFT which uses
AVX2 is just too punishing. If I use OCCT's "CPU" test (instead of it's
LinPack mode) then my CPU stays around 70C @ 1.344v, which I think is
fine; the only bench I have that makes LLC climb to 1.360v is Prime95.
LLC I think is supposed to LOWER the voltage when the CPU is really busy
so that temps stay constant; it's pretty cool to see MB manufacturers
switch it around and supply MORE volts when instead, so that the
overclock remains stable.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
oO-( )-Oo Compile, run, curse. Recompile, rerun, recurse.
[snip]
> That heatsink seems to give pretty good performance here,
> in terms of the degreesC/W.
so concentrated over a small area that I wonder if it really makes a
difference if you buy a more expensive heatsink over the cheaper EVO.
> If the internal case air is at 35C, the Noctua added 20C (to get to Tcase),
> you added another 25C case to junction (unknown, guess), the
> Tjunction should be around 80C. Check and make sure the
> computer case has a big enough fan on the back (to keep it at 35C).
> If you notice that the CPU temp *drops* when you take
> the side off the PC, that means your case cooling
> sucks :-)
(Thermaltake Urban T81 full-tower).
> My case cooling isn't the greatest (since I fitted a
> quieter fan on the back). I'm using the GPU as a proxy
slower thus less noise.
> for case air temp here. The hard drive is in the front of
> the computer and hasn't budged. You can see the GPU temp
> coming up a tiny bit, so I'd have to leave it running
> for a bit longer, to get a true final temp. You can see
> I could push this thing harder, but I'd have to fix
> VCore somehow (too hot).
> https://s28.postimg.org/ve04pgfl9/prime95_test.gif
I know it's not 45C. And 1.088v is really low (for my CPU anyway, dunno
for your Ivy Bridge 6-Core). Strange that your HD0 temp goes up along
with the CPU, you'd think it would stay cool being in front of the case
away from the back exhaust...
> That's cooled by a Noctua NH-D15 (with the center fan fitted).
> I de-tuned the PC a bit when I got it, as the default
> Asus settings were cooking the Vcore regulator. I turned
> off the Turbo (which wouldn't make any difference in this
> case, as it cannot Turbo if all cores are 100%). The measured
> power, using a clamp-on ammeter, was 156W, before I turned
> down the setting. That's the power feeding into VCore (ATX12V).
> I have no easy way to measure the current flowing on VCore
> itself. My clamp-on ammeter cannot fit around the copper
> plane :-)
"Package" power is between 1W-2W; I wish it would read correctly, my
Watts must be through the roof when it hits 1.360v with all cores
running AVX2 in Prime95.
> So my conditions aren't even close to a match for yours.
> My Vcore heatsink was running 65C with the original settings
> (with Prime95 as the test), so I couldn't leave the
> thing that way. It probably won't go into thermal
> runaway, but I'm not taking a chance. If only I'd
> remembered to check the size of the VCore heatsink
> before I bought the motherboard :-( Mistake.
Intel Burn Test). I don't know why you "couldn't leave things that way"
- 65C is just fine, unless you plan on running Lin-pack 24/7 for the
next 12 years...
> Take a picture and show the dynamics of your setup,
> from "idle" to running the test for 7-8 minutes. Just
> to give some idea what hops up first.
https://www.dropbox.com/s/u69v3u4479k03es/Prime95-SmallFFT-1344v.jpg?dl=0
Prime95 128K FFT (this is as high as it goes, if I leave it running
longer it goes up a degree or two):
https://www.dropbox.com/s/odkzpw8yjz826ex/Prime95-128KFFT-1360v.jpg?dl=0
> Things that matter:
>
> 1) Bent CPU.
but I'd have to take the thing apart to check :-(
> 2) Condition of the Intel TIM, between the silicon die
> and the lid. Intel used low-temperatore solder between
> the CPU and lid in the LGA775 days. To avoid "conflict
> minerals", they went back to conventional dough-like
> material on the newer processors.
found the lid would crack after hot/cold cycles if they used solder
instead of the goo they now use.
> 3) It would take a generous, gushing layer of AS5 to insulate.
