Intel Ivy Bridge Processor IPC and Overclocking @ [H]

It is not done for better temps. It is to protect the cpu die. Amd used to have open die designs, back with socket A. A lot of people killed their cpus when installing a heatsink. For the cost of adding a heat spreader is rather low, and prevents death of a good cpu with improper installation of a heatsink.

It has admittedly been a while, but I'm pretty sure that Intel's stated reason for moving to integrated heat spreaders was to help with heat dissipation. The core cracking thing was a problem and the IHS was a solution to it, but all the comments Intel made at that time were in relation to cooling the smaller and smaller dies. Hence the name integrated heat spreader and not integrated core protector or something. I was trying to find the link to the initial IHS announcement, but can't find it now.
 
Intel also had open die cpu's. They just held up better on the P3. Dont know why but they were hard like the H.
 
Hence the name integrated heat spreader and not integrated core protector or something. I was trying to find the link to the initial IHS announcement, but can't find it now.

Although, I've heard it referred to as Idiot Handling Shield on more than one occasion when I worked in computer repair. :D
 
It has admittedly been a while, but I'm pretty sure that Intel's stated reason for moving to integrated heat spreaders was to help with heat dissipation. The core cracking thing was a problem and the IHS was a solution to it, but all the comments Intel made at that time were in relation to cooling the smaller and smaller dies. Hence the name integrated heat spreader and not integrated core protector or something. I was trying to find the link to the initial IHS announcement, but can't find it now.

Common sense man, sticking a plate between your die and the heatsink is not going to help dissipate heat. They do it to protect the die, and they say something else.
 
Common sense man, sticking a plate between your die and the heatsink is not going to help dissipate heat. They do it to protect the die, and they say something else.

Well, at least this one test found that it really didn't make a difference. It really depends how they attach it - if it is soldered, then it is essentially the same as having a larger core, which is going to help temps by providing more surface area for heat dissipation.

http://www.overclock.net/t/1249419/pcevaluation-intel-i7-3770k-temperature-measured-without-ihs
 
Curious, for Sunday, does anyone think that they will come up early in the morning or like 8AM PDT?
 
So if I just sat there tomorrow night and it turns midnight, I should be able to order a CPU around 12AM Sunday morning/Saturday night?
 
They might be. This is AMD's chance to make up ground lost with Bulldozer. If Piledriver can come out with a big jump in performance, they wont be faster than IB but maybe they'll be fast enough to be a viable alternative then with some nice pricing like a $130 Black Edition quad core or $160 six core, they could be competitive especially since it looks like Intel is sticking to its $250 minimum to overclock.

After all the huge hype and major disappointment for BD I'm not holding my breath for Piledriver. It would be nice if they could hit near SB/IB numbers at a great price though.

NetBurst Bridge

These sum up my hope, doubt, and reason for hope, respectively. To put it through another member's words, "AMD will never get a chance like the Pentium 4 again". This is not a chance like the P4 was, but it's the closest thing in a long time.
 
Well, at least this one test found that it really didn't make a difference. It really depends how they attach it - if it is soldered, then it is essentially the same as having a larger core, which is going to help temps by providing more surface area for heat dissipation.

http://www.overclock.net/t/1249419/pcevaluation-intel-i7-3770k-temperature-measured-without-ihs

You can't solder directly to a microprocessor. Common sense, heating the die to hundreds of degrees is not going to help down the line.

Plus, you could never remove a soldered IHS, so it's a moot argument as part of a test.

I'll just ignore the spurious argument about surface area.
 

A little aside here, but I'd like to add that's a very nice wallpaper of Saya of Blood+ you have there. :D

Other than that, very lucky person to get a 3770K early than a lot of [H] users (outside of those that tested the processor here on [H]).
 
A little aside here, but I'd like to add that's a very nice wallpaper of Saya of Blood+ you have there. :D

Other than that, very lucky person to get a 3770K early than a lot of [H] users (outside of those that tested the processor here on [H]).

Funny, I knew it was Blood+ before reading your post lol.
 
You can't solder directly to a microprocessor. Common sense, heating the die to hundreds of degrees is not going to help down the line.

