Haswell Benchmarks?

[StrangerGuy]

Totally unrealistic these scores. Ivy Bridge improved IPC by 3% to Sandy Bridge. And look what Intel changed, only a few minor changes were made for Ivy Bridge. Haswell on the other side has much bigger changes included from the infos we got from IDF last year. So the IPC gain must be much bigger than from Sandy to Ivy.

You don't get it, if AMD have it's BD moment than Intel must have one too in this generation, hur hur hur. *sarcasm* Yeah maybe that will happen if Haswell got a red flag like having a 4 port scheduler shared between 2 cores but how the hell will it get slower when EVERY CPU resource got better from IB.

 

[inf64]

I doubt there will be a single workload that will run slower on Haswell than on IB/SB.

 

[Makaveli]

So I should maybe skip haswell and wait for the next generation before I upgrade my 2500k?

This may be alittle out there and some may say crazy but.....

How about waiting for reliable benchmarks from a site you trust ??

:biggrin:

 

[mikk]

haswell is very healthy in our labs, we don't see any case where it is slower than IvyB. .ru numbers are funky!

https://twitter.com/FPiednoel/status/297005032992284673

 

[frozentundra123456]

https://twitter.com/FPiednoel/status/297005032992284673

Should I know who this is?? Does he work for intel or something? Just wondering why he had the chip and was testing.

He really didnt say much though. I would be shocked if it was slower, but he did not say it was faster either, although perhaps he cant talk about that.

 

[Eeqmcsq]

Should I know who this is?? Does he work for intel or something? Just wondering why he had the chip and was testing.

He really didnt say much though. I would be shocked if it was slower, but he did not say it was faster either, although perhaps he cant talk about that.

Apparently, he does work for Intel. He's a "Principal Engineer / Performance Architect at @Intel".

https://twitter.com/FPiednoel

 

[inf64]

That is "thewho" now infamous intel employee . He was active during conroe and penryn launches,posting on various forums and getting into fights with other forum users. Not a best way to represent intel IMO. Their CPUs rock'n'roll ,they don't need a fanatical guy to belittle and personally attack other forumers who have a different opinion than him. Just my 2 cents.

Also it's worth saying he probably tells the truth there and Haswell is just fine.

 

[mikk]

Should I know who this is?? Does he work for intel or something? Just wondering why he had the chip and was testing.

Yes he is Principal Engineer, Senior Performance Analyst / System Architect from Intel.

http://www.linkedin.com/in/francoispiednoel
https://twitter.com/FPiednoel

 

[videogames101]

this thread does not need to exist

 

[ShintaiDK]

That is "thewho" now infamous intel employee . He was active during conroe and penryn launches,posting on various forums and getting into fights with other forum users. Not a best way to represent intel IMO. Their CPUs rock'n'roll ,they don't need a fanatical guy to belittle and personally attack other forumers who have a different opinion than him. Just my 2 cents.

Also it's worth saying he probably tells the truth there and Haswell is just fine.

Dont twist the truth. It was people who attacked him because they didnt want to accept the realities.

Phenom was also supposed to be 50% faster than Core 2 before the hype settled. And Core 2 was overclocked Opterons...

 

[Nemesis 1]

That is "thewho" now infamous intel employee . He was active during conroe and penryn launches,posting on various forums and getting into fights with other forum users. Not a best way to represent intel IMO. Their CPUs rock'n'roll ,they don't need a fanatical guy to belittle and personally attack other forumers who have a different opinion than him. Just my 2 cents.

Also it's worth saying he probably tells the truth there and Haswell is just fine.

Care to give us a link of Who picking a fight . I seen it the other way around . Where Who was attacked by liars. Links to show otherwise

 

[Sweepr]

Haswell will probably bring the usual 10% IPC boost (vs Ivy Bridge) we would expect from an Intel tock. This.. coupled with more agressive Turbo and likely higher clocks (especially mobile parts) could bring a 10-25% performance uplift (legacy).Then there's AVX2 and the new GT2/GT3 IGPs.. Even if desktops dont get higher clocked parts, +10% better performance per clock than SB + improved OC capability (rumoured) should convince many people to upgrade.