>
> On the "too lean" side, you should use an inspection mirror from
> the side, and look for the color of the AS5 wetting the junction.
> Depending on the size of the heatsink, it might not be possible
> to eyeball it. I apply half a rice grain bit of AS, compress, verify
> the spread diameter, then adjust the dose for the second and final
> installation. Then, hold the inspection mirror around the side,
> and look for a wetted junction. Mine don't generally gush all
> over as a result. I've under-dosed at least once, and then
> it was third-time-lucky.
http://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62&limit=1&limitstart=4
I used it once, then found out I'd put the heatsink in reverse (there
are little holes around the cooper posts, these I think should be facing
the fan, not the other way around,) so I removed it and did it again.
When I removed it, the paste was perfect, not too much, not too little,
spread nice and thin all over with very little run-off on the sides.
> I stopped spreading it with a credit card, a long time ago :-)
> That's just a waste.
I did some more googling, this time I added "Small FFT" to my terms, and
most people seem to agree that the latest Prime95 Small FFT which uses
AVX2 is just too punishing. If I use OCCT's "CPU" test (instead of it's
LinPack mode) then my CPU stays around 70C @ 1.344v, which I think is
fine; the only bench I have that makes LLC climb to 1.360v is Prime95.
LLC I think is supposed to LOWER the voltage when the CPU is really busy
so that temps stay constant; it's pretty cool to see MB manufacturers
switch it around and supply MORE volts when instead, so that the
overclock remains stable.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
B00ze
Jan 25, 2017, 12:21:49 AM1/25/17
to
On 2017-01-24 19:35, hanky liu <hank...@gmail.com> wrote:
> I have my 6700k at 4.5, I just use the xtu and tweak the memory timing and multiplier. When I run adobe media encoder or matlab sim, both keep the CPU at 95-100%, for couple hours at the time, my CPU temp doesn't go over 70.. and voltage is around 1.21-1.29 ish.... I have an Itx case with 980ti which the gpu generates more heat than anything other components..
I'd have to try those - I don't have them, but if I run OCCT in CPU mode
> I have my 6700k at 4.5, I just use the xtu and tweak the memory timing and multiplier. When I run adobe media encoder or matlab sim, both keep the CPU at 95-100%, for couple hours at the time, my CPU temp doesn't go over 70.. and voltage is around 1.21-1.29 ish.... I have an Itx case with 980ti which the gpu generates more heat than anything other components..
then I stay around 70C (LinPack mode climbs to 88C). It really depends
on the stress program you use. I will install HandBrake next and try
that on a 4K video; apparently your overclock could be 100% stable and
still crash in HandBrake with a large video...
Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
Paul
Jan 25, 2017, 5:15:28 AM1/25/17
to
almost as if the cooler is being "overwhelmed".
Heatpipes have a limited thermal pumping capacity.
If you manage to keep all the working liquid in the
vapor phase (no chance to condense), then the temperature
shoots up. The heatpipe has some sort of "watt rating",
over which it's probably not going to pump heat like it should.
Each heatpipe has two drops of fluid in it - there isn't
a couple of ounces of the stuff. Heatpipes can also leak and
fail, but that hasn't been a problem for years and years.
Manufacturers have figured out how to make good heatpipes.
When heatpipes were still a novelty, at least one product
shipped with no fluid in any of its heatpipes. Making it
gutless.
You could try a more powerful fan on the cooler.
The other issue with these designs, is "no cowling".
Mine doesn't have one either. A cowling on the heatsink,
would ensure that any static pressure the fan can develop,
goes through the fins.
Since I own a 37.5mm thick fan, if I had to attempt a retrofit,
I'd do a sheet metal cowling, plus fit the 37.5mm thick fan.
That's to get as much air velocity as possible through the
fins. I've only done one cowling/shroud for a CPU cooler here,
in the time I've worked on computers.
Air cooling goes asymptotic over about 800LFM (linear feet per minute).
Run of the mill electronics, might use around 200LFM. So there
is a point, where "mo fan" does "no good".
Paul
hanky liu
Jan 25, 2017, 7:51:16 PM1/25/17
to
When I did 4K encoding with adobe I have it running overnight a lot, it's stable plus able to do work while it's running encoding, the CPU is at 100 but system is responsive... one thing I've noticed, it really depend on your ram, you can't install more than ddr4 2400....
