Haswell 4c8t @2.6 Fritz Chess Benchmark <Update:2.8G Haswell>

[Nemesis 1]

10.6% over Ivy (Which had about 5% over sandy) is pretty good. This isn't AVX2 optimized or anything.

Oh I understand that this doesn't include AVX2 or Tsx . But thats a wait and see , Sure it helps make haswell future proof . I not going to go the way of others point at future apps that will use next generation apps and recompiles. Of course I would like to see intel 8 real cores against the AMD processor with 8 cores . That is no longer how we measure performance . We now measure performance based on price. Because No one can come close to 8 real intel cores. It was so easy in 2006 . I new this game was going to played this way . As I stated long befor it happened in real life . Intel played its core hand vary well . That doesn't change the fact I disappointed in only 10% gain in that benchmark . As I was expecting HT performance improving . With this benchmark I don't know where the improvement is as it uses HT.

 

[Nemesis 1]

I don't find the other benchmark all that questionable. Maybe this benchmark is questionable lol

You wouldn't . I will say this to you one time and one time only . Who is not a liar . Who would not put his job in jeopardy to lie about intels performance . If Who says those benchies are bogus they are bogus. We seen what happened to Tim for going public with larrabee.

 

[Nemesis 1]

Are most people "happy" with a 14-15% RAW IPC Increase from haswell over SB on legacy code?


Damn.

I'm not.

Unless MT gains are heavily better - i'm disappointed.

And i didn't mean to start of a intel vs AMD thing - i just made a small jab at how projected peformance for SR would suddenly make them performance kings in one "fix up" generation.

Clearly people can't take jest atm - my bad.

You didn't start nothing . The Op did in his first post . They should have used IVB. or SB and not compared against AMD.

 

[Nemesis 1]

lol, how would it be me when I didn't post above what you quoted?

Ya I know see I edited befor you replied. The OP introduced AMD to this thread . I missed that . Mac was the first to mention AMD but he did not introduce AMD to this Topic.

 

[Nemesis 1]

Hopefully not!

But it's still a relative indication of Legacy code performance.
And i would have preferred just a little more oomph out of that.


I know it's a weird mindset - but i sort of expect Intel to move both in the complex arena for computation and the legacy way.

IMHO, i'm all for Intel being the big bad untoucheable giant - as long as they deliver each generation - but they gotta deliver.

None of this incremental shit on a tock

Thats all well and fine. You have to keep in mind intel is behind on the gpu side of things . they have to adderess this first. If GT3e is as good as it appears to be in the 15 watt and below it will be untouchable . I honestly not happy about no GT3e on desktops . I would love playing around with one of these at say 1.5 GHz on the Gpu and 5.5ghz on the Cpu.

Just goes to show you intel either doesn't care if its good enough for desktop or there are agreements inplace that don't allow Intel to put these on desktops .OR we have another possiability

Gt3E has 40 eu Intel has not said what GT2 is or GT1. But what is it we know? We know that intel adds another slice to the GT3e . To make 40 eu . We know a single slice is 20 eu, So I believe GT1 is 20 eu and GT2 is 40 eu without the Edram . Intel says they improved memory subsystem so there are no bank conflicts . It could very will be That A GT2 on desktop walks all over A GT3e. It would run at almost 2x the speed with O/Cing it will run 2x faster. With hi frequency DDr3 . Really, I at the point of not caring about haswell other than its graphics and intel is BRAGGING . This sucks normally by now I would know all these things . But Bob wouldn't give me a haswell. Maybe I might get a broadwell. What Robert delivered at xmas was good enough .

 

[parvadomus]

I can only say one thing: if mhz delta between AMD and intel remains the same, an 8 core SR will kick i7-4770K's butt.

 

[Nemesis 1]

I can only say one thing: if mhz delta between AMD and intel remains the same, an 8 core SR will kick i7-4770K's butt.

I don't care I will always buy intel /nv. When and if the time comes preferrably just intel. I would like to see pcie slots go the way of the Do Do Bird on desktops and on servers I preferr to see a Cpu type socket for the Accelerator. Intel says this is what there doing The FTC said intel could do this in 2013 so do it intel do it.