That is "thewho" now infamous intel employee . He was active during conroe and penryn launches,posting on various forums and getting into fights with other forum users. Not a best way to represent intel IMO. Their CPUs rock'n'roll ,they don't need a fanatical guy to belittle and personally attack other forumers who have a different opinion than him. Just my 2 cents.

Also it's worth saying he probably tells the truth there and Haswell is just fine.

A quick Google search shows me the other way around. Seems like the other brand's fans loved to attack him.

http://pc1.stonearch.net/forums/showpost.php?p=108531&postcount=17

 

[Ferzerp]

propaganda backfire, lol

 

[inf64]

Well Llano and Bobcat did great . Only fail was bulldozer,so 2 out of 3 which is not bad.

 

[CHADBOGA]

That is "thewho" now infamous intel employee . He was active during conroe and penryn launches,posting on various forums and getting into fights with other forum users. Not a best way to represent intel IMO. Their CPUs rock'n'roll ,they don't need a fanatical guy to belittle and personally attack other forumers who have a different opinion than him. Just my 2 cents.

Also it's worth saying he probably tells the truth there and Haswell is just fine.

Sadly your definition of fanatical involves someone who attempts to correct the most egregious lies by demented AMD boosters.

 

[WhoBeDaPlaya]

What I want from Haswell is lower power consumption and better graphics, single threaded performance is fine as it is.

**** no! I did not wait 2+ years for no IPC increase :|
My target : 5GHz HW = (1.1 * 1.03 * 1.1) = 6.2GHz Nehalem

 

[ShintaiDK]

Well Llano and Bobcat did great . Only fail was bulldozer,so 2 out of 3 which is not bad.

You mean the cancelled Wichita product with the completely off estimate, the flopped Bulldozer and then Llano that didnt sell enough and had a 100M writeoff?

Looks to be 3 wrong out of 3.

 

[NTMBK]

You mean the cancelled Wichita product with the completely off estimate, the flopped Bulldozer and then Llano that didnt sell enough and had a 100M writeoff?

Looks to be 3 wrong out of 3.

Wichita was cancelled, but the original Bobcat(s) sold gangbusters- and Wichita was cancelled due to process woes at GloFo, not due to inherent problems with the design. Bobcat was far from a failure.

 

[ShintaiDK]

Wichita was cancelled, but the original Bobcat(s) sold gangbusters- and Wichita was cancelled due to process woes at GloFo, not due to inherent problems with the design. Bobcat was far from a failure.

The original statement didnt include bobcat.

 

[ShintaiDK]

The first retain Haswell products should be finished in mid march and starting to ship to distributors. Then we might see something more.

 

[Charles Kozierok]

 

[NTMBK]

The original statement didnt include bobcat.

I presume the "original statement" would be this- "Well Llano and Bobcat did great . Only fail was bulldozer,so 2 out of 3 which is not bad." Unless I missed something?

 

[ShintaiDK]

I presume the "original statement" would be this- "Well Llano and Bobcat did great . Only fail was bulldozer,so 2 out of 3 which is not bad." Unless I missed something?

http://pc1.stonearch.net/forums/showpost.php?p=108531&postcount=17

 

[sm625]

But what could cause SuperPi 1M to be so much worse while the other scores are generally better? What could possibly cause that?

And the WPrime 1024M scores make absolutely no sense at ALL. An i5 2500k will complete that benchmark in around 250-275 seconds. There are scores posted all over the web. So how the hell are they getting 430 seconds? An i3 first gen will do it faster than that! I call total bull.

If I had logmein installed on my i5 machine I would run that test right now. I bet it blows that 430 seconds score into next week...

 

[tweakboy]

Well if they use something else but math programs your going to get similar results,, clock for clock ivy 2.8 and haswell 2.8 are the same in performance. Haswell will just run cooler and have super fast GPU inside CPU. Haswell is more then a tick ,, what tock ,, when benchmarks show their the same......... thx gl