Intel "Haswell" Quad-Core CPU Benchmarked, Compared Clock-for-Clock with "Ivy Bridge"

[iLLz]

Just stop. I don't think these are legit at all. Anandtech's Haswell Architecture article would leave me to believe we can expect 10-20% IPC increase depending on workload.

Also this: https://twitter.com/FPiednoel/status/296459612377468928

Edit: Also check out some of his other tweets in his timeline, he clearly states that if you have a "healthy" Haswell, under no circumstance will it be slower than IVB. He is Principal Engineer / Performance Architect at Intel so I am inclined to believe him.

[TheMailMan78]

Well this old 2600K isn't going anywhere, anytime soon I guess. Mobo.......that might be a different story. Looking forward to a Cadaveca review to judge.

[phanbuey]

Haswell is for power efficiency at the same performance, with a lower thermal envelope. I am sure if they wanted to, for the same TDP they could have crammed 4 more cores in there then it would have whooped the IB in wprime.

[iLLz]

One thing to note guys is Haswell's main goal is to increase single threading workloads as well as lower power consumption. According to the twitter feed I posted earlier from the Intel Architect, single threaded performance helps with the perceived snapiness. And don't forget if the single threaded performance is better, multicore performance increase as well.

[cadaveca]

Originally Posted by TheMailMan78

Well this old 2600K isn't going anywhere, anytime soon I guess. Mobo.......that might be a different story. Looking forward to a Cadaveca review to judge.

I hope I get a chip before the launch.

I don't really think Haswell be be all that bad, but it strikes me as odd that they'd go back to Nehalem clock domain design for Haswell, unless they want OC to be more interesting. The only reason they'd invest that effort, and further complicate the chip, to me, is because clock-for-clock, Haswell is a bit worse off.

Many parts of internal cache has been doubled, L3 is changed too, lots of other changes...Haswell should be good...better than IVB...but for me it is too early to even guess what those improvements will bring. I'd be happy with exact same performance, faster ram clocks(to help cache subsystem), and lower power draw. IVB is pretty damn fast as is.

[brandonwh64]

Originally Posted by TheMailMan78

Well this old 2600K isn't going anywhere, anytime soon I guess. Mobo.......that might be a different story. Looking forward to a Cadaveca review to judge.

Why change boards if you do not OC?

[cadaveca]

Originally Posted by brandonwh64

Why change boards if you do not OC?

His current board lacks proper Windows8 drivers?

[TheMailMan78]

Originally Posted by cadaveca

His current board lacks proper Windows8 drivers?

This. Windows 8 just doesn't seem to like this board and as always Asus likes to forget the last generation of boards.

[brandonwh64]

Originally Posted by cadaveca

His current board lacks proper Windows8 drivers?

Ahhh good point. Same with my WLAN drivers on windows 8

[Samskip]

Originally Posted by Ikaruga

If I have to guess, I think it will be about much lower power consumption instead, since that's the area Intel is focusing on the most lately.

Am I seeing this wrong? The 3770K has a TDP of 77w and the 4770K will have a TPD of 84w.
That's not really what I call lower power consumption. Or does Intel only mean the mobile ones?
Kinda weird.

[Covert_Death]

Originally Posted by ...PACMAN...

You're wrong, clock to clock is the best way to benchmark and show differences in the architecture of two different chips at the same speed. It's all about IPC.

I understand where you are coming from with regards overclocking headroom but that normally forms part of a FULL review also.

Needless to say, my FX 4100@3.6 is miles behind a 2600K@3.6

its the ONLY point to doing clock to clock though ( to see arch differences). but if your trying to determine the better CPU you must do STOCK vs STOCK or MAX OC vs MAX OC

[jihadjoe]

Originally Posted by Samskip

Am I seeing this wrong? The 3770K has a TDP of 77w and the 4770K will have a TPD of 84w.
That's not really what I call lower power consumption. Or does Intel only mean the mobile ones?
Kinda weird.

Power regulation is included in Haswell, so it adds a bit to the TDP. But that's a component that isn't on the motherboard anymore, so total platform power consumption should still be less than Ivy.

[Darkleoco]

Guess the 2600K will stay where it is for quite some time, might be time to get a top of the line motherboard though.

Power Consumption be damned when I don't get to use my desktop half the year.

[Aquinus]

Originally Posted by Darkleoco

Power Consumption be damned when I don't get to use my desktop half the year.

Power consumption be damned anyways. I didn't get a SB-E chip to sip power. I bought it to shovel power down like a little kid with a bucket of candy... all year round... 24/7.

All things considered though, if Haswell lowers power consumption even further I might be in the market for a new laptop after Haswell mobile CPUs start becoming mainstream.

[erocker]

Pretty sure this is fake.

[Jstn7477]

I'd love to pick up a Haswell soon. Power consumption really matters for distributed computing since the machine is under full load 24/7, and I'm glad that they keep advancing the CPUs in that way even if performance increases are minimal. It's why I've stopped buying AMD processors, because their power consumption is rather high and the performance is lower in many cases. Sure, the upfront price is "cheap," but you end up paying for it with your electric bill and cooling costs especially if you live in a hot region.

[Darkleoco]

Originally Posted by Aquinus

Power consumption be damned anyways. I didn't get a SB-E chip to sip power. I bought it to shovel power down like a little kid with a bucket of candy... all year round... 24/7.

[happita]

I won't mind if performance isn't increased too much because if I can compare my current SB setup to a future Haswell setup, I'm sure the performance bump will be there regardless just because of the fact that it is on a smaller manufacturing process coupled with the fact that it will draw less power (higher OC potential). However, I do wish that they release different iterations of Haswell without the graphics part of it on the die.
A 4770S will be perfect for a modern HTPC that I plan on building for my HDTV in the near future.