Plus, you could never remove a soldered IHS, so it's a moot argument as part of a test.

I'll just ignore the spurious argument about surface area.

Ummm, what do you think the whole discussion about Ivy Bridge using TIM instead of solder is about? The IHS on Sandy Bridge is soldered to the core using low-temp flux-less solder - that's kind of the whole point of the complaints about Ivy switching back to TIM. And you can remove a soldered IHS, you just have to be ballsy with a propane torch. Maybe next time hit up Google before posting.
 
Last edited:
You can't solder directly to a microprocessor. Common sense, heating the die to hundreds of degrees is not going to help down the line.

Plus, you could never remove a soldered IHS, so it's a moot argument as part of a test.

I'll just ignore the spurious argument about surface area.

:rolleyes:

Ummm, what do you think the whole discussion about Ivy Bridge using TIM instead of solder is about? The IHS on Sandy Bridge is soldered to the core using low-temp flux-less solder - that's kind of the whole point of the complaints about Ivy switching back to TIM. And you can remove a soldered IHS, you just have to be ballsy with a propane torch. Maybe next time hit up Google before posting.

Don't you love it when someone gets all mighty feeling and they obviously have no clue what they are talking about?
 
This is getting silly, sitting with my i7 920 @ 4.3ghz, my biggest reason to upgrade is basically to get onboard SATA III.

Has there been a bigger bargain in desktop history than that fricking chip, it cost £178 3 years ago. I had the money for an extreme but it was pointless and now I'm sitting here with $4000 to spend after cancelling my last upgrade and still I'm tempted to just get a pair of GTX680's and forget the rest.

My usual trigger point is a 30% improvement in performance, it looks like it'd be close but it's painful up here on the fence
 
This is getting silly, sitting with my i7 920 @ 4.3ghz, my biggest reason to upgrade is basically to get onboard SATA III.

Has there been a bigger bargain in desktop history than that fricking chip, it cost £178 3 years ago. I had the money for an extreme but it was pointless and now I'm sitting here with $4000 to spend after cancelling my last upgrade and still I'm tempted to just get a pair of GTX680's and forget the rest.

My usual trigger point is a 30% improvement in performance, it looks like it'd be close but it's painful up here on the fence

Upgrading that CPU is only worth it if you play one of a very very short list of games that is highly CPU limited.

Red Orchestra 2 is one of these.

For most uses - however - that chip is still overkill.
 
I am currently using the software MediaCoder, which is free (though it nags you to donate sometimes) and supports Intel QuickSync encoding. It has a lot of features compared to many of the commercial ones listed in the Intel slide.

Perhaps you guys could try evaluating that in your reviews too? I'm currently getting about 11-12x encoding speeds on my i5-2400 @ 3.6GHz as compared to 4x without QuickSync enabled. I'm really happy I went with the Z68 instead of the P67 when I upgraded. Probably there's much more tangible gains to be had when using QuickSync on Ivy Bridge?

But of course, I'd love to have QuickSync support for Handbrake...
 
Zarathustra[H];1038665193 said:
Upgrading that CPU is only worth it if you play one of a very very short list of games that is highly CPU limited.

Red Orchestra 2 is one of these.

For most uses - however - that chip is still overkill.

What about emulators? My Athlon II X3 lags like a crazy playing a lot of games on Dolphin and PCSX2 at 1x native resolution with no AA or AF and I am growing more and more intolerant of it as the days go by. My inner graphics Nazi is telling me to go for i5-2500K or i5-3570K so I can play games at higher native resolutions with all the graphical bells and whistles turned OFF/[ON], but I can't decide between which of the two CPUs. I'm planning to OC to 4.8 GHz with the Turbo multiplier set to 5.2 GHz with two active cores since emulators aren't all that optimized for the quads. I'm guessing those emulators won't take advantage of Ivy Bridge's architecture any time soon? Or am I better off waiting for Haswell in order to get an even bigger upgrade?
 
