Sign in with
Sign up | Sign in
Your question

Ivy-Bridge Processors Seem to run hotter than SB

Last response: in CPUs
Share
March 15, 2012 10:28:39 AM

I have been waiting for the Ivy-Bridge processors to come out, and have been looking around on what temps they get.

Couple of people got their hands on one early, and they are getting 100 degrees.
If this is the case im going to stick with getting a Core i5 2500k

Thoughts on this ?
a b à CPUs
March 15, 2012 10:47:25 AM

matt2155 said:
I have been waiting for the Ivy-Bridge processors to come out, and have been looking around on what temps they get.

Couple of people got their hands on one early, and they are getting 100 degrees.
If this is the case im going to stick with getting a Core i5 2500k

Thoughts on this ?


Until I see a review or have one in my greedy little paws it is still speculation and you can not make an imformed decision on that. ;) 
m
0
l
March 15, 2012 11:00:44 AM

arthurh said:
Until I see a review or have one in my greedy little paws it is still speculation and you can not make an imformed decision on that. ;) 


Yeah thats true.

Will 2500k's be discontinued straight away after the release ?
m
0
l
Related resources
a b à CPUs
March 15, 2012 11:06:56 AM

Quote:
Will 2500k's be discontinued straight away after the release ?


Hard to tell til we actually see what Intel releases. Everyone assumes it will be unlocked, but maybe it won't. Either way we still have to wait and see.
m
0
l
a c 82 à CPUs
March 15, 2012 12:59:24 PM

i've been thinking about this, SB is a 95W part with a 216mm2 die = 0.44W/mm2 of heat dissipation. IVB is 77W with 160mms die = 0.48W/mm2 of heat dissipation.

I'm not sure how that transfers into deg C but there needs to be a higher heat flux between the chip and the package and the package and the heatsink (although this secondary stage is mitigated by the heat spreading across the die. I would therefore suggest that the contact between the die and the package is having to carry a bigger heat flow, and it may not be as capable of this, hence a greater temp rise, until that temp delta increases the efficiency of the heat flow.
m
0
l
a b à CPUs
March 15, 2012 1:05:31 PM

aqe040466 said:
In Anandtech Forum, they have this Engineering sample of IB CPU and they run and tested it is hotter then SB CPU.
http://forums.anandtech.com/showthread.php?t=2231710


Good find there but they actually provide a LINK to xtremesystems.org in which you have to be a member to read. So not exactly open to the ramdon public for comment.

I'm sticking to my earlier statement that we need a open review of the retail product to make any imformed decision. :sol: 
m
0
l
March 15, 2012 1:06:31 PM

Really disappointed about Ivy-Bridge from the things I have heard so far.
I am going to wait until it comes out first, but most likely I will go with 2500k

I will save money, because the Z77 chipset motherboards would not be as cheap as the Z68 motherboards aswell.

I would prefer a much cooler cpu than a 10% performance boost
m
0
l
a b à CPUs
March 15, 2012 1:15:48 PM

I am using a P67 M/BD and one of my BIOS updates allow the IB to run on my M/BD. I know my M/BD doesn't support Gen 3 but if that is the only thing that might not hold me back from trying it out just for OCing potientail. My hobbie. :sol: 
m
0
l
a c 341 à CPUs
March 15, 2012 1:20:25 PM

You have to remember that the report was on an engineering sample that was highly overclocked.
I saw no details on the cooler involved.

It the planned build time is anywhere near launch date, it will pay to wait and see.

