Intel's 8th Generation Core Family - Coffee Lake (LGA 1151, 6C/12T)

Where do you expect Core i7-8700K's Turbo to land?

  • 3.8/3.9 GHz

    Votes: 0 0.0%

  • 4.0/4.1 GHz

    Votes: 3 23.1%

  • 4.2/4.3 GHz

    Votes: 6 46.2%

  • 4.4/4.5 GHz

    Votes: 3 23.1%

  • 4.6/4.7 GHz

    Votes: 1 7.7%
  • Total voters
    13
  • Poll closed .
Dayman said:
That was private! I trusted you :(
Click to expand...
My bad!!

Post a good link, like that Intel has figured out how to do solder like AMD on their cpu's! (Weird, my cpu has solder and the ones before? and run fine)
 
Shintai said:
Infinity fabric is a massive bottleneck. But if you chose to use applications that scales extremely well and doesn't have any main threads. Then you dont see it as an issue (CB uses tile based render). That's why companies with weak interconnect likes to demo these. There is a reason why you can run faster by disabling a cluster or increase the infinity fabric. Just look at EPYC, there are more latency on a single EPYC chip than a 4 node Xeon. And one of the reasons why EPYC have sold 100-200 chips and Xeons have sold 1M+

There is a reason why Ryzen already got huge price cuts and more to come. And then we dont have to talk about the RMA hell and bugs that plaque Ryzen with machine halts and data corruption because someone skipped close to a year of QA.
Click to expand...

so basically if amd fix up their infinity fabric issue they are good to go right? ryzen 2/3 show us what you can do thanks.


juanrga said:
https://overclocking.guide/the-truth-about-cpu-soldering/

From the link:
Click to expand...

that thread looks smart but kinda contradicted what he wanted to tell us CONSUMERS.

quote: Micro cracks occur after about 200 to 300 thermal cycles. A thermal cycle is performed by going from -55 °C to 125 °C while each temperature is hold for 15 minutes.

honestly, 200-300 thermal cycles and if thermal cycle is defined what he mentioned, i will hit exactly 0 times after 5-10 yrs of usage and most extreme temp i'll be in is around 90c at possibly less than few mins at most. ultimately intel decides to cheap out, can easily give to consumers knowing full well this will never happen and cpu is long lasting, even more so for servers as well eqiuped cooling there simply no excuse.

intel is at it like they do best, milk hard.
 
Why is the price everywhere for 8700k $414? I thought MSRP was projected to be $359.99? Seems bogus. This doesn't happen in other product markets. I don't want to pay a $55 markup when Intel said it was gonna cost less. When will it settle down to the originally projected price?
 
MIDIBoss said:
Why is the price everywhere for 8700k $414? I thought MSRP was projected to be $359.99? Seems bogus. This doesn't happen in other product markets. I don't want to pay a $55 markup when Intel said it was gonna cost less. When will it settle down to the originally projected price?
Click to expand...

It'll taper off when supply outstrips demand.
 
Dayman said:
How much does screen size matter in comparing Ryzen Mobile and Kaby Lake-R battery life? - The Tech Report

TL : DR

KBL R is 2.4x~ better on external screens with a smaller battery.

r52500Uextbattery.png

r5batterynorm.png




Intel 10nm is better than everyone thinks - Fudzilla

TL : DR

According to Fudzilla's "well placed sources" Intel's 10nm is in good health and ready for mass production - HOWEVER they are waiting for even higher yields to maximize profits on the node. Basically arguing economics
Click to expand...

And those numbers comes from a site whos owner is working from AMD. What a complete disaster, no wonder 8th gen chips are everywhere and RR nowhere.
 
MIDIBoss said:
Why is the price everywhere for 8700k $414? I thought MSRP was projected to be $359.99? Seems bogus. This doesn't happen in other product markets. I don't want to pay a $55 markup when Intel said it was gonna cost less. When will it settle down to the originally projected price?
Click to expand...

Its not. Just check more stores ;)
 
Shintai said:
Its not. Just check more stores ;)
Click to expand...

Yep, plenty of shops selling them at exactly MSRP in Europe, would be surprised that it's any different in the US. Had zero issue getting one. Even when they say out of stock you'll usually get it in a couple of days (I got mine ONE day after I placed my order despite them saying "up to 10-15 days and I'm not the only one).

