First "real" Haswell CPU preview, 4670K

[WhoBeDaPlaya]

I see this becoming a meme in a matter of weeks

 

[Borealis7]

"What are you machine spec again? and Can it run Crysis?"

 

[aigomorla]

I see this becoming a meme in a matter of weeks

I used to be king of this game. :biggrin:
Can i be a guest host of it now since i am retired?

 

[Grooveriding]

The 4670K is the 3570K replacement being reviewed right ? I need one of these for my wife's machine. It looks like the results were pulled though, was it the expected 5-10% over a 3570K ?

 

[tential]

And thats exactly what i would like to know, if the CPU will grand me any gains at the resolution I and most of the people play and thats at 1080p. Bench the games at 720p but also show us if we will get any gains at 1080p as well(obviously not with low settings).

Would you consider upgrading a CPU+Motherboard if you would know that you will only gain a 2% at 1080p ??

I see your point.

It's why I find most reviews to never really give me the information I need. But then again, you might as well ask them to test every single combination out there. It's a hard life haha.

 

[Magic Carpet]

I see your point.

It's why I find most reviews to never really give me the information I need. But then again, you might as well ask them to test every single combination out there. It's a hard life haha.

Most people aren't as much bothered with it. I have friends who dont give a damn about these things, as long as their games run well. Sometimes you do get a feeling that you need above average close to most expensive gear to play games. That's of course not always the case.

 

[galego]

That's why it says "up to". Pretty much how the supposed FX scheduler hotfix was "up to 10%" but proved to be 0.5% all around (). Except in this case the all around benefits will be around 25%

What are your machine specs again?

There is no "up to" in the poster whom I was replying. Good that you mention the FX hotfixes. They provided up to 10% performance gain in some game, 5% in several games, 2% in some...

Average gain depends of the applications selected, but 2-5% gain is normal. This is far from the average 1-5% gaming gain provided by Haswell over Ivy.

 

[Enigmoid]

There is no "up to" in the poster whom I was replying. Good that you mention the FX hotfixes. They provided up to 10% performance gain in some game, 5% in several games, 2% in some...

Average gain depends of the applications selected, but 2-5% gain is normal. This is far from the average 1-5% gaming gain provided by Haswell over Ivy.

For 8150

Average over windows auto updates is ..... -1%. Windows 8 is slightly ahead and more efficient. But overall it looks like the hotfixes were a step backwards.

The only gaming test where there is a substantial difference is dirt showdown. However, the hotfixes reduce performance (1080p) by ~10% over windows updates!

Sometimes it appears to be more

But that review is dated Oct 2011 before any auto-updates have occurred.

With auto updates. (There was a wow expansion but I don't think it changed this part of the game, correct me if i'm wrong. Frame rates appear to be lower in MOP so the actual gain might be greater than directly seen).

Anyway, fps has improved from 71.5 to 77 fps.

 

[Grooveriding]

http://www.chinadiy.com.cn/html/42/n-9142-3.html

Hopefully that is not reflective of performance. 1.45V for 4.8Ghz and it hits 100C temps.

 

[Enigmoid]

http://www.chinadiy.com.cn/html/42/n-9142-3.html

Hopefully that is not reflective of performance. 1.45V for 4.8Ghz and it hits 100C temps.

Stock cooler.

 

[Mallibu]

Average gain depends of the applications selected, but 2-5% gain is normal. This is far from the average 1-5% gaming gain provided by Haswell over Ivy.

Are you claiming the FX hotfix which was +1% performance is a bigger upgrade than IB -> Haswell?
I've seen bias, but this is straight stupidity.

 

[Grooveriding]

Stock cooler.

Not so much the temps as the voltage. These chips won't be able to take that voltage. We're probably looking at 1.3v max safe rating or a little less. Still, it's just one sample and maybe he just cranked voltage with no adjustments.

 

[Borealis7]

mmm...stable 175MHz BCLK looks yummy!

 

[AtenRa]

Translation : Coming in 604,800,000,000,000 nanoseconds

 

[VirtualLarry]

mmm...stable 175MHz BCLK looks yummy!

Is the BCLK on the K models unlocked in 1Mhz increments, or does that depend on the mobo? Can you OC with BCLK on non-K models?

 

[ShintaiDK]

Is the BCLK on the K models unlocked in 1Mhz increments, or does that depend on the mobo? Can you OC with BCLK on non-K models?

BLCK straps, most likely K models only. 100, 125 and 167Mhz.

 

[WhoBeDaPlaya]

Is the BCLK on the K models unlocked in 1Mhz increments, or does that depend on the mobo? Can you OC with BCLK on non-K models?