B00ze
Jan 25, 2017, 9:10:19 PM1/25/17
to
On 2017-01-25 05:15, Paul <nos...@needed.invalid> wrote:
> Your curve seems to have a longer rise time than mine,
> almost as if the cooler is being "overwhelmed".
It's a "bug" in SpeedFan - I tried again and this time I added the CPU
> Your curve seems to have a longer rise time than mine,
> almost as if the cooler is being "overwhelmed".
Cores to the graph: They shoot-up to 88C immediately, while the line for
"CPU" slowly slowly slowly rises.
> Heatpipes have a limited thermal pumping capacity.
> If you manage to keep all the working liquid in the
> vapor phase (no chance to condense), then the temperature
> shoots up. The heatpipe has some sort of "watt rating",
> over which it's probably not going to pump heat like it should.
> Each heatpipe has two drops of fluid in it - there isn't
> a couple of ounces of the stuff. Heatpipes can also leak and
> fail, but that hasn't been a problem for years and years.
> Manufacturers have figured out how to make good heatpipes.
> When heatpipes were still a novelty, at least one product
> shipped with no fluid in any of its heatpipes. Making it
> gutless.
laptop the exhaust vent becomes really hot immediately when I run
Prime95, not so on the Noctua cooler in my rig, it just becomes warm'ish
- I mean it's higher than room temp, but I can leave my hand there
forever and never feel the need to remove it because it's too hot...
> You could try a more powerful fan on the cooler.
don't know, Noctua fans are supposed to be good. Oh well, I guess I will
live with it. The only thing that worries me is whether I have bent the
CPU's PCB or not - if not bent then fine, it's how it is. And if I did
bend it, well I'm not buying another one, lol, so I will just accept
that it gets hot under Prime95.
> The other issue with these designs, is "no cowling".
> Mine doesn't have one either. A cowling on the heatsink,
> would ensure that any static pressure the fan can develop,
> goes through the fins.
>
> Since I own a 37.5mm thick fan, if I had to attempt a retrofit,
> I'd do a sheet metal cowling, plus fit the 37.5mm thick fan.
> That's to get as much air velocity as possible through the
> fins. I've only done one cowling/shroud for a CPU cooler here,
> in the time I've worked on computers.
>
> Air cooling goes asymptotic over about 800LFM (linear feet per minute).
> Run of the mill electronics, might use around 200LFM. So there
> is a point, where "mo fan" does "no good".
Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
oO-( )-Oo I need a practice target ... Wesley! -Worf
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
Paul
Jan 26, 2017, 12:59:44 AM1/26/17
to
Likely leaving no choice but to take it apart and verify things.
I agree that the silicon die on your processor, is probably
pretty tiny. (All that FINFET goodness.) Which puts an extra
responsibility on the sole plate of your Noctua. If you had
one of those coolers where the bare pipes touched the processor,
then a couple of the pipes might be carrying all the heat. Yours
has a sole plate, so that gives a bit of spreading. The newer
processors have a thicker lid, which was added for spreading
angle too.
6700K - de-lid
https://hardforum.com/threads/6700k-delid-before-and-after-results.1878870/
Technique demonstrated on a 3770K
https://forums.anandtech.com/threads/bare-die-testing-a-delidded-3770k-an-h100-and-9-different-tims.2285595/
The first one, shows the motherboard measurement tool showing
177W when a bit more than your voltage is applied.
The second article, shows just how thick the IHS is getting
on these processors. It's there to get spreading angle
and make the thermal footprint larger on the lid surface.
For best results, the heatsink should be oriented such
that it had the feet of the pipes going across the
thin dimension. If just one pipe runs along the die, that
would give poor cooler response.
In this picture, you could compare the three screws of the
normal socket mounting hardware, to the orientation of
your setup. I would guess they don't arbitrarily rotate
the die, between 3770K and 6700K :-) But who knows.
http://i272.photobucket.com/albums/jj163/idontcare_photo_bucket/Intel%20Core%20i7-3770K/Bare-die3770kinsocket.jpg
My CPU would be a bit different, as the die is square
on mine. And the lid is soldered down. I believe at the
time, that was part of my buying decision (4930K or 5930K).