I do not expect the 2500K to be discontinued for a while. It needs to be there for the replacement market.
Accordingly, do not expect any big price drops at retail. Intel does not seem to do that. Price drops on the used market are likely to reach the price/performance curve of ivy bridge.
m
0
l
March 15, 2012 1:20:49 PM

arthurh said:
I am using a P67 M/BD and one of my BIOS updates allow the IB to run on my M/BD. I know my M/BD doesn't support Gen 3 but if that is the only thing that might not hold me back from trying it out just for OCing potientail. My hobbie. :sol: 


yeah truee

I can't afford though buying a Ivy-Bridge, and testing it out, and if it is running hot, then I cannot take back and I wasted $235
And I have no job, and I am on a budget, so thats even worse lol
m
0
l
a b à CPUs
March 15, 2012 1:30:17 PM

The speculation about the discontinuance of the 2500K seems to me to be very misguided in this thread. The 2500K is Intel's biggest seller in the gamer/enthusiast market, and one of their biggest sellers over all. It would be killing the Goose that lays the golden eggs. BTW the profit margin on the 2500K have either stayed the same or improved since its introduction.

I would speculate that Intel is more likely to kill off the i7 2600K or 2700K before the 2500K. I doubt that the 2500K will be killed off before '15 or '16, and Intel will reduce profit margin at the end of the life cycle.
m
0
l
March 15, 2012 1:30:42 PM

geofelt said:
You have to remember that the report was on an engineering sample that was highly overclocked.
I saw no details on the cooler involved.

It the planned build time is anywhere near launch date, it will pay to wait and see.

I do not expect the 2500K to be discontinued for a while. It needs to be there for the replacement market.
Accordingly, do not expect any big price drops at retail. Intel does not seem to do that. Price drops on the used market are likely to reach the price/performance curve of ivy bridge.


Intel are too money hungry, infact this is what I think they are going to do.
I personally think that IB will be hotter than SB, but I think they will bring out IB-E later on, to get the money out of us. Just like when they had SB and then after a while they released SB-E.

I hate intel
m
0
l
March 15, 2012 1:34:18 PM

Also, motherboards are not going to be futureproof, with DDR4 memory coming out next year
m
0
l
a c 82 à CPUs
March 15, 2012 1:48:34 PM

13thmonkey said:
i've been thinking about this, SB is a 95W part with a 216mm2 die = 0.44W/mm2 of heat dissipation. IVB is 77W with 160mms die = 0.48W/mm2 of heat dissipation.

I'm not sure how that transfers into deg C but there needs to be a higher heat flux between the chip and the package and the package and the heatsink (although this secondary stage is mitigated by the heat spreading across the die. I would therefore suggest that the contact between the die and the package is having to carry a bigger heat flow, and it may not be as capable of this, hence a greater temp rise, until that temp delta increases the efficiency of the heat flow.



looks like the anand people are thinking about die size too.

but their thoughts on the heat sink being better are probably not true, the only thing that the HSF has to do is maintain as a big a delta T between the package and the sink as possible. This has to be bigger than normal because of the greater W/mm2, but the heatsink itself is not near capacity. Its just can it cannot work as efficiently.
m
0
l
a b à CPUs
March 15, 2012 1:56:32 PM

matt2155 said:
Also, motherboards are not going to be futureproof, with DDR4 memory coming out next year


we (the consumer) will never be in front. Like a dog chasing its tail, albeit from a spectator’s view the dog seems to be having so much fun ^_^

Just spend what/when you can and enjoy the upgrades and the opportunities of handling components (:
m
0
l
a c 87 à CPUs
March 15, 2012 2:06:03 PM

Also keep in mind we/they are dealing with engineering samples. They run, but not always well. Could be a big difference between the ES and retail chips.
m
0
l
a c 127 à CPUs
March 15, 2012 3:28:29 PM

4745454b said:
Also keep in mind we/they are dealing with engineering samples. They run, but not always well. Could be a big difference between the ES and retail chips.


Most ES samples are great for overclocking as they can handle the higher voltages.

That said, they always have issues with them which is why any reviews right now cannot be taken at 100%.

When Intel/AMD sends out a press kit to the review sites, thats when we take it at 100% because those are the same stepping we will see and as well all the drivers and such are proper. I don't imagine IB will run much hotter than SB, but it could as it has less die space to spread the heat out.
m
0
l
a c 87 à CPUs
March 15, 2012 3:30:02 PM

Never said or meant to say that they couldn't clock. I meant that as an ES they aren't a good indicator for retail heat output. (other then maybe the K series CPUs.)
m
0
l
March 15, 2012 4:27:07 PM

matt2155 said:
Intel are too money hungry, infact this is what I think they are going to do.
I personally think that IB will be hotter than SB, but I think they will bring out IB-E later on, to get the money out of us. Just like when they had SB and then after a while they released SB-E.