That's the million dollar question. If the 4770 can do BCLK (straps/increments), I'm getting it instead of the 4770K. Nothing like saving some $$$ _and_ getting a superior CPU (+ TSX)

 

[galego]

For 8150

Average over windows auto updates is ..... -1%. Windows 8 is slightly ahead and more efficient. But overall it looks like the hotfixes were a step backwards.

They are comparing Windows 7 plus hotfixes installed manually to Windows 7 with auto-updates:

The last time we looked at the impact of Microsoft's hotfixes was almost a year ago. In today's comparison, our baseline Windows 7 machine is loaded with a lot more patches from Windows Update, and they cumulatively seem to have a larger impact than the two manually-installed tweaks specific to AMD's Bulldozer architecture. In fact, overall performance is better without the hotfixes applied.

Either the auto-updates include improved versions of the hotfixes or include some other improvement on the W7 scheduler. That is why there is a 1% improvement over Windows 7 with the hotfixes installed manually. Windows 8 includes improvements on the scheduler beyond those in the pair of hotfixes for W7.

Performance goes as

W8 > W7 + automatic patches > W7 + manual patches > W7

 

[Enigmoid]

They are comparing Windows 7 plus hotfixes installed manually to Windows 7 with auto-updates:



Either the auto-updates include improved versions of the hotfixes or include some other improvement on the W7 scheduler. That is why there is a 1% improvement over Windows 7 with the hotfixes installed manually. Windows 8 includes improvements on the scheduler beyond those in the pair of hotfixes for W7.

Performance goes as

W8 > W7 + automatic patches > W7 + manual patches > W7

You still don't get it. THERE ARE NO WINDOWS 7 MACHINES WITHOUT THE UPDATES unless

1) Its pirated (and they deserve whatever they get).

2) You have either disabled updates or not hooked up the computer to the internet in a year. (I'm assuming there are very few gamers doing this).

Every valid win 7 machine has received the autoupdates; there is no baseline win 7 without updates. All the reviews on the internet include an up to date copy of windows and so are technically benchmarking under the windows 7 + autoupdates category (note that article is months old so other updates could have improved performance to win 8 levels). Now compared to the old windows 7 before updates the hotfixes caused a small performance gain, however that has already be made irrelevant due to the fact that windows 7 has received updates that has improved performance beyond what the hotfixes did by themselves.

So we can conclude

1) The hotfixes do nothing for performance compared to an up-to-date machine and increase power use. (You said it there so please do not bring this up in another thread. Hotfix == BAD).

2) Windows 8 is slightly better than up-to-date windows 7.

 

[tweakboy]

Agreed with the iGPU thing. Unplayable. But they wanted to make sure to have a GPU bottleneck with the iGPU and a CPU bottleneck when testing the CPU part. So it makes sense, especially if you just want to find out the performance improvement over the predecessor.

I get 18.6 fps in the x264 benchmark with a 3930K@4.2 GHz, hyperthreading off.
Fits perfectly. If Haswell overclocks better than my Sandy-E, I will replace it.

Why on earth would you shut off HT. It gives 15 percent off the bat more power.....then if you use a app that understands HT then that 15 percent is gonna be 35 or 40 percent. gl

 

[galego]

You still don't get it. THERE ARE NO WINDOWS 7 MACHINES WITHOUT THE UPDATES

I wrote "patches"

W8 > W7 + automatic patches > W7 + manual patches > W7

the rest of stuff included in the updates was assumed in "W7" and "W8".

Anandtech has a list of gains offered by the manual FX hotfixes, several games showed 5% gain. The maximum obtained by improving the scheduler in W8 was 10%.

 

[guskline]

galego, have you personally used either a Bulldozer 8150 or a PilerDriver 8350 machine? A simple yes or no will do.

 

[Enigmoid]

I wrote "patches"

the rest of stuff included in the updates was assumed in "W7" and "W8".

Anandtech has a list of gains offered by the manual FX hotfixes, several games showed 5% gain. The maximum obtained by improving the scheduler in W8 was 10%.

The patches do not improve performance over an up to date win 7 computer. Before updates they did but not since the updates were rolled out. Since now the updates are better than the patches you CANNOT measure performance now on a up to date computer (basically the test systems of reviews) and then add supposed gains for hotfixes.

 

[WhoBeDaPlaya]

galego, have you personally used either a Bulldozer 8150 or a PilerDriver 8350 machine? A simple yes or no will do.

IDC is right, it is starting to catch on

 

[Sleepingforest]

IDC is right, it is starting to catch on [snip image]

Guskline, the dogged, determined man that he is, has been pursuing this question across something like 5 threads and 2 months now. :thumbsup: He deserves a medal for persistence in the face of being completely ignored nearly every time (I think he got a tangential sarcastic response back once).