Haswell vs. Ivy Bridge Overclocking

[Computer Bottleneck]

How well do you think Haswell desktop will overclock compared to Ivy Bridge desktop? (Assume the average max OC using large tower air coolers for both processors)

[Ferzerp]

Umm, no one knows (that isn't an Intel employee)...... What is this thread for?

[Computer Bottleneck]

Originally Posted by Ferzerp

Umm, no one knows (that isn't an Intel employee)...... What is this thread for?

It is just a speculation thread.

Some things to consider:

1. Material used under Haswell's heatspreader? Will it be TIM (as in Ivy Bridge) or solder?

2. Haswell specific changes to the 22nm process node?

http://hexus.net/tech/news/cpu/45489...rs-ivy-bridge/

Most surprisingly, perhaps, is that Intel has revealed that despite being a tock release, Haswell will in fact feature a different type of transistor to the one in Ivy Bridge. Has the firm realised a design even more efficient (power, space?) than the 3D Transistors it last featured? Or perhaps has it found a design better capable of extreme overclocking? There were suggestions that Intel's move away from fluxless-solder to attach the heat-spreader could have been due to transistor fragility.

http://www.theinquirer.net/inquirer/...ge-transistors

Intel has revealed that the transistors used in Haswell chips will not be the same as those in Ivy Bridge processors, despite being fabbed on the same 22nm process node.

The trio also stressed that advances came from both process node and architecture developments.

As Intel's Haswell will be a 'tock' processor, on a macro level it shares the same process node as Ivy Bridge. However the firm said at the low level there are changes and said, "We have done a lot of work with our manufacturing friends [...] in order we can get into these lower power envelopes. So we can tune the process for higher performance or lower power. [...] Definitely, don't have the mindset that because it is 22nm it is the same as Ivy Bridge. The process is evolving constantly, especially if we're targeting different products."

[Vectronic]

There were suggestions that Intel's move away from fluxless-solder to attach the heat-spreader could have been due to transistor fragility.

Due to laziness in figuring out how to solder it... "our previous method won't work... what should we do?"... "ahh screw it, just toss goo on it".

[bunnyfubbles]

couple factors going on here

1. Intel has more experience with 22nm
2. IIRC, it has been mentioned that Intel is going back to solder
3. It has been made known that Haswell will have overclocking friendly features more than unlocked multiplier made available

Really, the only thing that could potentially go against Haswell is that its possible the design isn't as friendly to higher clockrates, however that would then most likely mean its IPC is that much more beastly and people are going to be happy regardless - for instance, even with all the OC advantages, if Haswell can only muster ~4.6GHz but yet is 10-15% faster that's basically like having a 5-5.2GHz Ivy, not bad.

That being said, I don't necessarily think IPC improvements will be quit that good, however I am leaning towards 4.8+GHz being more of a norm.

[FlanK3r]

I think, Haswell will be OC similar as Sandy, so between 4700-5100 MHz stable

[MisterMac]

I'm hoping for 5 GHZ average newb overclocking as opposed to 4,6-4,7 ranges.


Without too much voltage - but i'm not thinking it's realistic :C

[2is]

I hope it's a terrible over-clocker so that I'm not tempted to upgrade. :p

[dma0991]

I'm guessing 5GHz will be the norm with Haswell but then again, that's the same prediction I made with IB and it didn't turn out that way. Assuming the 22nm process has matured since IB, 5Ghz should be possible. With my lemon Core i5 3570K, I could go 5GHz with no temp constraints but it will still be unstable at suicide run voltages.

[ShintaiDK]

Originally Posted by 2is

I hope it's a terrible over-clocker so that I'm not tempted to upgrade. :p

LoL

[BTRY B 529th FA BN]

if I can get 5.29Ghz stable, then i'll be happy. lol

[ShintaiDK]

[Idontcare]

Originally Posted by dma0991

I'm guessing 5GHz will be the norm with Haswell but then again, that's the same prediction I made with IB and it didn't turn out that way. Assuming the 22nm process has matured since IB, 5Ghz should be possible. With my lemon Core i5 3570K, I could go 5GHz with no temp constraints but it will still be unstable at suicide run voltages.

Given that the development resources for Haswell were devoted to prioritizing performance scaling at the low-end of power and clockspeed, I very much doubt we are going to see much added clockspeed headroom over that of what we already see with IB.

For starters, where is the money in it for Intel? The portion of the market that comprises enthusiast OC'ers who are going to have make-or-break purchasing decisions on the basis of whether or not a 3.5GHz Haswell can overclock beyond 5GHz on air has got to be silly, truly silly, small.

So how can we, the six of us who care, really expect Intel to have devoted resources towards engineering Haswell to have ">5GHz OC'ing headroom, ya know, for those six dudes who won't buy Haswell if it doesn't do that"?

Meanwhile they are looking at a market with the potential revenue of billions of dollars if they can get the performance/watt tuned to fit within the right max wattage...that opportunity justifies expending R&D dollars and it is not an opportunity that requires tuning neither the microarchitecture nor the process technology to function in the >5GHz region of the shmoo plot.

[MisterMac]

Originally Posted by Idontcare

Given that the development resources for Haswell were devoted to prioritizing performance scaling at the low-end of power and clockspeed, I very much doubt we are going to see much added clockspeed headroom over that of what we already see with IB.

For starters, where is the money in it for Intel? The portion of the market that comprises enthusiast OC'ers who are going to have make-or-break purchasing decisions on the basis of whether or not a 3.5GHz Haswell can overclock beyond 5GHz on air has got to be silly, truly silly, small.