Amazon seems to be doing some milking though (at least amazon.fr) judging by the delivery estimates they are probably selling them from the US warehouses or something.
 
I just fired my 8700K system up and played some PUBG. This system is stupid fast compared to my Ryzen setup @ 4.0. Night and day difference running a 1080 Ti. Ridiculous.
 
kirbyrj said:
That's all fine and good and I don't have a problem with not using solder in theory...

BUT...

Why is it then that a $30 3d printed delid tool with a little thermal paste can lower temps 10-20C over Intel's stock configuration?

espanol.jpeg
Click to expand...

Hint: Delidding.


Nightfire said:
The 1920x gets 1.094 using this graph and your formula, so it is clear PCPEr used garbage Ram/bios for this. That is an 18% difference now.


I've seen Ryzen get over 1800 on CB which puts it in the same IPC as SKL-X such as the 7800x running 4.8 ghz (easy comparison - both 8 core)
Click to expand...

No. The 1920x gets 1.035, which is virtually identical to the 1.025 got by RyZen. The difference is less than 1%.

i7-8700K

IPC = (1/3032.11) / (6C * 4.3GHz) = 1.278 * 10^(-5)

R7-1800X

IPC = (1/3297.1) / (8C * 3.7GHz) = 1.025 * 10^(-5)

TR-1920X

IPC = (1/2176.58) / (12C * 3.7GHz) = 1.035 * 10^(-5)

So AMD has about 25% lower IPC on Blender.

I have said you twice that Cinebench is an exception, not the rule. CineBench is a favorable case for Zen.

PCPer found that 1800X is 13% faster than 8700k in CB15. This is reduced to 8% on POV-Ray 3.7.1. Reduced to 5% on Handbrake. Tie in X264. And then loses in Blender and Audacity. 1800X is 10% slower in Blender. So CB is a favorable case for RyZen.

And rendering/encoding, which is a kind of task well-suited for the throughput-optimized Zen microarchitecture. On latency workloads the gap is higher and Zen is worse.
 
Last edited:
I just assumed CineBench was the most popular synthetic, but if it really favors AMD the most, that might explain why I see it getting thrown around everywhere.
 
I miss my threadripper but NOT for gaming in the least.

My 8600k is hitting 202 single thread in cinebench since were referencing CBench. 7820x is hitting like 198.

The difference is nil. It is just MUCH less expensive if all you need is gaming to go with 8x00k chips. But I need horsepower for other tasks not just gaming hence x299 is for me.

I promise if you sit aroind and gasp at a $50 dollar premium but you can snag a chip your missing out. Coffee Lake IS THE GAMERS chip hands down.

Forgot to add...
Edit... my 8600k is faster at single threaded work than the 8700k was and at a lower clock speed. I cant answer why but it could have something to so with SMT adding latency as the chip decides what thread is assigned to what core during the workload. I.e. single thread operation doesnt mean its start to finish on the same one core the whole time the job will bounce around the chip all over the place.
 
Last edited:
TaintedSquirrel said:
I just assumed CineBench was the most popular synthetic, but if it really favors AMD the most, that might explain why I see it getting thrown around everywhere.
Click to expand...

Its been used for years to show IPC improvements at the same clock speeds since it uses the same extensions. The only thing that might effect it some now is the growing size of the cpu cache. AMD does better in multithreading then Intel and that is why it tends to score better. Some programs leverage different instructions and close that gap and then you factor in the clock speed advantage Intel has as well. It's just a metric like any other program it's just used as a good bench since it has not been continually optimized. Most programs are optimized for Intel and not AMD so this a program that gives a more clear look at the IPC of each company.
 
tangoseal said:
I miss my threadripper but NOT for gaming in the least.

My 8600k is hitting 202 single thread in cinebench since were referencing CBench. 7820x is hitting like 198.

The difference is nil. It is just MUCH less expensive if all you need is gaming to go with 8x00k chips. But I need horsepower for other tasks not just gaming hence x299 is for me.

I promise if you sit aroind and gasp at a $50 dollar premium but you can snag a chip your missing out. Coffee Lake IS THE GAMERS chip hands down.