Too bad I was such an idiot on motherboard selection, and
am now limited by a gutless VCore.
http://i.imgur.com/ZnOF6xh.jpg
Paul
hanky liu
Jan 26, 2017, 11:26:01 AM1/26/17
to
Have you considered liquid cooling? What mobo are you using, and the build?
B00ze
Jan 26, 2017, 8:21:51 PM1/26/17
to
On 2017-01-26 00:59, Paul <nos...@needed.invalid> wrote:
> B00ze wrote:
>> On 2017-01-25 05:15, Paul <nos...@needed.invalid> wrote:
>>
>>> Your curve seems to have a longer rise time than mine,
>>> almost as if the cooler is being "overwhelmed".
>>
>> It's a "bug" in SpeedFan - I tried again and this time I added the CPU
>> Cores to the graph: They shoot-up to 88C immediately, while the line
>> for "CPU" slowly slowly slowly rises.
>>
>> [snip]
> B00ze wrote:
>> On 2017-01-25 05:15, Paul <nos...@needed.invalid> wrote:
>>
>>> Your curve seems to have a longer rise time than mine,
>>> almost as if the cooler is being "overwhelmed".
>>
>> It's a "bug" in SpeedFan - I tried again and this time I added the CPU
>> Cores to the graph: They shoot-up to 88C immediately, while the line
>> for "CPU" slowly slowly slowly rises.
>>
to Prime95 in normal use.
> I agree that the silicon die on your processor, is probably
> pretty tiny. (All that FINFET goodness.) Which puts an extra
> responsibility on the sole plate of your Noctua. If you had
> one of those coolers where the bare pipes touched the processor,
> then a couple of the pipes might be carrying all the heat. Yours
> has a sole plate, so that gives a bit of spreading. The newer
> processors have a thicker lid, which was added for spreading
> angle too.
>
> 6700K - de-lid
> https://hardforum.com/threads/6700k-delid-before-and-after-results.1878870/
ok, even tho he reaches 1.4V to my 1.360V maximum. On Large FFTs @
1.344v I hover around 67C which is fine. He dropped 20C with the de-lid,
amazing. Btw, I fixed the Watts reading not working by reEnabling SVID;
HWMonitor says I pull 135W @ 1.360V when I hit 92C (nowhere near the
177W that guy is burning).
> Technique demonstrated on a 3770K
> https://forums.anandtech.com/threads/bare-die-testing-a-delidded-3770k-an-h100-and-9-different-tims.2285595/
> The first one, shows the motherboard measurement tool showing
> 177W when a bit more than your voltage is applied.
work.
> The second article, shows just how thick the IHS is getting
> on these processors. It's there to get spreading angle
> and make the thermal footprint larger on the lid surface.
> For best results, the heatsink should be oriented such
> that it had the feet of the pipes going across the
> thin dimension. If just one pipe runs along the die, that
> would give poor cooler response.
> In this picture, you could compare the three screws of the
> normal socket mounting hardware, to the orientation of
> your setup. I would guess they don't arbitrarily rotate
> the die, between 3770K and 6700K :-) But who knows.
>
> http://i272.photobucket.com/albums/jj163/idontcare_photo_bucket/Intel%20Core%20i7-3770K/Bare-die3770kinsocket.jpg
> My CPU would be a bit different, as the die is square
> on mine. And the lid is soldered down. I believe at the
> time, that was part of my buying decision (4930K or 5930K).
> Too bad I was such an idiot on motherboard selection, and
> am now limited by a gutless VCore.
>
> http://i.imgur.com/ZnOF6xh.jpg
>
> Paul
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
oO-( )-Oo If there is no God, who pops up the next kleenex?
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
B00ze
Jan 26, 2017, 8:27:43 PM1/26/17
to
On 2017-01-26 11:26, hanky liu <hank...@gmail.com> wrote:
> Have you considered liquid cooling? What mobo are you using, and the build?
