Intel Core i7-7700K CPU Synthetic Benchmark Sneak Peek @ [H]

I was going to ask if they worked in existing Z170 boards, but after reading this article I don't care, even slightly. I knew this release was geared towards lowering TDP per clock speed so they could up the notebook chip speeds a bit at the same TDP but this is essentially no improvement on the desktop at all.

Are we heading in to a period where computers aren't getting any faster? Is that it? I guess that's good for my wallet if nothing else...
 
Brackle said:
Just curious if you even heard anything from the CPU department of AMD about possibly getting a review sample? Just don't mention Rory and I think you might get one!
Click to expand...

Hey Kyle,


Sorry for delay. Yes we do have plans to keep you and your team up to date on all our items on processor side of the fence.


We will keep you posted on all details as they come available. You have any questions let me know.

Now what exactly that means in terms of timeliness and samples, I am not quite sure of. That said, I nor no other HardOCP editors were invited to the next event. Of course it is going to stream live so there is that. Sampling is the lynch pin, unless we get our own out of China and start posting review material before embargo.
 
HeadRusch said:
Who knew then we'd be building our last PC's.... at this point the only reason any of us will have to upgrade will be the motherboards finally burning out, caps popping, or driver updates drying up.
Click to expand...

Tell me about it. I'm still on my i7-3930k I bought in the end of 2011 on launch, and while I am considering upgrading in the next 6 months, it has absolutely nothing to do with performance.

It hits 4.8Ghz, and gives me a Cinebench 11.5 single threaded performance of 1.92, which places me in pretty good company, even today, 5 years later.

It just gets so damned hot. A 32nm chip at 1.45v and 4.8Ghz can double as a space heater.

So, yeah, I'm looking to upgrade for:

  • Temperature/fan noise
  • Modern features (NVMe, proper validated PCIe Gen 3 support*, etc.)
  • Because the plastic bits on my motherboard are starting to get brittle and snap off

(*x79 and Sandy-Bridge-E got unofficial PCIe 3.0 support, but there was no PCIe 3.0 hardware to validate it with at the time, so it was never validated, and to this day can suffer minor timing issues in PCIe 3.0)

I miss the days when I was on the "new CPU every 8-12 months, New Motherboard every 12-18 months. new GPU every 6 - 8 months" plan back in the good old days when hardware actually moved forward at a decent pace.
 
Burticus said:
Wow remember when we all bitched about 10% upgrades? Intel has scaled it down to 1%. Progress!
Click to expand...
This honestly reeks of opportunism. They don't have a true upgrade for at least another year and are hoping a name change and slight clock speed upgrade will trick people into upgrading.
 
drescherjm said:
Wasn't there an update the next day that showed this to be related to 1 specific motherboard?
Click to expand...

I stand behind our numbers 100%. These results are NOT specific to this motherboard and there has not been an i7 microcode update in about 2 weeks. We might see another update between now and launch, but the fact of the matter is that there is no possible way for it to significantly change these results in my opinion.
 
I've been holding out on upgrading 3570k because I thought Kaby Lake would be a game changer. Now it looks like I'll just wait for the 6850k to drop so I can get me a x99 setup.

Hopefully I won't be as disappointed once we see 1080ti. I desperately need to upgrade this gtx680... even though I can still run most games fine.
 
Last edited:
Why did they even bother? I've waited this long, i might as well just hold out for mainstream 6 or 8 core i7's whenever that happens. Intel isn't seeing a dime from me until then, money better spent on gpu in any case.

Hopefully amd delivers on zen as i do plan on building many budget pc's in the near future.
 
Honestly, we should all know by now when Intel abandoned the Tick-Tock model (hinted July 2015, official March 2016) that we would be receiving an optimization of the same architecture. Search your feelings; you KNOW it to be true! In terms of respect for AMD, I lost that 8-10 years ago. Here's to building with 4790K or 6700K. Cheers.
Tik-Tok10.jpg
vader.jpg
 
So when haswell got updated we got devil's canyon and a very small SKU update 4770k to 4790k. Back then Intel promised 5ghz on water. Fast forward to now and they jump a complete generation with 0% improvement and have the balls to sell it as new. Even the 390x had 4gb of extra RAM from the base SKU. This is rebranding perfected. Also why would Skylake prices go down if the sellers know the performance is there, prices will stagnate.
 
Golly Gee! I guess i will be dooping the Bulldozer for another 2-3 years. . .
 
d50man said:
Zen will be slower than what you are running now 4770k will do it in games.
Click to expand...
If Zen hits its rumored performance numbers it'll actually be nearly identical in games. All we can do is wait and see at this point if it lives up to the hype.
 