Forgot to add...
Edit... my 8600k is faster at single threaded work than the 8700k was and at a lower clock speed. I cant answer why but it could have something to so with SMT adding latency as the chip decides what thread is assigned to what core during the workload. I.e. single thread operation doesnt mean its start to finish on the same one core the whole time the job will bounce around the chip all over the place.
Click to expand...

At what clock speed? And what ram timings?

Well there is a thread around here talking about HT hurting scores. I'll see if I can find it. Think it's on win7 and it may have been a core parking issue?

I just tested 7700k at 4.8ghz and get 207 single thread with HT on and 208 with HT off(multiple tests of each)
 
Last edited:
Gideon said:
Its been used for years to show IPC improvements at the same clock speeds since it uses the same extensions. The only thing that might effect it some now is the growing size of the cpu cache. AMD does better in multithreading then Intel and that is why it tends to score better. Some programs leverage different instructions and close that gap and then you factor in the clock speed advantage Intel has as well. It's just a metric like any other program it's just used as a good bench since it has not been continually optimized. Most programs are optimized for Intel and not AMD so this a program that gives a more clear look at the IPC of each company.
Click to expand...

It doesn't show IPC improvements. Its so limited and running SSE loops that the only thing left to test is a tiny subset and uncore parts.
 
4.8 ghz on the 8600K I5. 4.8 on my 7920x 3000mhz ram @ cas 16 on the 7820x and 3200mhz @ 16 on the CoffeeLake.

Cache on 7820x is 3ghz

Windows 10 64. I am not benching with a clean install and nothing running. I have all kinds of shit running so my scores are going to naturally lower than a fresh install bench right away. That is how it should be... realistic usage not a clean install with nil running as background processes.
 
Shintai said:
Yet they dont perform the same. So you can see how pointless it is.
Click to expand...

No not at all ... the 7820x might have a bacteria sized difference in single threaded performance but the 7820x absolutely CRUSHES Coffee Lake in everything else except raw high end clock speed and even that makes no difference in the much much higher overall performance and throughput of the 7820x and higher chips.
 
tangoseal said:
4.8 ghz on the 8600K I5. 4.8 on my 7920x 3000mhz ram @ cas 16 on the 7820x and 3200mhz @ 16 on the CoffeeLake.

Cache on 7820x is 3ghz

Windows 10 64. I am not benching with a clean install and nothing running. I have all kinds of shit running so my scores are going to naturally lower than a fresh install bench right away. That is how it should be... realistic usage not a clean install with nil running as background processes.
Click to expand...
Ok only difference for me is 3000mhz CAS 15, was just curious
 
Speedeu4ia said:
Ok only difference for me is 3000mhz CAS 15, was just curious
Click to expand...

I bet that extra speed makes the difference. Or maybe you just are running less background processes than I. Remember that the 7820x uses Mesh and your 7700k uses ring bus. The ring bus is faster for your four cores whereas the mesh is faster at sharing between a large amount of cores but not as fast between just 4 cores.
 
tangoseal said:
No not at all ... the 7820x might have a bacteria sized difference in single threaded performance but the 7820x absolutely CRUSHES Coffee Lake in everything else except raw high end clock speed and even that makes no difference in the much much higher overall performance and throughput of the 7820x and higher chips.
Click to expand...

8700K is the gaming king, no SKL-X can reach it. And there are other parts as well.

But its not really a surprise since the 2 designs got a different target audience. There is a reason why a CFL can clock its uncore over 50% higher than a SKL-X.
 
juanrga said:
Hint: Delidding.




No. The 1920x gets 1.035, which is virtually identical to the 1.025 got by RyZen. The difference is less than 1%.

i7-8700K

IPC = (1/3032.11) / (6C * 4.3GHz) = 1.278 * 10^(-5)

R7-1800X

IPC = (1/3297.1) / (8C * 3.7GHz) = 1.025 * 10^(-5)

TR-1920X

IPC = (1/2176.58) / (12C * 3.7GHz) = 1.035 * 10^(-5)

So AMD has about 25% lower IPC on Blender.

I have said you twice that Cinebench is an exception, not the rule. CineBench is a favorable case for Zen.