I hate intel


Does it matter? They are not going to sell a processor that cooks itself out of the box. Unless your room is where you house your arctic penguin, I don't see what the issue is. And then the pointless rant about this being hotter, and Intel trying to get money out of you, and now you hate intel? WTF? So Intel is supposed to make a chip that overclocks for you, cheaper, because you don't have a job and want to buy a new pc? Are you even overclocking this computer you haven't bought yet? Please don't troll midway through your discussion.

You should have started off with, Hey, I'm planning on getting a new PC soon, and I'm wondering about the overclocking potential of a 2500K against IVB. To which we would have replied, we don't know yet, there are only sparse reports from people who have ES samples, so no one truly knows. But someone supposedly got one up to 7Ghz on liquid cooling.
m
0
l
a c 479 à CPUs
March 15, 2012 5:08:16 PM

chesteracorgi said:
The speculation about the discontinuance of the 2500K seems to me to be very misguided in this thread. The 2500K is Intel's biggest seller in the gamer/enthusiast market, and one of their biggest sellers over all. It would be killing the Goose that lays the golden eggs. BTW the profit margin on the 2500K have either stayed the same or improved since its introduction.

I would speculate that Intel is more likely to kill off the i7 2600K or 2700K before the 2500K. I doubt that the 2500K will be killed off before '15 or '16, and Intel will reduce profit margin at the end of the life cycle.


There really is no need for Intel to continue to produce Sandy Bridge CPUs once Ivy Bridge CPUs have been released. Running a production line takes money continuing to manufacture Sandy Bridge CPU can cannibalize sales of Ivy Bridge CPUs. The right thing to do is to cease all production of Sandy Bridge CPUs, then retool the fabs for future CPU production.
m
0
l
a c 479 à CPUs
March 15, 2012 5:26:49 PM

matt2155 said:
Intel are too money hungry, infact this is what I think they are going to do.
I personally think that IB will be hotter than SB, but I think they will bring out IB-E later on, to get the money out of us. Just like when they had SB and then after a while they released SB-E.

I hate intel


Not everyone needs a 6 core CPU. If you feel that whatever programs you use will benefit from 6 cores, then that's different.

Based on benchmarks that I remember seeing, the only game the really benefited from 6 cores (i7-3960X vs. i7-2600k) was Civilization 5. So games gain little benefit from more than 4 cores. The Sandy Bridge-E series are for the enthusiasts; it's not mainstream. Therefore, if you are an enthusiast and you have the money to spare, then you buy it. If you consider yourself an enthusiast, but you are not willing to spend the money, then perhaps you are not a "true enthusiast" who's willing to pay the price the get the fastest CPU around.

Do you hate Lamborghini, because you cannot afford to buy their cars?

I suppose I have been reading too many rants, about too many different things lately (not just IT related), therefore my suggestion is to switch over to AMD so that you can buy an 8 core CPU at a fraction of what Intel charges for their enthusiast level CPUs. Intel won't miss you and AMD will probably not notice you.

m
0
l
a b à CPUs
March 15, 2012 5:32:34 PM

Idk, i doubt Intel would release a CPU that they didn't test them selves first for heat Issues or power issues. Kinda seems sketchy
m
0
l
March 15, 2012 5:44:44 PM

And they will get more profit by selling Ivy variants because they are smaller!
I am guite sure that heat will not be big problem if you just don't puch Ivy too far. With conservative clocks, Ivy will be just fine!
m
0
l
a b à CPUs
March 15, 2012 8:18:21 PM

matt2155 said:
Yeah thats true.

Will 2500k's be discontinued straight away after the release ?