Bahanime said:
Which according to the NDA leak is using more power than Skylake at the same clocks. WTF is Intel doing..
Click to expand...
They're choosing to improve the iGPU more over CPU performance.

If the iGPU actually performed decently it'd be nice but realistically given the die sizes and TDP's they have to work with + the main system RAM limitation of dual channel DDR4 only this is probably impossible. So IMO its mostly wasted effort and given the reactions in thread so far most people would probably agree.
 
  • Like
Reactions: blkt
like this
mesyn191 said:
If Zen hits its rumored performance numbers it'll actually be nearly identical in games. All we can do is wait and see at this point if it lives up to the hype.
Click to expand...

Nope. I've done the calculations over and over. If they meet the "excavator +40% IPC" claims, which is the only official word we have except for those BS blender benchmarks, IPC will be slightly behind Sandy Bridge. Overall performance may be better, depending on where the clock speed winds up.

Depending on final clocks , we are looking at something that is comparable to anywhere in between a Gulftown 990X and an Ivy Bridge 3770K.

Anything higher than a 3770K per thread, is highly unlikely.
 
mesyn191 said:
They're choosing to improve the iGPU more over CPU performance.

If the iGPU actually performed decently it'd be nice but realistically given the die sizes and TDP's they have to work with + the main system RAM limitation of dual channel DDR4 only this is probably impossible. So IMO its mostly wasted effort and given the reactions in thread so far most people would probably agree.
Click to expand...
iGPU has more to do with the their mobile business, and laptop makers are desperate for a real performance boost which has been lacking.
 
Zarathustra[H] said:
Nope. I've done the calculations over and over. If they meet the "excavator +40% IPC" claims, which is the only official word we have except for those BS blender benchmarks, IPC will be slightly behind Sandy Bridge. Overall performance may be better, depending on where the clock speed winds up.

Depending on final clocks , we are looking at something that is comparable to anywhere in between a Gulftown 990X and an Ivy Bridge 3770K.

Anything higher than a 3770K per thread, is highly unlikely.
Click to expand...
Seems a bit uncharitable. Even AMD knows there is no way they are going to compete with performance that many years behind.
 
Zarathustra[H] said:
Nope. I've done the calculations over and over.
Click to expand...
No offense but I wouldn't trust your numbers at all there.

Presbytier said:
iGPU has more to do with the their mobile business, and laptop makers are desperate for a real performance boost which has been lacking.
Click to expand...
Sure but that doesn't matter here since Intel isn't providing much of a iGPU boost and they couldn't even if they really wanted to for the reasons I gave.

They're going to have to go back to something like Crystalwell style on package cache or go to quad channel DDR4 to fix the bandwidth problem and then they'll have to give either more of the die share to the iGPU OR increase TDP's dramatically OR some combo of both to get the real dramatic performance boost needed to make the laptop guys happy.
 
Ran that test on my setup and got 90 percentile. Score 4411. CPU and memory got 4 out of 5 stars. Which is a bit disappointing. If people really need an upgrade looks like the non k 7700 would be the best bet. Price vs performance.
 
Looks like my Z97/4790k gonna soldier on a good bit more.

Runnin' this thing at 4.7, don't really need a platform update, do want a 6 core HEDT chip, don't have any serious use cases for that kind of power...


Awesome :x
 
After reading this I'm not expecting much, but I would like to see Handbrake comparisons using both CPU encoding as well as QuickSync. I'd like to see if QuickSync is getting better.

Additionally I'd like to see benchmarks of the GPU. Is this using the eRAM or or whatever the integrated DRAM was on the previous IRIS GPU's?

It's obvious the CPU performance is nothing worth talking about, but there may be some other aspects worth investigating.

But otherwise these benchmarks are with in "noise" differences. 1% difference? Who cares?
 
westrock2000 said:
encoding as well as QuickSync
Click to expand...
Know any easy Quick Sync apps to encode and decode with that we can use. I am not experienced with it, but if there is something easy to run, I am down for it.
 
AMD stated a 40% IPC improvement quite some time ago. I don't buy that it really is 60% to 70% or even more.
 
Last edited:
Presbytier said:
Seems a bit uncharitable. Even AMD knows there is no way they are going to compete with performance that many years behind.
Click to expand...

Well, We are in wait and see mode now, so nothing is for certain, all I'm saying is that if they meet their own published figures, that is where they will fall in final performance.

I cant make any assessment as to how likely it is that they actually meet those figures, or even exceed them, all I can do, is take excavator IPC and multiply by 1.4 and see where we land, and that's what I did.

