Wow, Heat Issue with Ivy just stinks, what to do now?

I recommend just undervolting / underclocking over purchasing a factory underclocked / undervolted processors. However some users will prefer the guaranteed underclock / undervolt and will pay more for that.
3770K MSRP 313$, 3770/3770S/3770T 278$ :rolleyes:

I'm unsure of how Turbo mode with these chips works when underclocked/overclocked. Does Turbo still push it to the normal turbo speeds, is it disabled, or does it increase it by a specific increment above the underclocked speed?
I haven't bought a new processor to play around with in some years so haven't had a reason to look into this. I maybe be in the market soon though so I'm honestly curious. I suppose I could google around.

3770T for example has normal clocks of 2.5Ghz and a maximum Turbo of 3.7Ghz.
If needed, the processor will operate at a higher clockspeed depending on four factors that I have included on the bottom of this post.
*snip*
But atleast with 2500K/2600K it was:
3.3/3.4Ghz base speed
four cores loaded -> +1bin over base aka 3.4/3.5Ghz
three cores loaded->+2bins over base aka 3.5/3.6Ghz
two cores loaded->+3bins over base aka 3.6/3.7Ghz
one core loaded->+4bins over base aka 3.7/3.8Ghz

You can read more here.
http://www.intel.com/content/www/us/en/architecture-and-technology/turbo-boost/turbo-boost-technology.html
http://en.wikipedia.org/wiki/Intel_Turbo_Boost
Code:
"Intel Turbo Boost Technology 2.0 is activated when the Operating System
(OS) requests the highest processor performance state (P0).

The maximum frequency of Intel Turbo Boost Technology 2.0 is dependent on the
number of active cores. The amount of time the processor spends in the Intel Turbo
Boost Technology 2.0 state depends on the workload and operating environment.

Any of the following can set the upper limit of Intel Turbo Boost Technology 2.0 on a given workload:
*Number of active cores
*Estimated current consumption
*Estimated power consumption
*Processor temperature

^When the processor is operating below these limits and the user's workload demands
additional performance, the processor frequency will dynamically increase until the upper
limit of frequency is reached. Intel Turbo Boost Technology 2.0 has multiple algorithms
operating in parallel to manage current, power, and temperature to maximize
performance and energy efficiency. Note: Intel Turbo Boost Technology 2.0 allows
the processor to operate at a power level that is higher than its rated upper power limit
(TDP) for short durations to maximize performance."
Turbo boost really is a dream come true for laptops and other low power rigs :D

E:
Well I found an older review of 2500T (only I could find with turbo states actually told..) and the Turbo works like this on the 2500T
table3.png

So with the 3770T it would work something like this (just guessing)
1-2 threads loaded 3.7
3-4 threads loaded 3.5-3.6
5-6 threads loaded 3.0-3.1
7-8 threads loaded 2.6-2.7Ghz
 
Last edited:
3770K MSRP 313$, 3770/3770S/3770T 278$ :rolleyes:



3770T for example has normal clocks of 2.5Ghz and a maximum Turbo of 3.7Ghz.
If needed, the processor will operate at a higher clockspeed depending on four factors that I have included on the bottom of this post.
*snip*
But atleast with 2500K/2600K it was:
3.3/3.4Ghz base speed
four cores loaded -> +1bin over base aka 3.4/3.5Ghz
three cores loaded->+2bins over base aka 3.5/3.6Ghz
two cores loaded->+3bins over base aka 3.6/3.7Ghz
one core loaded->+4bins over base aka 3.7/3.8Ghz

You can read more here.
http://www.intel.com/content/www/us/en/architecture-and-technology/turbo-boost/turbo-boost-technology.html
http://en.wikipedia.org/wiki/Intel_Turbo_Boost
Code:
"Intel Turbo Boost Technology 2.0 is activated when the Operating System
(OS) requests the highest processor performance state (P0).