Looks like this will be a good CPU for my next rig. I'll likely be upgrading from my I7-920 main rig to something that also includes SATA-3, USB 3.0, and a more robust motherboard in general so making the jump from x58 isn't too bad in this case. Although I am not running into any problems with my current setup. Would just like a little more speed out of the attached components. So why not upgrade across the board and re-purpose the good 'ol 920 to server duty.

I do have to say though, that the i7-920 has been my favorite CPU of all time as far as longevity in my main rigs.

Thanks for the write-up Kyle!.

QFT, i had mine running at about 3.6 ghz for the longest time, then i got an H80 and bumped that to 4.2 ghz and feel ZERO need to get a new processor for anything but benchmark scores.

I've had the 920 since 2009...
 
QFT, i had mine running at about 3.6 ghz for the longest time, then i got an H80 and bumped that to 4.2 ghz and feel ZERO need to get a new processor for anything but benchmark scores.

I've had the 920 since 2009...

lol, almost exactly what I did. Had mine at 4.0 on air for 2.5 years but decided to ditch my heatsink when I emigrated (shipping with a 700g weight attached to the socket seemed silly) Got a H100 and now have mine at 4.5ghz after a bit of work this week.

Gonna buy a GTX 690 and am a bit worried in case the lack of pci/e 3.0 becomes an issue but otherwise I'm going to be happy waiting for Haswell. I think I'm going to rename my PC Methuselah.
 
oc.jpg


Got mine yesterday, OC'ed a little. Gonna keep it at this.

Also, is it me or is there a rather big temperature difference between cores with ivy and are they a bit more jumpy? I come from a Q9450 and those were all around the same temperature, and hardly moved when idling/browsing in windows. But this 3570k looks like a rollercoaster ride.
Both have the Thermalright Trueblack as cooler. Idle is between 31-40 (lowest and highest core). On load in Prime I hit 65ish. This is with all the fans on the lowest setting, no noise. hehe
 
whoa thats nice, my 2700k needs almost .3 more volts to hit 4.8 though i only ever hit 72C under load
 
old story, everyone knows ....so WTF you wanna proof with it? This topic is bout IB overclocking and not about theories why its hotter compared SB. im getting tired of these posts.
 
I'm just going to wait for haswell
you think haswell gonna give you clocks like SB or even better temps ? That cpu will be even smaller and most likely will utilize trigate too , all smashed into to a tiny processor.
my bet is haswell k stable Oclocks will stop @+ 800mhz
 
well maybe by then the ipc will have improved that an only 800mhz boost will be as fast or faster than SB, either way im rocking a wicked fast SB setup so im not complaining
 
you think haswell gonna give you clocks like SB or even better temps ? That cpu will be even smaller and most likely will utilize trigate too , all smashed into to a tiny processor.
my bet is haswell k stable Oclocks will stop @+ 800mhz

Haswell will be 22nm also, so it probably won't be much smaller than Ivy. Larger, most likely.
 
you think haswell gonna give you clocks like SB or even better temps ? That cpu will be even smaller and most likely will utilize trigate too , all smashed into to a tiny processor.
my bet is haswell k stable Oclocks will stop @+ 800mhz

Lol everyone is so spoiled by SB overclocking that they figure the next release will be even better.

Reality check, intel is trying to make its products more efficient at stock speeds. I highly doubt they design their products with a mindset of "i want users to be able to overclock this with minimal effort and get huge gains"
 
I just purchased one of the new SB processors for my main PC that does a lot of video coding, and file packing/unpacking. All I have to say is this thing is an absolute monster. I don't see upgrading from this for a WHILE.
 
BTW, I got the unlocked version SB, and it's probably the easiest piece of hardware to overclock that I've ever worked with. I suppose my Mobo's UEIF helps though...
 
Anyone here with a prolimatech megahelms push pull for their Ivy Bridge and overclocks over 4.2ghz??
 
Hello, does anyone know if there's any info about a recall for the 3770K to replace the TIM that's causing high temps?

The temps are only high when overvolting so no. At least it doesn't get as hot as a bulldozer cpu when overclocking.

I'm sure that they lost some sales over that decision though. Even still, I'm tempted to pick up a 3570k just to pop the IHS and see what I can do with it.
 
Back
Top