I certainly hope they blast through their expectations and hit one out of park, but based on the little credible information we have right now, I'm not too enthusiastic.

What I will say is this. If Zen overclocked on water can beat my Sandy-E [email protected] on water in single threaded benchmarks (in other words, a 1.92 in single threaded Cinebench 11.5), I'll probably buy one. If it can't I probably won't.

I want to support AMD, and I'd really like something that runs cooler than this heat monster, but I can't bring myself to do a per core performance downgrade, regardless of how many cores I get. I'd probably even get one if it is a per core performance side grade, but I can't do a downgrade.

According to my calculations (again, Excavator IPC *1.4) this means that Zen will have to hit 5.1Ghz to beat my [email protected], and I somehow don't see this as being all that likely, unfortunately.

Call me a pessimist if you will, but I go by numbers, not by feelings, and this is what the numbers are saying right now.
 
40% should be an average. Meaning that some task the IPC improvement will be less and some much more.
 
Zarathustra[H] said:
Call me a pessimist if you will, but I go by numbers, not by feelings, and this is what the numbers are saying right now.
Click to expand...
I like to go by numbers too but getting the correct ones can be real tricky. Especially for IPC which is a broad average. If you want to just use a handful of bechmarks or applications and multiply the numbers by 1.4 you'll probably be off by a fair amount.

Given the Cinibench numbers released a while back I think there is good reason to be at least a bit optimistic here. Other than being more multi threaded friendly than most software there wasn't anything particularly cheezy about using it to showcase Zen IMO.
 
Presbytier said:
This honestly reeks of opportunism. They don't have a true upgrade for at least another year and are hoping a name change and slight clock speed upgrade will trick people into upgrading.
Click to expand...
I'm not sure I'd call it a trick. When they announced they were delaying Canonlake (or whatever the original follow up was), we kinda knew it wasn't going to be a big change. I was certainly hoping for better than what we've seen so far, but I wasn't expecting to upgrade Skylake and if I had a 47xx chip, I doubt I would have gotten skylake. Hell, I probably could have stuck with my 860, but after almost 6 years, i felt it was time for an upgrade.

I hope AMD gets competitive, but it seems like a decent possibility that this is as good as it gets at 14nm
 
AMD will get Intel to do what they already planned. To have a 6 core mainstream CPU in 2018. I am not sure anything more than that with AMD having lower IPC and likely lower clocks than Intel.
 
I get the feeling that Intel may have held a bit back with this one as a just in case.

Zen perhaps gives them a run in the mid range and perhaps even at the top tiers. Then again perhaps not. Still I feel Intel may have held just a bit back so that if needed they can drop another chip in the spring to claim the title of benchmark winner if need be.

From what I'm reading though... damn even if your a Intel fan boy you best be preying for AMD. If not we are going to get CPU release after release that each offer 1-2% more performance and a new DRM. lol
 
I'm waiting on Zen benchs.
Hopefully they come close.
This is not an upgrade for me.
 
I would honestly be shocked if AMD can beat the most advanced semiconductor company there is. No one has the resources, physical or financial, that Intel does.

I know they did it once before, but Intel got themselves into a very unfortunate position at that time and payed dearly for it. They are not in a bad place right now.
 
I bought a 6600k a Gigabyte z170-hd3 and gskill 3200 ram over the black friday weekend.
I wanted to wait and see what zen and kaby lake could do but I could not resist the deal I had on cpu + mobo and my sister will get my 3570k/mobo/ddr3 ram.
But seeing this I'm happy I didn't wait especially since intel/resellers usually milk the early adopter.


Now If I could only received the mobo that took forever to ship for amazon and an eternity to get to me via Canada Post I'll be happy!
 
westrock2000 said:
They are not in a bad place right now.
Click to expand...
If they do its because Intel let them by doing little to improve IPC or clockspeed for 3-4 generations.

Bad business decisions can screw a company just as thoroughly as technical ones.
 
just a heads up. Not sure if this should even be here, but Kyle just broke the news that Intel and AMD came to an agreement for Intel to license using AMD GPU tech in Intel iGPU's in the future....

So I guess Kaby Lake having a beefer iGPU won't matter in the future? This really makes Kaby Lake that much more horrible of a purchase over Skylake.
 
This is really disappointing. I was only expecting 5% IPC this year and 1% next year. The main feature that Coffee Lake is supposed to have over Kaby Lake is 4-6 cores and twice the iGPU execution units, so if you were hoping for any performance improvement, expect to wait 2 years for the next architecture.
 
You must log in or register to reply here.
Back
Top