The maximum frequency of Intel Turbo Boost Technology 2.0 is dependent on the
number of active cores. The amount of time the processor spends in the Intel Turbo
Boost Technology 2.0 state depends on the workload and operating environment.

Any of the following can set the upper limit of Intel Turbo Boost Technology 2.0 on a given workload:
*Number of active cores
*Estimated current consumption
*Estimated power consumption
*Processor temperature

^When the processor is operating below these limits and the user's workload demands
additional performance, the processor frequency will dynamically increase until the upper
limit of frequency is reached. Intel Turbo Boost Technology 2.0 has multiple algorithms
operating in parallel to manage current, power, and temperature to maximize
performance and energy efficiency. Note: Intel Turbo Boost Technology 2.0 allows
the processor to operate at a power level that is higher than its rated upper power limit
(TDP) for short durations to maximize performance."
Turbo boost really is a dream come true for laptops and other low power rigs :D

E:
Well I found an older review of 2500T (only I could find with turbo states actually told..) and the Turbo works like this on the 2500T
table3.png

So with the 3770T it would work something like this (just guessing)
1-2 threads loaded 3.7
3-4 threads loaded 3.5-3.6
5-6 threads loaded 3.0-3.1
7-8 threads loaded 2.6-2.7Ghz

more numbers to process, where is the genius pelo? help with this thread please!
 
temps will be hotter than SB due to smaller die area to disperse power over. think about it if you disburse 100W over one inches square and 100W over 1/2 inches square. Just an example but you get the idea that it's harder to dissipate same power over smaller areas. No biggie. Just keep it less than 90 or so.

I like your setup. Must be really silent and that's whole reason I watercool too. fans like 600-800 and you hear nothing.
 
temps will be hotter than SB due to smaller die area to disperse power over. think about it if you disburse 100W over one inches square and 100W over 1/2 inches square. Just an example but you get the idea that it's harder to dissipate same power over smaller areas. No biggie. Just keep it less than 90 or so.

I like your setup. Must be really silent and that's whole reason I watercool too. fans like 600-800 and you hear nothing.

I am so close to pulling the trigger on an XSPC single loop, know the owner of Performance PCS local and if the H100 can't handle the heat of Ivy Bridge, that could be the next step, thanks man...
 
I am so close to pulling the trigger on an XSPC single loop, know the owner of Performance PCS local and if the H100 can't handle the heat of Ivy Bridge, that could be the next step, thanks man...

Since the H100 can handle the 3930K overclocked, Ivybridge will be a cakewalk.
 
A buddy of mine who kind of follows this stuff knew IB was just released and asked what the "current" best OC CPU is, as he went to SB 2500K when it was released and thought maybe it was time to upgrade.

When I reported "his current" CPU was best bang for buck with OC, he just laughed....
 
The cables were looking pretty shabby, cleaned them up a little, not finalizing until the new mainboard arrives which now looks like it will be a hard wait, since Asus has not yet announced the ROG Maximus 5 Extreme yet. The Maximus 5 Gene is out, the Maximus 5 Formula will be out in May and JJ from Asus has mentioned in two videos about two boards coming out with native thunderbolt connectivity, not the Asus add-in card, and one was announced today the P8Z77-V Premium which is a combination of the P8 Deluxe and the P8 Workstation. The other sku he said was on the ROG side, so it has to be the M5E. The ROG boards being black and red, I have for months been sort of building around that mobo and a new processor while waiting and waiting, again the only downside to loving this hobby, is the perpetual waiting. Please leave any suggestions for my build that you feel would be helpful and Thanks for stopping by.


IMG_0678.jpg
 
Last edited:
on a side note, a set of Black Vengeance mixed with the Red Vengeance would look kick ass!
 
3770/3770S/3770T 278$

In the past the "Energy Efficient" models used to cost more although I admit that was when they were basically a top of the line chip running at a lower frequency to achieve a lower maximum heat output.
 
on a side note, a set of Black Vengeance mixed with the Red Vengeance would look kick ass!

Awesome man!

