Coffeelake thread, benchmarks, reviews, input, everything.

[moonbogg]

Hmmm

Oh boy, not this scenario again. TWICE now we've said "It can't be so" with both Ryzen and X299, but it was so. Is it so again? Gaming performance less than it should be? Why would they mess with the mesh or cache on these chips? I think its a bunk slide.

 

[StinkyPinky]

Oh boy, not this scenario again. TWICE now we've said "It can't be so" with both Ryzen and X299, but it was so. Is it so again? Gaming performance less than it should be? Why would they mess with the mesh or cache on these chips? I think its a bunk slide.

Well we know sure enough in a few hours.

 

[coercitiv]

Oh boy, not this scenario again. TWICE now we've said "It can't be so" with both Ryzen and X299, but it was so. Is it so again? Gaming performance less than it should be? Why would they mess with the mesh or cache on these chips? I think its a bunk slide.

My money is not on the cache, but rather on the same type of rushed launch as X299 that gave mobo makers little time to prepare on the software side. We'll probably see locked turbo ratios or other weird turbo behavior due to partially working power management, which is fine on one side since this can be soon addressed via updates, but not fine in the context of first impression.

Let's hope the hardware Intel sent to reviewers is in a different shape BIOS wise than what smaller review sites managed to buy and review before the NDA expires.

 

[FlanK3r]

guys, logicaly, depends at games, settings of games etc And also at setting BIOS, because we have edhanced turbo settings vs Intel spec settings. Tahts mean in first case, CPU working in all scenarios (without AVX stress) at max turboobost, so at 4.7GHz! With Intel spec CPU working at lower specs. So you will see mixed review, higher numbers will be with edhnaced multicore turbo (simply if Cinebench will be around the 1450+points at stock) and lower scores at Intel specs BIOS settings (13xx points in R15).

 

[Blue Storm]

Well, as I wrote in the post with these results (that table are my results of 8700K), some numbers of game tests are weird and I will try to find the problem today and re-test it with clean OS of it changes. If not, there is bottleneck somewhere... but it isn't secret that Ubisoft games are generally not so well optimised and as you can see results of 7900X, they maybe don't like high-core-count Intel CPUs. Will see...

 

[majord]

https://videocardz.com/73089/intel-8th-gen-core-coffee-lake-s-review-roundup

early reviews begining to roll


http://www.expreview.com/57166.html
http://www.itocp.com/htmls/18/n-8218.html
http://www.pcekspert.com/clanak/intel-i7-8700k-recenzija/
https://wasd.ro/hardware-reviews/mobo-cpu/intel-core-i7-8700k-review/

 

[epsilon84]

https://videocardz.com/73089/intel-8th-gen-core-coffee-lake-s-review-roundup

early reviews begining to roll


http://www.expreview.com/57166.html
http://www.itocp.com/htmls/18/n-8218.html
http://www.pcekspert.com/clanak/intel-i7-8700k-recenzija/
https://wasd.ro/hardware-reviews/mobo-cpu/intel-core-i7-8700k-review/

Do you know when the 'big' review sites like Anandtech and Tomshardware will go live?

 

[jpiniero]

Do you know when the 'big' review sites like Anandtech and Tomshardware will go live?

9 Eastern (2 hours from now). Maybe 10.

 

[wchang99]

https://videocardz.com/73089/intel-8th-gen-core-coffee-lake-s-review-roundup

early reviews begining to roll

That videocardz.com review roundup/meta-list (have they done it for a long time?) is really nice to have.

 

[justoh]

early reviews begining to roll

https://wasd.ro/hardware-reviews/mobo-cpu/intel-core-i7-8700k-review/

But I'm telling you a secret. In terms of gaming, the star is not 8700K, but 8600K overclock. 8600K is a real beast. It's cool and can run stable at 5.3 GHz using an AIO cooler.

Sounds promising. Guess I'll have a few weeks to reconsider getting an 8700k, since I can't get either at the moment, apparently. Most etailers in CH referencing a 15.10 stock date. Currently have a 2500 non k. Plan on using the input here and in the other thread to work out the rest of the components. Thanks in advance.