[sergionography]

Originally Posted by qubit

So there's almost no performance boost with a "brand new" architecture?

Nice to see AMD providing stiff competition to Intel.

We'll only gain if it clocks higher and has a proper soldered heatspreader and that remains to be seen.

from my understanding of the haswell presentations alot of the focus was on the new voltage regulation and how haswell can scale much better to different tdps, ivy bridge wasnt designed with 22nm in mind it was just a die shrink, with haswell intel can use all the bells and whistles of the process node

Originally Posted by Ikaruga

If I have to guess, I think it will be about much lower power consumption instead, since that's the area Intel is focusing on the most lately.

its more than just power consumption, its scaling. ivy bridge brought much better performance/watt over sandy bridge due to the shrink but couldnt scale any higher than sandy did because its the same architecture

Originally Posted by Aquinus

A number of confirmed changes to Haswell could support this. Intel definitely is playing the power consumption card and they're going to beet it to death. Intel's CPUs are plenty fast already. I think they're working on the easier things to improve at this point because you can only get clock speeds and your IPC so high before you run into the diminishing returns problem.

If Intel can get a CPU to consume less power but perform just as well, that's a win.

and this is what amd saw when designing bulldozer, except they jumped ship a bit too early, and this is when their winner multicore scaling design starts to pay off, intel on the other hand working on their strengths and making all new instruction sets multicore ready

Originally Posted by Easy Rhino

am i missing something? take cpu A at 2.8 ghz. take cpu B which can do much faster than that and bring it down to the speed of cpu A. how is that a good comparison of the two cpus? After all you are spending your money on what the processor can do... It's not like i am going to buy cpu B and downclock it and then act disappointed at the results...

no one here is disapointed at the result, only at the progress if this is true, but like many mentioned here this doesnt tell us anything about power consumption and tdp because that sure is very important, ivy brought substantial performance/watt over sandy bridge, but enthusiats didnt benefit much because ivy didnt scale well, there is a reason the highest ivy bridge is rated at 77w and not 95 like sandy, and no its not the new standard, its because the architecture doesnt scale well to higher tdps without running into binning problems or what not, the only way to get there is by just adding more cores, and what was impossible with sandy bridge(8 core cpu) might now be possible with ivy bridge extreme when it comes out, maybe there people will realize the benefit of performance/watt
haswell on the other hand if i intel isnt bluffing is supposed to have a much wider range of operating voltage, that doesnt only mean lower tdp like some understood it, but also higher tdp for the higher voltage models, that means higher stable clocks without running into problems

Originally Posted by cadaveca

Not in this instance. For IVB, CPU cache speed is directly linked to core clock...they run the same speed.


SO by downclocking an IVB chip, you are not reporting actual performance. you are reporting a gimped performance, with L3 running at a lower speed than intended.




Haswell breaks this design, and has L3 clocked independently, so C2C compare at low clocks doesn't tell you anything, but what a broken IVB does vs a non-broken Haswell.



Which makes this compare stupid, and that's why it was allowed. It's not a "real" performance compare.

so are you suggesting ivy bridge has really bad performance scaling with clockspeed? :P well this is exactly why haswell took the other approach, so no its a fair comparison, as fair as it can get actually because not every ivy bridge chip is sold at 3.4ghz clockspeed

Originally Posted by Samskip

Am I seeing this wrong? The 3770K has a TDP of 77w and the 4770K will have a TPD of 84w.
That's not really what I call lower power consumption. Or does Intel only mean the mobile ones?
Kinda weird.

i second what i said earlier, there is a reason the highest quad core ivy is rated at 77watt and not 95watt like sandy, its because clocking it accordingly to the 95watt envelope will pretty much get it near its limit and run into binning and stability problems for comfort, not to mention dissapointment for overclockers(who already saw no benefit from moving from sandy) because of a cpu clocked near its limit, haswell is supposed to address that issue

[Patriot]

431s for wprime 1024m ... Darn thats slow.


I can do it in ~30s

[Ikaruga]

Originally Posted by Samskip

Am I seeing this wrong? The 3770K has a TDP of 77w and the 4770K will have a TPD of 84w.
That's not really what I call lower power consumption. Or does Intel only mean the mobile ones?
Kinda weird.

I'm aware of those numbers, I just made a guess about how things will go with the power consumption perhaps

[Darksword]

I was hoping for a better IPC increase.

Guess my 4.0GHz i7-930 will stick around for whatever comes after Haswell.

[TheGuruStud]

If the real results are anywhere close, I want an apology from every fan boy in the world claiming 40% increase from nehalem (like there was anything from nehalem to IVB LOL).

Originally Posted by Darksword

I was hoping for a better IPC increase.

Guess my 4.0GHz i7-930 will stick around for whatever comes after Haswell.

That's what I've been recommending for everyone. Either that or just upgrade to IVB now if you want a newer CPU/ MB b/c haswell isn't bringing anything new.

[esrever]

Originally Posted by Patriot

431s for wprime 1024m ... Darn thats slow.


I can do it in ~30s

you sure are good at math

[Jack1n]

Originally Posted by Samskip

Am I seeing this wrong? The 3770K has a TDP of 77w and the 4770K will have a TPD of 84w.
That's not really what I call lower power consumption. Or does Intel only mean the mobile ones?
Kinda weird.

TDP stands for Thermal design power,it means how much heat the chip creates,if i remember correctly the 4770k is slightly higher clocked which could account for the higher TDP,although i suspect temp will be lower if Intel does the IHS proper this time around.