MB is a Maximus VIII Hero. I'm fine with Air - I just wanted to increase
> Have you considered liquid cooling? What mobo are you using, and the build?
the MHz as much as I could on Air and still get a 100% stable system, I
was just surprised at the high temps on some tests. I tried 4.6GHz
without touching my current setup but Prime95 fails; I would have to
increase base voltage from 1.30V to 1.35V and tweak LLC again, but
4.5GHz is fine.
Btw, I installed HandBrake and tried to transcode some 4K video ->
Instant HandBrake crash. So I dialed down to plain vanilla 4.0GHz and
still crash. A few more crashes later, I installed it on the laptop and
tried there, and it still crashes. Works fine for 1080p video but no at
all with a 4K source. Suspect HandBrake problem rather than both my
machines having issues.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
B00ze
Feb 10, 2017, 10:43:06 PM2/10/17
to
On 2017-01-23 20:55, B00ze <B00...@hotmail.com> wrote:
> Good day.
>
> So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
> LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
> wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
> are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
> your overclocked CPU? Because on mine that particular test can make the
> temps climb all the way to 93C. The heat-sink is a really good one
> (NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
> warmer than room temp, and I used good Arctic Silver TIM. While I may
> have placed a tiny tiny bit too much I was real careful with the stuff
> (maybe I didn't put enough?) Under normal testing I have good temps, but
> under that extreme Small FFT Prime95 test, my CPU really cooks - do you
> all get the same temps on air?
Good day Paul, Hanky Liu, all.
> Good day.
>
> So I've overclocked my 6700K to 4.5GHz; this required 1.3V base and with
> LLC=6 it goes up to 1.360V in extreme situations, and this is where I am
> wondering if I screwed-up the heat-sink/termal-paste somehow. What temps
> are you guys seeing when you run Prime95's Small FFT (Maximum Heat) on
> your overclocked CPU? Because on mine that particular test can make the
> temps climb all the way to 93C. The heat-sink is a really good one
> (NOCTUA NH-U12S) but it doesn't even get hot to the touch, it is barely
> warmer than room temp, and I used good Arctic Silver TIM. While I may
> have placed a tiny tiny bit too much I was real careful with the stuff
> (maybe I didn't put enough?) Under normal testing I have good temps, but
> under that extreme Small FFT Prime95 test, my CPU really cooks - do you
> all get the same temps on air?
So I took the plunge - I had to open the desktop and move some hard
drives around, so I took the opportunity to remove the CPU heatsink and
have a look: Good news; I had, as I expected, just a tiny tiny tiny bit
too much TIM and the CPU PCB was NOT bent. Cleaned-up with 100% alcohol
(NOT rubbing alcohol) and re-did the TIM and put the heatsink back
(pretty much with the same amount of TIM). Since this is Arctic Silver
5, I need to wait a couple days for it to settle before I test again,
but I ran OCCT real quick and I get the same temps as before (70-80
degrees under the CPU stress test (NOT the Linpack)). We'll see in a few
days, I will report back.
PS: I had to move a drive because it was heating-up, it's a 6TB with
many platters so it gets hot. Turns out what I thought was a really good
feature of my Thermaltake Urban T81, those two 200mm front fans, is also
a problem: The cylinder in the center of the fans that holds the blades
is really big, and the hard drives that sit at that position do not get
any air because of that. Seems smaller fans are better...
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
Paul
Feb 11, 2017, 1:22:31 AM2/11/17
to
through that drive tray and rack design ?
http://www.thermaltake.com/products-model_gallery.aspx?id=C_00002239
The air from the top 200, it going to squirt over the top rack,
instead of going through it.
http://www.overclockersclub.com/siteimages/articles/thermaltake_urban_t81/t81-21.jpg
And I'm surprised there isn't a noise problem, from the
200mm fan being so close to the metal of the drive racks.
On a high velocity air cooled design at work, when a slot
is not filled in the chassis, we provide a "filler" card.
That's a box that fills the slot, and prevents cooling air from
leaking through the hole. (The air is then forced to go
up through the powered and operational cards.) The I-81
design looks like it's in need of both "dremeling and filling" :-)
Of course, I say that, because I love to mod stuff like
that.
Paul
B00ze
Feb 13, 2017, 10:56:10 PM2/13/17
to
On 2017-02-11 01:22, Paul <nos...@needed.invalid> wrote:
> If I look at image 28 here, how is the air supposed to get
> through that drive tray and rack design ?