I doubt it. From what I've heard, every Ivy Bridge that Intel can produce is already sold to an OEM or retailer. This is going to be as big if not bigger than Sandy Bridge was last year. So Intel will keep the 32nm fab lines open until sales drop as more 22nm volume comes on line. Which could mean problems getting one of the top bin Ivy's in May or June..

m
0
l
a b à CPUs
March 15, 2012 8:39:33 PM

matt2155 said:
Intel are too money hungry, infact this is what I think they are going to do.
I personally think that IB will be hotter than SB, but I think they will bring out IB-E later on, to get the money out of us. Just like when they had SB and then after a while they released SB-E.

I hate intel



You hate Intel but you are using Intel platforms. We can't do anything about it, we are the consumers we are bound to pay(Intel products) what we want. Since the first time I used a PC it has been Intel all thruoghout way back in the 90's e.g. Intel 386, 485, 586 tyhen came PENTIUM 1, 2, 3, 4 and the rest is history and until now still Intel, maybe I'm an Intel fanboy?
m
0
l
a b à CPUs
March 15, 2012 9:29:51 PM

read the page

http://news.techeye.net/chips/intels-ivy-bridge-specs-l...

the 100oC is under a 4.8ghz overclock

the reason it is hotter than normal is it is drawinf more current

basic theory

power = volts times current

if the voltage is the same then it will draw more current. more current = higher temperatures

this is the same reason why in your house you need a thicker cable for sockets than for lights


i get 50oC with my i5 2500 i doubt temps would double with ivy bridge and if they did i would turn ir off and reseat heatsink and fan


people have had high temperatures like this when they havent put the fan on right usually with way too much thermal paste
m
0
l
March 15, 2012 9:35:21 PM

jaguarskx said:
[..] games gain little benefit from more than 4 cores. The Sandy Bridge-E series are for the enthusiasts; it's not mainstream. Therefore, if you are an enthusiast and you have the money to spare, then you buy it.


For the enthusiasts .... and for the number-crunchers who have real work to do. In the latter case it's not so much money to spare as just sensible to pay the motherboard and processor price differential and get 50% more performance out of a single box. ;-)
m
0
l
a b à CPUs
March 15, 2012 10:51:11 PM

jaguarskx said:
There really is no need for Intel to continue to produce Sandy Bridge CPUs once Ivy Bridge CPUs have been released. Running a production line takes money continuing to manufacture Sandy Bridge CPU can cannibalize sales of Ivy Bridge CPUs. The right thing to do is to cease all production of Sandy Bridge CPUs, then retool the fabs for future CPU production.


You do realize that fabs can run multiple platforms at the same time. The last wafer fab I worked in was running 200 different chips in the same fab. No re-tooling required. Only thing that changes is the photo masks and the scheduling of the different processes.
m
0
l
March 29, 2012 9:07:37 AM

I was planning to build a new rig with intel i7 2600k. But then i came to know about the release of ivy bridge. I thought i'd wait for it to release. So after reading this whole thing about heat issues and all, i feel like i should stick with 2600k. What are your opinions? Should i buy ivy bridge? Is that heat thing really a big problem? If its not then i will buy ib.
m
0
l
a c 82 à CPUs
March 29, 2012 9:30:01 AM

Its not a lot of heat, and it can be clocked 10% lower to get the same performance (roughly speaking, ok very roughly speaking, based on the generic statement of a 10-15% improvement in performance) so a 4.1Ghz OC would be similar perhaps to a 4.5Ghz OC.

The problem is in my opinion (based on engineering principles) the interface between the die and the heat spreader, because the die is smaller, that interface has to be 'better' per mm2 to get the same result. I'd hope that at this stage they are testing various materials, and working to improve this.

m
0
l
a b à CPUs
March 29, 2012 1:55:27 PM

Well, the die is packed smaller, so that alone would add heat. Throw in the increase in GPU performance, and that could explain the difference.
m
0
l
a c 82 à CPUs
March 29, 2012 2:31:49 PM

it'll get hotter until the DT is high enough to provide enough additional heat flux, thats what the temp will stabilise at.
m
0
l
!