8700k benchmarks leaked.

[ravenshrike]

19 minutes ago, Dash Lambda said:

.. Huh?

And where is that information from exactly? I have trouble imagining Zen2 running at 6Ghz.

He meant IPC not clocks. GloFo is saying that you can get a 40% performance boost at the same efficiency or a 60% efficiency boost at the same performance in their die shrink from 14-7nm. I figure knock 10% at least off of each of those for real world numbers. It remains to be seen whether these are real performance gains or Intel style performance gains however.

[ravenshrike]

1 hour ago, Brooksie359 said:

How do you figure? It would be more like comparing the performance difference between a r5 1600 and an i7 7700k. The 1700 still has 2 cores on the 8700k. 

So look at the relative performance of the 1600X and 1500X, then apply the same delta percentage you see between the 1500X and 7700k to the 1600X. Should get pretty close numbers. Then, for when Intel's X4xx lineup comes out, you can assume the same relationship between the 8 core part and the 1800X.

[Coaxialgamer]

2 hours ago, MageTank said:

Maybe now the forum will finally believe me. Been saying there was going to be no IPC increase with CFL, and people kept throwing out "7% higher IPC" just because Intel said it was "7% faster". Intel also said Kaby was 13% faster than Skylake, lol.

 

I see Intel did pull the 8700k's base clock back by 100mhz (initial info was 3.8ghz base), they might have had a difficult time hitting TDP ratings otherwise. Either that, or this ES is 100mhz slower than what will launch, but it doesn't matter. I completely expect 4.8-5ghz to be possible with these chips due to the slightly larger die making heat transfer easier (assuming your cooling is not the bottleneck).

 

As I said before, if you already own a 6700k or 7700k, and are not in need of additional threads, you have no need to worry about upgrading to CFL. Your CPU will be plenty fast for gaming, and you didn't make a terrible purchase decision just because Intel came out with more cores later. As long as you feel you got your money's worth, then that's all that matters. 

That was also my belief ever since coffee lake was announced. People were expecting intel to pull IPC improvements overnight. 

Given how recently it was announced, there was no way intel could have done anything more than a refreshed kaby lake. 

[mr moose]

Think I'll just wait until I see this CPU in the real world and know what it will actually sell for before I decide if it is "pointless", "worthwhile" or even "just an excuse to justify skylake x".

 

 

[ravenshrike]

2 minutes ago, mr moose said:

Think I'll just wait until I see this CPU in the real world and know what it will actually sell for before I decide if it is "pointless", "worthwhile" or even "just an excuse to justify skylake x".

Kaby Lake was the excuse to justify Skylake X. Coffee Lake is what Kaby Lake should have been in the first place.

[dexT]

1 hour ago, Dabombinable said:

This is only good news to me if my 4790K is really dead (woke my computer up from standby, it woke for a second then died-and I was running everything at stock)

I'll test it for you if you wan't. 

[mr moose]

3 minutes ago, ravenshrike said:

Kaby Lake was the excuse to justify Skylake X. Coffee Lake is what Kaby Lake should have been in the first place.

If that's what you want to believe. 

[Prysin]

4 hours ago, DocSwag said:

It is? I was always under the impression that Intel was planning for MCM with 10nm, with that being the main innovation.

Ice lake is supposed to be a new architecture. Maybe a large redesign of the "core" architecture 

[Dabombinable]

41 minutes ago, dexT said:

I'll test it for you if you wan't. 

I'll just have to test it inside the rig I built for my friend, since its got everything that was replaced by the 4790K+Z97 Sabertooth MKII. Its got to be the CPU though as I could actually get the motherboard to give out a beep before it'd go back to not doing anything (apart from powering up for a second), the RAM is fine (in fact only 1 stick isn't brand new), and the PSU is powering my extremely temperamental LGA775 machine (its really picky when it comes to RAM+the slots where expansion cards/graphics card are installed).