>
> http://www.thermaltake.com/products-model_gallery.aspx?id=C_00002239
Well there /are/ holes in the racks, and slits between each drive. But
> If I look at image 28 here, how is the air supposed to get
> through that drive tray and rack design ?
>
> http://www.thermaltake.com/products-model_gallery.aspx?id=C_00002239
don't get me started on that case. In the last 2 pictures on that
webpage, you see cool blue arrows showing incoming air on the side of
the front panel (where the 3 rows of rectangular slots are in the
plastic, on the side of the front panel). Well, air WOULD come into
those slots if the fans were further inside the case, but they are not,
they are further to the front of the case. Those air slots are mostly
useless. Then there is the front door - there is no real way for air to
come into the case when the door is closed (except from the air slots at
the very botton of the front panel) - I am using a bent paper clip to
keep the door ajar. It's a great case because it is huge, but that's
about it; too many stupid design choices all over.
> The air from the top 200, it going to squirt over the top rack,
> instead of going through it.
>
> http://www.overclockersclub.com/siteimages/articles/thermaltake_urban_t81/t81-21.jpg
top (bottom of first HD cage). It is now empty, the 3 drives for the
RAID are all in the 2nd cage and staying cool. That 3rd slot from the
top just so happens to be right smack in the center of the first fan...
> And I'm surprised there isn't a noise problem, from the
> 200mm fan being so close to the metal of the drive racks.
> On a high velocity air cooled design at work, when a slot
> is not filled in the chassis, we provide a "filler" card.
> That's a box that fills the slot, and prevents cooling air from
> leaking through the hole. (The air is then forced to go
> up through the powered and operational cards.) The I-81
> design looks like it's in need of both "dremeling and filling" :-)
> Of course, I say that, because I love to mod stuff like
> that.
case! Looks easy, but there are many variables! I would not buy this
again, even tho I'm somewhat OK with the purchase. I'd look for some
kind of server case instead. Or just give up and get the Cosmos II lol
(but I didn't like the fact that the top of that case is not flat; can't
put USB drives on top).
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
B00ze
Feb 27, 2017, 11:42:44 PM2/27/17
to
On 2017-02-10 22:42, B00ze <B00...@hotmail.com> wrote:
> So I took the plunge - I had to open the desktop and move some hard
> drives around, so I took the opportunity to remove the CPU heatsink and
> have a look: Good news; I had, as I expected, just a tiny tiny tiny bit
> too much TIM and the CPU PCB was NOT bent. Cleaned-up with 100% alcohol
> (NOT rubbing alcohol) and re-did the TIM and put the heatsink back
> (pretty much with the same amount of TIM). Since this is Arctic Silver
> 5, I need to wait a couple days for it to settle before I test again,
> but I ran OCCT real quick and I get the same temps as before (70-80
> degrees under the CPU stress test (NOT the Linpack)). We'll see in a few
> days, I will report back.
Alright, I tested the PC with Prime95 again, after having redone the TIM
> So I took the plunge - I had to open the desktop and move some hard
> drives around, so I took the opportunity to remove the CPU heatsink and
> have a look: Good news; I had, as I expected, just a tiny tiny tiny bit
> too much TIM and the CPU PCB was NOT bent. Cleaned-up with 100% alcohol
> (NOT rubbing alcohol) and re-did the TIM and put the heatsink back
> (pretty much with the same amount of TIM). Since this is Arctic Silver
> 5, I need to wait a couple days for it to settle before I test again,
> but I ran OCCT real quick and I get the same temps as before (70-80
> degrees under the CPU stress test (NOT the Linpack)). We'll see in a few
> days, I will report back.
and reseated the CPU -> Same results; Prime95 (the latest version) can
make some cores climb to 94C for some tests; we're still hovering at
around 80C in OCCT's CPU test. One thing that HAS changed is that OCCT
crashes the first time I run it and start the CPU test. Run it again, no
matter how many times you like, and it works fine, but start it for the
first time that day and it crashes. So the situation is actually worse
than before. I'm not going to take the cooler off and reseat the CPU
again tho, unless I find some other symptoms/programs crashing
(Thunderbird did start crashing randomly last week but I re-installed it
and it's been doing fine since). I will re-install OCCT and see what
that does; maybe I will have to increase base voltage from 1.30 to 1.35
to fix this; I shoulda left well enough alone, it was rock solid before!