Thanks for all the advice so far, pretty sure going with the 3770K instead of the 2700K (unless we hear some more information when the retail versions hit the streets next week) thanks for all the help, and I hope it helped others decide also. Opened an Asus ROG Maximus 5 Extreme thread over in intel motherboards, since it looks like I will be waiting for a while on the board to be available. I still need build suggestions, maybe turn the AX 850 over to expose the fan, or a new triple-slot video card and two more sticks of Vengeance memory for more "red" I dunno see how it progresses, gonna be fun.

I posted a really informative Asus video over there it’s about an hour and a half, worth watching.
 
I'm in a real similar pickle to you MeanBruce. I've been waiting for the Ivy Bridge to come out so I can upgrade from my Core 2 Duo. I was looking at the i5 3570k. Now that they're talking about high temps it's got me concerned and wondering if I should go down to the 2500k. I would probably try overclocking it but nothing beyond 4GHz. I'm not comfortable messing with voltages. I'm currently planning to get a fairly cheap but well-performing CPU cooler, the Cooler Master 212 Evo.
 
I'm in a real similar pickle to you MeanBruce. I've been waiting for the Ivy Bridge to come out so I can upgrade from my Core 2 Duo. I was looking at the i5 3570k. Now that they're talking about high temps it's got me concerned and wondering if I should go down to the 2500k. I would probably try overclocking it but nothing beyond 4GHz. I'm not comfortable messing with voltages. I'm currently planning to get a fairly cheap but well-performing CPU cooler, the Cooler Master 212 Evo.

At that low overclock just get the 3570K. It appears that as long as you don't increase the voltage the temps on Ivy don't go crazy, and it appears you can get up to around 4.4 or 4.5 without much of a voltage bump. It is above that where things start to go crazy.
 
At that low overclock just get the 3570K. It appears that as long as you don't increase the voltage the temps on Ivy don't go crazy, and it appears you can get up to around 4.4 or 4.5 without much of a voltage bump. It is above that where things start to go crazy.

I'm going with the same i5-3570K (not i5-2500K) due to the planned lower overclocking (at the same planned nice round 4 GHz OC, I'll have a higher IPC, with lower wattage, compared to the 2500K). The only issues are twofold - the wait (can't buy until Monday at the earliest, despite two MCs near, due to mass-transit issues) and possible pricing shenanigans (no idea what MC's price difference will be between Sandy and Ivy).
 
@MeanBruce Man what have you done to your AX850? The cable management looks awesome. I've been having so much trouble with the sleeved cables being stiff and I think this might be the solution I need. So each individual wire is red once you remove the sleeves or did you end up painting them or something?
 
I'm going with the same i5-3570K (not i5-2500K) due to the planned lower overclocking (at the same planned nice round 4 GHz OC, I'll have a higher IPC, with lower wattage, compared to the 2500K). The only issues are twofold - the wait (can't buy until Monday at the earliest, despite two MCs near, due to mass-transit issues) and possible pricing shenanigans (no idea what MC's price difference will be between Sandy and Ivy).
Yeah, I've been reading that AnandTech article on the voltages and overclocking and have come to the same conclusion. Going with the 3570k. I hope being an early adopter of the z77 chipset and ivy bridge processor doesn't come with any major problems though.
 
I'm in a real similar pickle to you MeanBruce. I've been waiting for the Ivy Bridge to come out so I can upgrade from my Core 2 Duo. I was looking at the i5 3570k. Now that they're talking about high temps it's got me concerned and wondering if I should go down to the 2500k. I would probably try overclocking it but nothing beyond 4GHz. I'm not comfortable messing with voltages. I'm currently planning to get a fairly cheap but well-performing CPU cooler, the Cooler Master 212 Evo.

Now they are talking about the Thermal Paste on or within the Ivy chips being at fault. So let's hope that is a simple thermal solution changing paste from the engineering samples to the retail chips.

