Intel Core i7-8700K Put Through Cinebench R15

[btarunr]

Intel's upcoming Core i7-8700K six-core mainstream desktop processor was put through Cinebench R15, where it was seen trading blows with much higher high-end desktop (HEDT) processors, thanks to its core-count and relatively high clock speeds. Unlike HEDT processors, the i7-8700K doesn't carry a bulky uncore, keeping its TDP low at 95W, enabling high clock speeds. This reflects in its single-threaded performance, where it was significantly faster than older chips, some of which are even HEDT, but since the "Coffee Lake" architecture is essentially a refresh of the "Kaby Lake" architecture, the chip could lose out on single-threaded performance to the Core i7-7700K on account of slightly lower clock speeds.

The multi-threaded test is where the action is. Bolstered by two more cores, four more threads, and 4 MB more L3 cache, the i7-8700K is proportionately faster than the quad-core chips it succeeds, and is even faster than older 6-core HEDT chips thanks to higher clock speeds, and a newer micro-architecture. The i7-8700K features 6 cores, 12 threads enabled by HyperThreading, Intel's newest Turbo Boost Max 3.0 technology introduced with its Core X family, and 12 MB of L3 cache. It launches on the 5th of October, at an expected price of around USD $380, if not more.



A video presentation by YouTuber Karl - MrTechQc, who tested the chip, follows

View at TechPowerUp Main Site

 

[dwade]

Ryan, under NDA, have this to say:

 

[Totally]

Ohh, a mystery! Now is the score bad in a good way or just simply bad i.e. hype.

 

[Xzibit]

Ryan, under NDA, have this to say:

I dont like it when PC Tech review personalities say something Pro or Against a product prior to release yet hide behind the NDA. Stay shut and let your "comprehensive" review speak for itself. It just make one self look like a child at a play ground saying, "Nah ah".

His next tweet just goes to show he wasn't even paying attention

"Oh, was it at a fixed clock speed? Didn't see that part."

 

[drade]

Ryan Shrout doesn't know jack squat about coffee lake at this moment. Why retweet such nonsense? Has he had the opportunity to utilize coffee lake? Where are his benchmarks? My point exactly. This goes back to the pre Ryzen era. Hype, hype, hype. Wait till the product drops with the right MOBO combo, and then post some benchmarks.

 

[RejZoR]

I wonder how my 5820K stacks up against this...

 

[Patriot]

Ryan Shrout doesn't know jack squat about coffee lake at this moment. Why retweet such nonsense? Has he had the opportunity to utilize coffee lake? Where are his benchmarks? My point exactly. This goes back to the pre Ryzen era. Hype, hype, hype. Wait till the product drops with the right MOBO combo, and then post some benchmarks.

Wow... Jon Snow knows more than you.

He is a proper reviewer and has samples and is behind NDA. He probably knows a good bit about coffee lake.

for those talking about freq lock... cinebench only shows base clock not turbo bins... now it could be set to turbo 2.0 and not 3.0 in the bios...

 

[dwade]

Ohh, a mystery! Now is the score bad in a good way or just simply bad i.e. hype.

Most likely something is wrong. Even a lower clocked 7800x has higher score than this.

 

[Xzibit]

Wow... Jon Snow knows more than you.

He is a proper reviewer and has samples and is behind NDA. He probably knows a good bit about coffee lake.

for those talking about freq lock... cinebench only shows base clock not turbo bins... now it could be set to turbo 2.0 and not 3.0 in the bios...

Right before Cinch is launched Task manager reads 4.40 -> 4.55 -> 4.44 -> 4.52

 

[MxPhenom 216]

Ryan Shrout doesn't know jack squat about coffee lake at this moment. Why retweet such nonsense? Has he had the opportunity to utilize coffee lake? Where are his benchmarks? My point exactly. This goes back to the pre Ryzen era. Hype, hype, hype. Wait till the product drops with the right MOBO combo, and then post some benchmarks.

There is something called NDA. You must not know much about that.

 

[JalleR]

my 5930K @4.1 is scoring 1214 so my guess is that it will be 5-10% better if Intel is doing business as usual

but maybe they actually made some changes to the cores or added in new stuff VEGA style..... LOL

 

[mark84]

To those dumping dirt on Mr Shrout, you guys need reality checks. Your trash talk sounds good but if you stand back for a minute and realise he does this stuff for a living, he's more than qualified to have a reason for saying what he did. Unlike you, he needs to stick to NDAs so he can do his job and release stuff the day they release for you guys to lap up.
Just because he can't give specifics right now doesn't mean he is wrong.

my 5930K @4.1 is scoring 1214 so my guess is that it will be 5-10% better if Intel is doing business as usual

Well... perhaps 20%
Don't forget the 6850K came in between.
Clock speeds alone will give it a good uptick in performance in single/dual thread stuff.

I'm glad 6 cores is coming to main stream. I've been on HEDT for many years pretty much only because the core counts I wanted weren't there on mainstream platform.
Now that 6 core has finally trickled down for Intel I can consider a cheaper system next upgrade!

 

[FlanK3r]

Its interesting, the CPU working at stock speeds, but in practice with BIOS updates, CPU wil boost at turboboost 2.0, so score will be higher (power consumption also)

 

[LemmingOverlord]

Ryan is right, don't know why @drade is talking rubbish.

Cinebench is reporting some low values for the 8700K, if anything because we know current-gen Intel CPUs are have higher IPC than AMD, and, at the same time, there's a reasonable clock gap between Intel (higher) and AMD (lower).