Anyway, except for stability issues, I'm giving up on this strange
cpu/heatsink combo, where the HS base doesn't really get hot even when
the CPU is running 90C (it becomes "warm" only). Looks like it is how
things are - unless I run a water cooler, I must expect very high temps
in Prime95.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
~misfit~
Apr 6, 2017, 11:03:42 PM4/6/17
to
sounds to me like the factory TIM between the CPU proper and its IHS
(Integrated Heat Spreader) could be faulty. Apparently it's not that
uncommon a thing...
--
Shaun.
"Humans will have advanced a long, long way when religious belief has a cozy
little classification in the DSM*."
David Melville (in r.a.s.f1)
(*Diagnostic and Statistical Manual of Mental Disorders)
B00ze
Apr 7, 2017, 10:52:06 PM4/7/17
to
On 2017-04-06 23:03, ~misfit~ <shaun.at...@gmail.com> wrote:
>> Alright, I tested the PC with Prime95 again, after having redone the
>> TIM and reseated the CPU -> Same results; Prime95 (the latest
>> version) can make some cores climb to 94C for some tests; we're still
>> hovering at around 80C in OCCT's CPU test. One thing that HAS changed
>> is that OCCT crashes the first time I run it and start the CPU test.
>> Run it again, no matter how many times you like, and it works fine,
>> but start it for the first time that day and it crashes. So the
>> situation is actually worse than before. [...]
>> Alright, I tested the PC with Prime95 again, after having redone the
>> TIM and reseated the CPU -> Same results; Prime95 (the latest
>> version) can make some cores climb to 94C for some tests; we're still
>> hovering at around 80C in OCCT's CPU test. One thing that HAS changed
>> is that OCCT crashes the first time I run it and start the CPU test.
>> Run it again, no matter how many times you like, and it works fine,
>> but start it for the first time that day and it crashes. So the
>
> I haven't been following your progress but from that last bit you wrote it
> sounds to me like the factory TIM between the CPU proper and its IHS
> (Integrated Heat Spreader) could be faulty. Apparently it's not that
> uncommon a thing...
Quite possible. For CERTAIN de-lidding would make this CPU run much
> I haven't been following your progress but from that last bit you wrote it
> sounds to me like the factory TIM between the CPU proper and its IHS
> (Integrated Heat Spreader) could be faulty. Apparently it's not that
> uncommon a thing...
cooler, but I'm not going there lol. Thunderbird crashed again upon
starting by the way, so I now had a few crashes of applications "upon
starting" which told me LLC was not boosting voltage fast enough. So I
changed by base voltage from 1.300 -> 1.310 (and kept LLC @ Level 6).
This has the effect that unless the CPU is doing absolutely nothing (and
running Windows is /something/) then it runs at 1.328v (instead of
1.312v as before). So far so good. The downside of course is that if I
run Prime95 v28 small FFT I now see a Max Temp of 97C because the volts
boost-up to 1.360v. I installed the recommended Prime 95 (v26) and I
stay around 80C on small FFT's - v28 uses FMA3/Avx and together with a
high LLC it means a lot of energy is used. I tried LLC @ 5 but then LLC
*lowers* the voltage when the CPU gets really busy. That's what LLC is
supposed to be for actually, but it doesn't work when you overclock.
I agree with you my CPU sample isn't great. The best overclock I get
without boosting volts too much, like I am doing now, is 4400MHz. At
that speed all I need is 1.260v base volts (an increase of 0.010v) and
LLC @ 7. Getting 4500MHz stable requires a lot more power.
If I become unstable as the years go by, or if I discover that I can't
encode a movie or play a game without the CPU climbing to 95C then I
will just fall back to 4400MHz.
Best Regards,
--
! _\|/_ Sylvain / B00...@hotmail.com
! (o o) Member:David-Suzuki-Fdn/EFF/Red+Cross/SPCA/Planetary-Society
0 new messages