[D13H4RD]

I’m going to have to adopt the “wait and see” approach.

 

Whenever a leaked benchmark shows and it’s positive, a lot of us tend to get excited, which I can understand. For me, though, I prefer to wait until an actual 8700K officially arrives to reviewers and gets benched and tested.

 

I’m not expecting much from what looks like a refresh but I would be happy to be proven wrong. If AMD’s Zen+ ends up being the killer offering next year, it’s gonna be a pretty good time to be a PC enthusiast, especially if both companies have put their best into their products.

[ZestyJalapeno]

5 hours ago, jnkokseby said:

Pretty impressive in my opinion. 2 cores and 4 threads more and only 4W TDP increase, hope they fixed the thermal issues from the kaby lake series (you're nearly forced to delid it if you want fair temps)

Do they say at what frequency is that TDP is measured at? The 4/4 i3 which has a base of 4.0 GHz has a 95W rating, whereas the 6/12 which boosts to 4.3/4.7 also has a 95W TDP.  What?

[Brooksie359]

2 hours ago, ravenshrike said:

So look at the relative performance of the 1600X and 1500X, then apply the same delta percentage you see between the 1500X and 7700k to the 1600X. Should get pretty close numbers. Then, for when Intel's X4xx lineup comes out, you can assume the same relationship between the 8 core part and the 1800X.

I'm not even going to try and figure out what you are trying to say. I said that comparing the r5 1500x to a 7700k is not going to be the same as comparing a 8700k to a 1700. the reason being the 1500x and the 7700k are both 4c8t cpus but the 7700k has the better single threaded performance. the 1700 vs the 8700k on the other hand have more going on. the 1700 has 2 more cores while the 8700k has better single threaded so both have their strengths and weaknesses so i would assume the 8700k will likely equal the 1700 in multi-threaded performance and have better single threaded. now compare that to comparing a 1500x to a 7700k and you cna see they aren't similar at all. the 7700k beats the 1500x across the board which makes sense because it is the more expensive cpu. 

[Brooksie359]

1 minute ago, Dogeystyle said:

Do they say at what frequency is that TDP is measured at? The 4/4 i3 which has a base of 4.0 GHz has a 95W rating, whereas the 6/12 which boosts to 4.3/4.7 also has a 95W TDP.  What?

Intel TDP is a rough estimate at best. I would assume all of them are on the conservative side. 

[Flavio hc 16]

I would not believe even seeing  it that the 8700k has a real tdp of 95w, if it is just an impovement of the node of kabylake  ( wich at 99%it is)  i think we are looking at 120-130w of real tdp

[ZestyJalapeno]

Just now, Brooksie359 said:

Intel TDP is a rough estimate at best. I would assume all of them are on the conservative side. 

It's very sneaky from a prospective buyer's POV, thinking his premium 150W air cooler is enough for the 6/12, then it turns out you need a 280mm rad with Gentle Typhoons to take advantage of the unlocked multiplier.

[Brooksie359]

Just now, Dogeystyle said:

It's very sneaky from a prospective buyer's POV, thinking his premium 150W air cooler is enough for the 6/12, then it turns out you need a 280mm rad with Gentle Typhoons to take advantage of the unlocked multiplier.

well yeah TDP is always based on stock clocks

[Dabombinable]

Actually, wouldn't an old i7 5820K be better value?

[mr moose]

4 minutes ago, Dogeystyle said:

Do they say at what frequency is that TDP is measured at? The 4/4 i3 which has a base of 4.0 GHz has a 95W rating, whereas the 6/12 which boosts to 4.3/4.7 also has a 95W TDP.  What?

Intel determine the TDP from the base clock rate with a predefined workload (whatever that is). 

 

2 minutes ago, Brooksie359 said:

Intel TDP is a rough estimate at best. I would assume all of them are on the conservative side. 

No it's not, it is a defined spec. It has a specific meaning in EE,  That is until the Internet came along and turned it into a pissing contest.