I would vary much also like to see AMD price their processors like they did befor they were conroed. I would love to see the cheapest AMD CpU at over $300 dollars Because intel will NOT lower their price I dime. Amd leading in performance maybe what you think you want . If AMd beats intel you will not want to pay the price.

 

[frozentundra123456]

This again is not an AMD vs. Intel who is better thread. We are talking about Haswell performance, not AMD and Intel's R&D.


If you want to have a VS discussion go find a thread from 2006.

Every thread morphs (degenerates??) into intel vs amd these days.

 

[ShintaiDK]

I can only say one thing: if mhz delta between AMD and intel remains the same, an 8 core SR will kick i7-4770K's butt.

And Phenom was 50% faster than Core 2 too.

 

[Rvenger]

Every thread morphs (degenerates??) into intel vs amd these days.

Ya, I should know that by now, right. :hmm:

 

[frozentundra123456]

Are most people "happy" with a 14-15% RAW IPC Increase from haswell over SB on legacy code?


Damn.

I'm not.

Unless MT gains are heavily better - i'm disappointed.

And i didn't mean to start of a intel vs AMD thing - i just made a small jab at how projected peformance for SR would suddenly make them performance kings in one "fix up" generation.

Clearly people can't take jest atm - my bad.

If it is 10% improvement vs IB across a wide variety of ST and MT applications, I think that would be a decent improvement, considering the SB and IB were already highly optimized.

 

[Rvenger]

And Phenom was 50% faster than Core 2 too.

How far away was Phenom from Core 2?

 

[ShintaiDK]

Really I am impressed they found an extra 10% that apparently Sandy/Ivy Bridge left on the table. Considering how good those chips are, to come in over the top of that with another 10% IPC is darn good.

Just makes me wonder what lengths they are going to have to go to in order to keep finding another 10% and then another 10% going forward. Pollack's rule and all that.

But you are fightning the instant gratification generation

I wonder how much left there is to gain, specially on the OoO concept.

Haswell is 4 issue wide, Poulson is 12 issue wide.

Looking at Haswell and the future in performance. It essentially already begs for 512bit units in Skylake.

 

[ShintaiDK]

How far away was Phenom from Core 2?

With or without the TLB bug?

Without you can find here:
http://www.anandtech.com/show/2378

And with here:
http://techreport.com/review/13741/phenom-tlb-patch-benchmarked/3

 

[Nemesis 1]

I can only say one thing: if mhz delta between AMD and intel remains the same, an 8 core SR will kick i7-4770K's butt.

Let me ask you something . If intel raised the wattage on a IVB to a 100watt TDC . What would the IVB running at in MHz. I know the answer , But I think you do not . Intel TDC is lower than AMDS Why ? So AMD could claim to be closer to intel than they actually are . Run both Cpus at the same exact wattage and see who has what . Amd friend isn't even in the game . AMD is playing a wattage game . Its like electricity is free when measuring the cost of owning a cpu . Now if you think A 1oo watts is only a few dollars a year you would be wrong . Besides this price differance between Intel AMD is amusing when I look at GPU pricies . Haswell and whatever AMD has to offer will push those GPU pricies even higher as The Intel graphics and the Amd graphics proove to be good enough for many one time discrete users. If Intel GT3e is all its cracked up to be . Amd has to come back with something even better. Discrete will die because it will not have enough support and the pricies will skyrocket . Supply and demand working for us all .

 

[inf64]

Phenom I vs C2Q (2.5Ghz 65nm vs 2.4Ghz 65nm vs 2.33Ghz 45nm):
100pts 9850 X4 vs 102.7pts C2Q (Conroe) vs 103.7pts C2Q(Penryn)
That's the average of 12 workloads including 4 cpu heavy games . Closest Phenom II in the chart is actually crippled(no L3) Propus based Athlon II X4 620 (2.6Ghz) at 104.4pts. All in all 65nm vs 65nm they were close and 45nm vs 45nm also pretty close. In both cases intel had a slight(~5-6%) IPC lead but OCed significantly better so that's why C2Q was overall better desktop chip. Not that Athlon II or phenom II were bad choices,for the money they were very nice alternative to C2Qs.