 

[LTC8K6]

We will probably see that most games still don't need or use more than 4 cores.

 

[Crono]

9 Eastern (2 hours from now). Maybe 10.

PCPer confirmed 9 AM ET.

 

[Glo.]

3 different reviews show power delta between Idle and Load for Core i5 8400@ around 45W. Thats is very good info. Also the CPu draws under load just 6W more than i5 7500.

 

[vissarix]

And to think there were people saying that Intel was finished...got caught with pants down, its all over and bla bla bla...

Yet here we have a 6 core oc to 5ghz scoring 1663 on cinebench MT which is higher then stock 1800X, so you need to overclock an 8 core Ryzen $500 flagship to catch up with a six core Intel which also costs a lot less...

And this is the best case scenario for Ryzen, i mean outside Cinebench you wont see that scaling on any application which will make coffeelake even more lethal...no wonder Amd drop prices like mad, selling 8 cores for $250

Threadcrapping and trolling are not allowed.
Markfw
Anandtech Moderator

 

[PhonakV30]

https://www.anandtech.com/show/1185...-8400-initial-numbers?utm_source=notification

 

[PhonakV30]

in game , 5930K is faster than 8700K? isn't it Haswell again CFL ?

 

[FIVR]

Geez, it gets absolutely demolished by even the base 1600 in AES. Why does it perform so poorly?

4.7Ghz getting beaten badly by 3.2Ghz AMD. I never thought I'd see that. It's like the reverse of bulldozer/sandy bridge

 

[PeterScott]

https://www.anandtech.com/show/1185...-8400-initial-numbers?utm_source=notification

Something is messed up with initial BIOS. Check the gaming scores, The i5-8400 is destroying the i7-8700K in a couple of games, while having less Clock Speed, Less Cache and half the threads? Slow frame spikes in GTA, are the worse of all processors!

 

[PhonakV30]

Geez, it gets absolutely demolished by even the base 1600 in AES. Why does it perform so poorly?

4.7Ghz getting beaten badly by 3.2Ghz AMD. I never thought I'd see that. It's like the reverse of bulldozer/sandy bridge

This is synthetic benchmark, just look at reality.Hard to believe that 5930K is faster than 8700K ( all cores at 4.3Ghz) , at least according to Anandtech
It must be Bios Fault.

 

[vissarix]

Lol at that poor Ryzen cpu, 30 to 50% slower on gaming

Threadcrapping and trolling are not allowed.
Markfw
Anandtech Moderator

 

[Burpo]

Wow, that sucked.. 8700 & Asus combo in cart, went & selected memory, added to cart & 8700 combo had been removed.. Sold out!

 

[FIVR]

This is synthetic benchmark, just look at reality.Hard to believe that 5930K is faster than 8700K ( all cores at 4.3Ghz) , at least according to Anandtech
It must be Bios Fault.

So it's a BIOS fault? I suppose that makes sense. To see 4 core 4ghz AMD matching 6 core 4.7Ghz intel is surprising to say the least.

 

[pantsaregood]

I'm interested to see if someone manages to inject Coffee Lake microcode into Z170/Z270 motherboards or if Skylake/Kaby Lake gets injected into Z370 motherboards.

There are some extra voltage pins on Coffee Lake, but the fact it was tested on Z270 motherboards means that it can theoretically run without them. It was even called a "firmware lockout," so I think seeing modified BIOS is quite likely.

 

[elhefegaming]

I don't doubt there are bios issues or something. this was clearly rushed.

 

[LTC8K6]

Geez, it gets absolutely demolished by even the base 1600 in AES. Why does it perform so poorly?

4.7Ghz getting beaten badly by 3.2Ghz AMD. I never thought I'd see that. It's like the reverse of bulldozer/sandy bridge

Looks correct to me given the 7700K numbers. SL-X and Ryzen are just better at this benchmark.