PCPer found that 1800X is 13% faster than 8700k in CB15. This is reduced to 8% on POV-Ray 3.7.1. Reduced to 5% on Handbrake. Tie in X264. And then loses in Blender and Audacity. 1800X is 10% slower in Blender. So CB is a favorable case for RyZen.

And rendering/encoding, which is a kind of task well-suited for the throughput-optimized Zen microarchitecture. On latency workloads the gap is higher and Zen is worse.
Click to expand...

I thought all core was 3.5 ghz on the 1920x but it looks like it was closer to 3.7 ghz.
 
Given the CB results, the r7 1700 was running closer to 3.2 ghz yet it was way closer than that for Gooseberry results.
Also, what is up with the BMW results. The R7 1800x,1700x, and 1700 were practically the same??

PCPer does a terrible job with giving the system setup specs. RAms speed and timing? Bios?
 
legcramp said:
I just fired my 8700K system up and played some PUBG. This system is stupid fast compared to my Ryzen setup @ 4.0. Night and day difference running a 1080 Ti. Ridiculous.
Click to expand...
That’s why I’m still using my 6700k and laugh at those who recommend ryzen for gaming builds. 6700k/7700k and especially 8700k now are no match for amd in gaming tasks. Simple as it is.

Btw nice review from gamer’s nexus showing that any Sky/Kaby i7 owner can sit tight and wait for next gen cpus. No real gains from coffee in gaming tasks and multitasking if compared to 8600k.

https://www.gamersnexus.net/hwreviews/3157-intel-i5-8600k-review-overclocking-vs-8700k-8400

But 8/16 i7 might be my next upgrade.
 
Gideon said:
Its been used for years to show IPC improvements at the same clock speeds since it uses the same extensions. The only thing that might effect it some now is the growing size of the cpu cache. AMD does better in multithreading then Intel and that is why it tends to score better. Some programs leverage different instructions and close that gap and then you factor in the clock speed advantage Intel has as well. It's just a metric like any other program it's just used as a good bench since it has not been continually optimized. Most programs are optimized for Intel and not AMD so this a program that gives a more clear look at the IPC of each company.
Click to expand...

The reason why RyZen gets good scores on CineBench is because CineBench doesn't represent average real-life performance, not because "AMD does better in multithreading". Even with only six-cores CoffeLake beats eight-core Zen in most multithreaded scenarios. From Arstechnica review of CoffeeLake:

Performance

As you'd expect, running six cores at 4.7GHz results in some stellar benchmark results (more so if you can brave 5.0GHz). Even though it has two fewer cores than the Ryzen 1800X (a CPU that costs a hefty £437), the 8700K comes in faster in many production workloads. It's four seconds quicker in Blender at stock, and 11 seconds quicker when overclocked. It's faster at Handbrake video encoding too, and miles ahead in 7-Zip's synthetic benchmark, which tends to favour clock speed even in multithreaded mode.

It's only in PovRay and Cinebench that 1800X comes out on top—and only then by a small amount.
Click to expand...
 
juanrga said:
The reason why RyZen gets good scores on CineBench is because CineBench doesn't represent average real-life performance, not because "AMD does better in multithreading". Even with only six-cores CoffeLake beats 8C Zen in most multithreaded scenarios. From Arstechnica review of CoffeeLake:
Click to expand...
But it does it at a higher price?
Price/performance?
 
Hagrid said:
But it does it at a higher price?
Price/performance?
Click to expand...

8700k --> 409€
1800X --> 417€

https://www.pccomponentes.com/intel-core-i7-8700k-37ghz-box
https://www.pccomponentes.com/amd-ryzen-7-1800x-40ghz-sin-disipador


TaintedSquirrel said:
Running 6 cores at 4.7 GHz is overclocked though.
Click to expand...

That is like pretending that XFR in Zen is overclocking.

Overclocking is when reviewers and users push the clocks on the Infinity Fabric interconnect via overclocking RAM to 3200MHz and higher on Zen systems.
 
  • Like
Reactions: N4CR
like this
Hagrid said:
1800x is 319.00
8700k is 414.00
Newegg.com
Click to expand...

Even if Newegg was to sell to my country I would avoid that store as the plague. CoffeLake is cheaper to people like me. No sure why you pretend to convince me of the contrary.
 
You must log in or register to reply here.
Back
Top