 

[ChronoReverse]

The only problem was that they were up against Nehalem (since it was the Phenom I that was late against Core 2).

 

[ShintaiDK]

Hillarious flashback. Since its a rerun event:
http://forums.anandtech.com/showthread.php?t=2134911

And yes, we are out of Haswell benchmarks for now. Not much drama when everything behaves as expected

 

[parvadomus]

Let me ask you something . If intel raised the wattage on a IVB to a 100watt TDC . What would the IVB running at in MHz. I know the answer , But I think you do not . Intel TDC is lower than AMDS Why ? So AMD could claim to be closer to intel than they actually are . Run both Cpus at the same exact wattage and see who has what . Amd friend isn't even in the game . AMD is playing a wattage game . Its like electricity is free when measuring the cost of owning a cpu . Now if you think A 1oo watts is only a few dollars a year you would be wrong . Besides this price differance between Intel AMD is amusing when I look at GPU pricies . Haswell and whatever AMD has to offer will push those GPU pricies even higher as The Intel graphics and the Amd graphics proove to be good enough for many one time discrete users. If Intel GT3e is all its cracked up to be . Amd has to come back with something even better. Discrete will die because it will not have enough support and the pricies will skyrocket . Supply and demand working for us all .

All you said is irrelevant, AMD is playing the price/performance, it cant compete vs 22nm with 28nm or 32nm on pr per watt thats easy to understand.
It will let intel release their crap first so they can tune their frecuency to match or even surpass intel @ MT performance, I think its simple to understand. You will see as I said 4770K getting rapped by SR if we see an SR 8-core CPU, at least @ MT tasks, maybe this time they can even reduce the high end part to a 6 core part, that would enough to reach i5 levels, which are they pr target.
PD : you can buy whatever you want. And its funny to see a lot of fanboys cry by writing just 1 line of text :awe:

 

[CHADBOGA]

Phenom I vs C2Q (2.5Ghz 65nm vs 2.4Ghz 65nm vs 2.33Ghz 45nm):
100pts 9850 X4 vs 102.7pts C2Q (Conroe) vs 103.7pts C2Q(Penryn)

Why are you using the 2.5Ghz 9850 X4 instead of the 2.4Ghz 9750 X4 in this comparison?

 

[Torn Mind]

All you said is irrelevant, AMD is playing the price/performance, it cant compete vs 22nm with 28nm or 32nm on pr per watt thats easy to understand.
It will let intel release their crap first so they can tune their frecuency to match or even surpass intel @ MT performance, I think its simple to understand. You will see as I said 4770K getting rapped by SR if we see an SR 8-core CPU, at least @ MT tasks, maybe this time they can even reduce the high end part to a 6 core part, that would enough to reach i5 levels, which are they pr target.
PD : you can buy whatever you want. And its funny to see a lot of fanboys cry by writing just 1 line of text :awe:

Steamroller's hexacore will have a difficult time besting Sandy Bridge-E. The octocore should surpass the 3960X in terms of multithreaded performance. Whatever combination of IPS(=IPCxclockspeed) improvement they make will have to be about 20% faster than Piledriver. To catch up with Haswell, AMD will need its own Conroe architecture.

 

[parvadomus]

Steamroller's hexacore will have a difficult time besting Sandy Bridge-E. The octocore should surpass the 3960X in terms of multithreaded performance. Whatever combination of IPS(=IPCxclockspeed) improvement they make will have to be about 20% faster than Piledriver. To catch up with Haswell, AMD will need its own Conroe architecture.

I didnt say anything about SB-E, its just in another league. AMD would have to release a high frecuency 10-core part to rival haswell intel hexa, it's such a waste of silicon.

 

[Edrick]

I can only say one thing: if mhz delta between AMD and intel remains the same, an 8 core SR will kick i7-4770K's butt.

Posts like this remind me of the days before BD specs started being leaked. Such high expectations.

 

[Genx87]

Going to be difficult to not upgrade my i5 2500K with this thing.

 

[Idontcare]

Posts like this remind me of the days before BD specs started being leaked. Such high expectations.

I have it on high confidence that IPC will not decrease with Steamroller!








(I seriously hope that doesn't jinx anything)