Hey man, join me over in the intel motherboards thread, working on my ROG build and need advice. That Cooler Master 212 Evo is an amazing cooler, I don't know how CM can make such efficient coolers at the price points, engineering I guess. Like you I am still on the fence with SB or IB thinking next week when the retail versions are tested real world we might I am so hoping get some much better thermals overall, stock and OC'd. :)
 
@MeanBruce Man what have you done to your AX850? The cable management looks awesome. I've been having so much trouble with the sleeved cables being stiff and I think this might be the solution I need. So each individual wire is red once you remove the sleeves or did you end up painting them or something?

Have had the Corsair cables two weeks they look great Corsair matches the wire color beneath with the sleeving on top, but so difficult to manage, crazy stiff, I hung them on a wire shelf for a week hoping gravity would kick in, it didn't. I ended up just using clean hands and bending all the kinks out, then when you install them its more bending, and oops closing the rear panel pushes them forward and separates them so more bending. They are not easy to position but when you get them right, they are sweet.:)
 
Last edited:
I doubt it really makes that much difference. A couple of degrees, sure, but these things run 20C hotter than SB when pushed, and I don't think thermal paste explains away that difference.

should the improved GPU in Ivy cause the extra heat?

i thought everyone said that when Ivy was announced to have a improved gpu.
 
A solder attach could have a heat conductivity in the range of 80 W/mK. A TIM paste could have a heat conductivity in the range of 5 W/mK. That’s your problem right there! Note that these values are not exact, as we don’t know the exact heat conductivity of Intel’s “Secret sauce”. However, these are values representative of solder or TIM paste, and there is a giant gap between how TIM paste and solder perform in regards to conducting heat. They are in different leagues...Most importantly here, if Intel is using TIM paste between the IHS and CPU die, the IHS effectively becomes a heat barrier rather than a heat spreader.
http://www.overclockers.com/ivy-bridge-temperatures
 
Interesting overclocking results here - don't really track with what other sites are reporting.

Using the automatic overclocking features of the Asus P8Z77-V Deluxe motherboard, we were able to easily overclock our test chip to 4.63GHz using the Intel XTS100H air cooler, while maintaining solid system stability. During our initial stability testing, running the processor at 100 percent load for over 10 minutes, the chip’s temperature peaked at 65 degrees Celsius. Compare that with last year’s test of the Core i7-2600K; that chip's temperature reached 81 degrees with the same cooler while overclocked to 4.5GHz.

http://computershopper.com/components/reviews/intel-core-i7-3770k/(page)/4#review-body

I'm very curious to see what happens when the chips release next week.
 
I doubt it really makes that much difference. A couple of degrees, sure, but these things run 20C hotter than SB when pushed, and I don't think thermal paste explains away that difference.

For what its worth, this wouldn't be just the difference between a high quality TIM and low quality TIM. Fluxless solder is in an entirely different league of heat transfer from regular paste/grease so while 20C might be a liberal estimate the change would probably be more than a couple degrees.
 
For what its worth, this wouldn't be just the difference between a high quality TIM and low quality TIM. Fluxless solder is in an entirely different league of heat transfer from regular paste/grease so while 20C might be a liberal estimate the change would probably be more than a couple degrees.

Yeah, valid point. I still don't believe that this is coming as a surprise to Intel though, which leads me to believe that either this isn't the final solution, or they have some other reason for not soldering the chips.
 
Yeah, valid point. I still don't believe that this is coming as a surprise to Intel though, which leads me to believe that either this isn't the final solution, or they have some other reason for not soldering the chips.

The CPUs sent review sites are Engineering Samples (ES) - thus not packed (in terms of the CPU itself) the same way retail CPUs are (though they may come in retail-type packaging, or even retail packaging). Wasn't the same, in fact, true with early Sandy Bridge review CPUs (also of the ES sort) and due to the same switchology (TIM was used in early SB ES CPUs, but solder was used in production)? The only retail Ivies to show up anywhere are in Europe - not the US; the closer Selling Sunday gets, the more nerve-wracking it gets.