Results from this video are clearly being impaired by something.

 

[Vayra86]

Getting pretty annoyed with all these Youtubers and 'reviewers' chasing sound bites and spreading misinformation. There is a chance here for the real tech press to make waves as the trustworthy medium. I don't need these emo-kids yelling about hardware at all, most of the time they haven't the slightest clue what they're saying, and they are too impatient to revisit the topic and correct themselves.

That 7700k hype for example, as opposed to what Ryzen had to offer back then, and the mostly utterly unimportant performance gaps (95% of games it mattered for ONLY those with high refresh + 1080p displays which is still a small minority)... it doesn't make any sense. Yes its fastest single thread... if you void your warranty and put overpriced cooling on it. Still doesn't make it a good chip. It'll be very interesting to see how the 8700K fares with its 4.7 Ghz boost out of the box, if it requires the same treatment, I'll pass. There are still 5775c's around that still kick major butt all the way up to 7700k perf levels, I'll get the 6c a few years down the line then. Hot headed CPUs in 2017, thats not stagnation, that's a step back.

 

[ps000000]

I will wait for 2018

 

[Gasaraki]

Getting pretty annoyed with all these Youtubers and 'reviewers' chasing sound bites and spreading misinformation. There is a chance here for the real tech press to make waves as the trustworthy medium. I don't need these emo-kids yelling about hardware at all, most of the time they haven't the slightest clue what they're saying, and they are too impatient to revisit the topic and correct themselves.

That 7700k hype for example, as opposed to what Ryzen had to offer back then, and the mostly utterly unimportant performance gaps (95% of games it mattered for ONLY those with high refresh + 1080p displays which is still a small minority)... it doesn't make any sense. Yes its fastest single thread... if you void your warranty and put overpriced cooling on it. Still doesn't make it a good chip. It'll be very interesting to see how the 8700K fares with its 4.7 Ghz boost out of the box, if it requires the same treatment, I'll pass. There are still 5775c's around that still kick major butt all the way up to 7700k perf levels, I'll get the 6c a few years down the line then. Hot headed CPUs in 2017, thats not stagnation, that's a step back.

Yeah, it seems Ryzen was overhyped. A 7700K can beat a 1800X easily in all games.

 

[B-Real]

Anyway, if we consider the cheaper possibility (389$ for the 8700K instead of the 420$ one), considering that it's around 7700K single bench and let's say in multi thread gets 20% better compared to the 1600, it will still be 85% more expensive... 12% FHD gaming performance and maybe 20% better performance in multi threads for 85% more cost, without the costlier Z370 mobos compared to the B350. Intel is putting a halter round its own neck.

Yeah, it seems Ryzen was overhyped. A 7700K can beat a 1800X easily in all games.

1. Nope, there ARE games that run better on Ryzen.
2. We are speaking (for R5 and R7) about a 10-12% performance (without considering updates which make games like Rise of the Tomb Raider run better on Ryzen) difference in FHD. Do you mean easily by that? lol

 

[dyonoctis]

Too bad that ninja reporter didn't had the insight to leave the system monitor up, so we could see the boost clock on all core, or the behavior of the system.
With the higher clock the score expected was 1300-1400 ?

 

[drade]

Okay okay you all are right I spewed rubbish. Didn't need to see a retweet of him stating he has info on something but has to wait well then keep pulling my arm because it's starting to hurt I want results !!

Too much natty ice last night

 

[GreiverBlade]

oh well .... X370 and R5 1600/1600X then ...

 

[phanbuey]

Anyway, if we consider the cheaper possibility (389$ for the 8700K instead of the 420$ one), considering that it's around 7700K single bench and let's say in multi thread gets 20% better compared to the 1600, it will still be 85% more expensive... 12% FHD gaming performance and maybe 20% better performance in multi threads for 85% more cost, without the costlier Z370 mobos compared to the B350. Intel is putting a halter round its own neck.



1. Nope, there ARE games that run better on Ryzen.
2. We are speaking (for R5 and R7) about a 10-12% performance (without considering updates which make games like Rise of the Tomb Raider run better on Ryzen) difference in FHD. Do you mean easily by that? lol

You're forgetting that this chip will OC to 4.8-5 ghz and for that you're going to be at 1700x performance multithreaded and faster singlethread with much lower memory latency. The 1600 might be better for the money, but it won't be in the same class at all.

I mean a pentium G is/was great for the money with an rx470 ... does that mean I want that for my rig? No.

 

[GhostRyder]

Well the score does seem a bit low so I will be waiting till official reviews are out.

 

[Franzen4Real]

I wonder how my 5820K stacks up against this...

Had the same thought myself when I read "This reflects in its single-threaded performance, where it was significantly faster than older chips, some of which are even HEDT"

I'm wondering if two less memory channels but higher clocks help/don't help for video encoding or 3D modeling&rendering? My 5820K has a pretty conservative overclock of 4.1ghz, all cores.

 

[chief-gunney]

You're forgetting that this chip will OC to 4.8-5 ghz and for that you're going to be at 1700x performance multithreaded and faster singlethread with much lower memory latency. The 1600 might be better for the money, but it won't be in the same class at all.

I mean a pentium G is/was great for the money with an rx470 ... does that mean I want that for my rig? No.

" this chip will OC to 4.8-5 ghz" - lol, on LN2 maybe, probably need serious water cooling to get it to 4.5 on all cores which might just be worth it.
Still, we haven't seen the price yet and in terms of price performance against AMD now, Intel has no hope of being competitive until they come up with something to rival Infinity fabric and it's yield cost.