 

 

 

 

 

 

[Brooksie359]

Just now, mr moose said:

Intel determine the TDP from the base clock rate with a predefined workload (whatever that is). 

 

No it's not, it is a defined spec. It has a specific meaning in EE,  That is until the Internet came along and turned it into a pissing contest.

 

 

 

 

 

 

I did not say TDP I said "Intel TDP"

[ZestyJalapeno]

5 minutes ago, mr moose said:

Intel determine the TDP from the base clock rate with a predefined workload (whatever that is). 

I see, so it means for someone who'll buy a K CPU along with a Z board, with the express intention of running it at Turbo frequency or OCd, the use-case power consumption is above 95W. How much by is left ambiguous. Lovely.

[DocSwag]

1 hour ago, Prysin said:

Ice lake is supposed to be a new architecture. Maybe a large redesign of the "core" architecture 

Is Ice Lake the 7nm one or the 10nm new arch? I don't remember  

[Dabombinable]

9 minutes ago, mr moose said:

Intel determine the TDP from the base clock rate with a predefined workload (whatever that is). 

 

No it's not, it is a defined spec. It has a specific meaning in EE,  That is until the Internet came along and turned it into a pissing contest.

 

 

 

 

 

 

Whatever the pre-defined workload is with CPU post Sandybridge, its got to the best case scenario by far as my 120W TDP Xeon X5450 (50-60W actual power consumption when stock) is handled perfectly by my 115W Freezer Pro 7 Rev 2.0, in fact the cooler handles it far better than the supposedly 88W TDP 4790K of mine which goes to show that Intel is not taking the TIM under the IHS into consideration in regards to the TDP.

Edit: BTW I'm currently using said Xeon under my CM Eclipse, and the cooler runs warmer and quieter than with my 4790K, meaning that heat is being dissipated far better from the Xeon.

[Brooksie359]

1 minute ago, Dogeystyle said:

I see, so it means for someone who'll buy a K CPU along with a Z board, with the express intention of running it at Turbo frequency or OCd, the use-case power consumption is left ambiguous. Lovely.

well its not like they could test that. overclocking is quite variable so thermals would be as well.

[ZestyJalapeno]

2 minutes ago, Brooksie359 said:

well its not like they could test that. overclocking is quite variable so thermals would be as well.

I'll just wait for Tom's or Anandtech to tell me what it really is and if I need to upgrade my cooling for the 8700k. I cant exactly fit a custom loop into my case and if it needs a delid for an OC then I'll buy the Zen+, not in the business of intentionally voiding warranty on expensive hardware.  Although I could hack my case to stuff a 360mm AIO in there...

[mr moose]

22 minutes ago, Brooksie359 said:

I did not say TDP I said "Intel TDP"

tdp is tdp, it's an industry spec just like horsepower or watts.  Intel do not "guess" the TDP, it is a measured and tested spec.

20 minutes ago, Dogeystyle said:

I see, so it means for someone who'll buy a K CPU along with a Z board, with the express intention of running it at Turbo frequency or OCd, the use-case power consumption is above 95W. How much by is left ambiguous. Lovely.

more or less,  there are too many parameters outside of Intel's control when enthusiasts overclock.   The idea of TDP is that pc designers can implement a suitable thermal solution for the conditions they expect, I.E if they are selling an overclocked machine they will know they need a cooler larger than the rated TDP to handle the output.

 

 

People seem to think that TDP is a concrete unchanging figure and that because the thermal output of an overclocked or overworked CPU is higher, then the TDP figure is wrong or "guessed", this is not the case, it is calculated, and if you are designing your PC to operate outside that definition then you need to understand how it is calculated.   I can only assume it is lucky aftermarket coolers are generally better than rated (marketed), because with the number of people who don't understand TDP we'd have a lot more people complaining about their CPU's unnecessarily throttling and not know knowing why.