(I don't know if the same is true with Kyle; he stated that when possible, he buys reviewed products from a retailer - in the case of [H], MicroCenter - like ordinary folks - to avoid the *cherry-picked* issue; however, was that, in fact, even possible with Ivy Bridge?)
 
What if next weeks retail chips run ice cold, oh please Lawd, I can't take any more bad news about Ivy!:(
 
The CPUs sent review sites are Engineering Samples (ES) - thus not packed (in terms of the CPU itself) the same way retail CPUs are (though they may come in retail-type packaging, or even retail packaging). Wasn't the same, in fact, true with early Sandy Bridge review CPUs (also of the ES sort) and due to the same switchology (TIM was used in early SB ES CPUs, but solder was used in production)? The only retail Ivies to show up anywhere are in Europe - not the US; the closer Selling Sunday gets, the more nerve-wracking it gets.

(I don't know if the same is true with Kyle; he stated that when possible, he buys reviewed products from a retailer - in the case of [H], MicroCenter - like ordinary folks - to avoid the *cherry-picked* issue; however, was that, in fact, even possible with Ivy Bridge?)

Some of the review CPUs are ES but some are not, look at the CPUz screenshots, it was about 50/50 from the 15ish reviews I looked at. No difference in average overclock speed/voltage either, 4.5-4.7 1.2-1.35v temps peaking 75-85c on core.
 
The CPUs sent review sites are Engineering Samples (ES) - thus not packed (in terms of the CPU itself) the same way retail CPUs are (though they may come in retail-type packaging, or even retail packaging). Wasn't the same, in fact, true with early Sandy Bridge review CPUs (also of the ES sort) and due to the same switchology (TIM was used in early SB ES CPUs, but solder was used in production)? The only retail Ivies to show up anywhere are in Europe - not the US; the closer Selling Sunday gets, the more nerve-wracking it gets.

(I don't know if the same is true with Kyle; he stated that when possible, he buys reviewed products from a retailer - in the case of [H], MicroCenter - like ordinary folks - to avoid the *cherry-picked* issue; however, was that, in fact, even possible with Ivy Bridge?)

People are beginning to purchase the Ivys now, maybe as Selling Sunday begins [H]ard members can post their thermal results... A new thread would make the information more readily accessible.
 
Z77 + 2600k/2700k, good processors with frickin solder. You know Intel, that stuff you used to use before you tried to save a damn penny or two. Jerks.

When do the mobos with thunderbolt arrive?
 
Last edited:
I am REALY shocked How hot these things get and they HIT awall at 5ghz like no ones buiss. WOW. Less Silicon more heat. Im reinstaling my 2600k.
 
Z77 + 2600k/2700k, good processors with frickin solder. You know Intel, that stuff you used to use before you tried to save a damn penny or two. Jerks.

When do the mobos with thunderbolt arrive?

Asus has about 9 Z77 boards out now with a Thunderbolt header on the lower left of the boards, you can buy an add-in card from Asus for $30 when you are ready to upgrade...

And they are also producing two mobos with thunderbolt native on the board a P8 Premium board and a ROG board.
 
Z77 + 2600k/2700k, good processors with frickin solder. You know Intel, that stuff you used to use before you tried to save a damn penny or two. Jerks.

When do the mobos with thunderbolt arrive?

I may go 2700K also still hoping better news about 3770K will arrive tomorrow and next week.:)
 
Z77 + 2600k/2700k, good processors with frickin solder. You know Intel, that stuff you used to use before you tried to save a damn penny or two. Jerks.

When do the mobos with thunderbolt arrive?

Why are you so upset about a change that doesnt seem to have made ANY appreciable difference? Relax, it's not like Intel raped your mom.
 
in the end i can live without OCing the CPU , honestly i dont give a flying shit bout it. What i do care about however is the lifetime of Ivy, if it serves me 4-5 years with its new TIM architecture its ok for me , if not id be mad cuz i waited a shit long time for this CPU.
 
Back
Top