So how can we, the six of us who care, really expect Intel to have devoted resources towards engineering Haswell to have ">5GHz OC'ing headroom, ya know, for those six dudes who won't buy Haswell if it doesn't do that"?

Meanwhile they are looking at a market with the potential revenue of billions of dollars if they can get the performance/watt tuned to fit within the right max wattage...that opportunity justifies expending R&D dollars and it is not an opportunity that requires tuning neither the microarchitecture nor the process technology to function in the >5GHz region of the shmoo plot.

...because of ePeen yo.

Gotz to hav dat biggest ePeen, yo.

AMDz ain't gotz nuttin on uz Intelz - guyz man com on!.



Atleast - that's how i imagine\hope the marketing head pitching it to Otellini

[Idontcare]

Originally Posted by MisterMac

...because of ePeen yo.

Gotz to hav dat biggest ePeen, yo.

AMDz ain't gotz nuttin on uz Intelz - guyz man com on!.



Atleast - that's how i imagine\hope the marketing head pitching it to Otellini

Ironically I imagine that is rather precisely how the internal marketing pitch for bulldozer went with Dirk.

Maximum Speed | AMD FX Sets a Guinness World Record

There are two stories from AMD today. One is of the years of technology development that have led to a Guinness World Record, an achievement made possible by the AMD FX processor, achieving the highest frequency for a computer processor to date. The second is one of devotion, of individuals who, for better or worse, have become computer hardware racers, test pilots of technology who push limits and often break hardware. And break records.

We also achieved clock frequencies well above 5GHz using only air or sub-$100 water cooling solutions.

http://blogs.amd.com/play/2011/09/09/guinness/

A truly epic tale of clockspeed and e-peen fanboys. A proud moment for Dirk no doubt.

[dma0991]

Originally Posted by Idontcare

Given that the development resources for Haswell were devoted to prioritizing performance scaling at the low-end of power and clockspeed, I very much doubt we are going to see much added clockspeed headroom over that of what we already see with IB.

For starters, where is the money in it for Intel? The portion of the market that comprises enthusiast OC'ers who are going to have make-or-break purchasing decisions on the basis of whether or not a 3.5GHz Haswell can overclock beyond 5GHz on air has got to be silly, truly silly, small.

So how can we, the six of us who care, really expect Intel to have devoted resources towards engineering Haswell to have ">5GHz OC'ing headroom, ya know, for those six dudes who won't buy Haswell if it doesn't do that"?

Meanwhile they are looking at a market with the potential revenue of billions of dollars if they can get the performance/watt tuned to fit within the right max wattage...that opportunity justifies expending R&D dollars and it is not an opportunity that requires tuning neither the microarchitecture nor the process technology to function in the >5GHz region of the shmoo plot.

I'm not quite the advocate of reaching such speeds. I'd be happy enough if Intel improves their Turbo Boost on desktop chips to reach near 1GHz(~4.4-4.5GHz). Current mobile chips do come with 1GHz Turbo Boost. Unless, Intel knows that Turbo Boost beyond 4GHz creates new problems or the likelihood of finding a golden chip that does that is difficult.

[NTMBK]

Originally Posted by MisterMac

...because of ePeen yo.

Gotz to hav dat biggest ePeen, yo.

AMDz ain't gotz nuttin on uz Intelz - guyz man com on!.



Atleast - that's how i imagine\hope the marketing head pitching it to Otellini

Hey, this is the company which brought you the #YOLO phone.

[Idontcare]

Originally Posted by dma0991

I'm not quite the advocate of reaching such speeds. I'd be happy enough if Intel improves their Turbo Boost on desktop chips to reach near 1GHz(~4.4-4.5GHz). Current mobile chips do come with 1GHz Turbo Boost. Unless, Intel knows that Turbo Boost beyond 4GHz creates new problems or the likelihood of finding a golden chip that does that is difficult.

That would be nice A 4.5Ghz turbo-boost model would be great.

[Nemesis 1]

My hope for IVB was 5.3ghz Couldn't get stable at 5.3ghz but its good at 5.2 ghz . For Haswell If I can get 5.5 ghz stable I will be more than happy.

[ShintaiDK]

Originally Posted by Idontcare

That would be nice A 4.5Ghz turbo-boost model would be great.

I am not so sure that a 1Ghz turboboost would be great. It leaves the possibility for too much performance delta between 2 different environment settings.

[IntelEnthusiast]

Originally Posted by Ferzerp

Umm, no one knows (that isn't an Intel employee)...... What is this thread for?

At this point I dont have a clue but I am looking forward to finding out.

BTW there is a new promo that will be lunching in the next couple days www.intelgamingpromo.com

[NTMBK]

Originally Posted by ShintaiDK

I am not so sure that a 1Ghz turboboost would be great. It leaves the possibility for too much performance delta between 2 different environment settings.

Hey, you live in Denmark, I live in Scotland- so at least we'd be on the right side of that delta!

[ShintaiDK]

Originally Posted by NTMBK

Hey, you live in Denmark, I live in Scotland- so at least we'd be on the right side of that delta!

LOL

Well in your case it might decrease next to the deep fryer

[Ferzerp]

Originally Posted by IntelEnthusiast

At this point I dont have a clue but I am looking forward to finding out.

BTW there is a new promo that will be lunching in the next couple days www.intelgamingpromo.com

It won't be a very effective promotion if it gets eaten.

[NTMBK]

Originally Posted by Ferzerp

It won't be a very effective promotion if it gets eaten.

Nonono, other way round. Free sammich with every i7.