Intel Skylake / Kaby Lake

[myocardia]

Wow, classy move... People like you are why the world is in such a mess. I had an issue with one poster, and you went out of your way to create more hate. Although you are a horrible person, on the plus side im sure your original post had no value and saved me from reading it.

You seem like a very paranoid individual.

lmfao, you think?

 

[CHADBOGA]

lmfao, you think?

LOL

 

[Luddite]

Looking to get SkyLake Xeon 16xx v.5 (4 or 6 cores) for a workstation.

1. Is this called 'Purley'?

2. When is the release time frame for SkyLake Xeons?

3. When would compatible mobos come out for these?

4. Anyone know what the clockspeeds and TDP would be?


Thanks.

 

[ShintaiDK]

Purley platform is 2017.

 

[phillyman36]

ShintaiDK I know we are still waiting for all the mob but are you leaning to the Evga or the Asrock mini-itx mobo?

 

[ShintaiDK]

ShintaiDK I know we are still waiting for all the mob but are you leaning to the Evga or the Asrock mini-itx mobo?

If the EVGA Z170 Stinger got a M.2 slot on the back for SSD. (There is a M.2 riser on the front for Wifi next to the battery). Then its no competition, I will get the EVGA Z170 Stinger.

However if it doesnt, then I have no idea yet what I will do. Either another mobo with the features I wish, or a compromise.

 

[Sweepr]

Core i5 6600K Geekbench 3 32-bit Score:
http://browser.primatelabs.com/geekbench3/3093800

 

[phillyman36]

Core i5 6600K Geekbench 3 32-bit Score:
http://browser.primatelabs.com/geekbench3/3093800

Just for some sort of a reference my i7 4770k
Single core 3960
Multi core 15100

http://browser.primatelabs.com/geekbench3/3102562

With Hyperthreading off
Single Core 3958
Multi Core 12579

http://browser.primatelabs.com/geekbench3/3102620

 

[Sweepr]

Just for some sort of a reference my i7 4770k
Single core 3960
Multi core 15100

http://browser.primatelabs.com/geekbench3/3102562

Need to turn of hyper threading

Thanks, please run with HT off too if you can.

 

[phillyman36]

Thanks, please run with HT off too if you can.

Done and added to original post.

 

[inf64]

@ phillyman36

That score of yours is done on stock 4770K? Hmm your ST score @ stock is on par with my OCed 4690K's score (4.3Ghz).

 

[Sweepr]

So Core i5 Skylake-S was 15.8% faster in ST and 17.7% faster in MT despite Core i7 Haswell's extra cache. Impressive.

 

[phillyman36]

@ phillyman36

That score of yours is done on stock 4770K? Hmm your ST score @ stock is on par with my OCed 4690K's score (4.3Ghz).

Yes no overclock. Not sure how the turbo effects it.

 

[inf64]

Yes no overclock. Not sure how the turbo effects it.

Your turbo should be constant at 3.9Ghz no matter the load. I know my 4690K runs at 3.9Ghz across all geekbench tests (ST or MT). Still I'm around 10% higher clocked yet scoring the same as you in ST tests for some reason.

 

[AtenRa]

Your turbo should be constant at 3.9Ghz no matter the load. I know my 4690K runs at 3.9Ghz across all geekbench tests (ST or MT). Still I'm around 10% higher clocked yet scoring the same as you in ST tests for some reason.

2mb more cache ??

 

[ShintaiDK]

Memory timings perhaps?

But it really shows how unreliable greekbench is to compare with.

 

[inf64]

@ AtenRa

I doubt 2MB more L3 cache could result in 10% better performance

@ Shintai

Maybe, but the memory score/subscores are basically identical between his and my run. Geekbench is a strange benchmark indeed. I looked at my other results (C11.5/R15, fritz chess, winrar, wprime etc) and they are all in line with ~4.3Ghz 4C/4T Haswell. Only GB3 shows this discrepancy. I also ran the test under Win10, the scores are within the margin of error Vs Win7.

 

[ShintaiDK]

Here is mine for compare:
http://browser.primatelabs.com/geekbench3/3103068

Windows 10 upgrade, so filled with crap.
Else all stock i5 4670 + DDR3 1600Mhz 9-9-9-24-37 (2T CMD) and a 480GB M500 SSD.

I also noticed Phillyman36 got a fast Samsung 850 Pro SSD. Maybe thats of influence too.

 

[inf64]

Here is mine for compare:
http://browser.primatelabs.com/geekbench3/3103068

Windows 10 upgrade, so filled with crap.
Else all stock i5 4670 + DDR3 1600Mhz 9-9-9-24-37 (2T CMD) and a 480GB M500 SSD.

I also noticed Phillyman36 got a fast Samsung 850 Pro SSD. Maybe thats of influence too.

Thanks.
Here is the comparison between yours and mine:
http://browser.primatelabs.com/geekbench3/compare/3103126?baseline=3103068
Assuming your chip maintains 3.7-3.8Ghz Turbo during all tests, the scaling is basically linear going from your chip to mine since my results are around 14-16% better than yours which is equival. to 4.3Ghz Haswell 46xx. BTW Win10(upgrade from WIn8.1 also filled with crap) is installed on Kingston HyperX Fury SSD (240GB).

 

[ShintaiDK]

Then there is only the cache left it seems. But that would be a crazy impact. On the other hand it explains Skylake with its improved caches.

 

[inf64]

Yeah probably. It's odd that GB3 is so cache sensitive, 33% more L3 gives you 10% more performance . Skylake probably has faster caches too which could explain the rather large boost we see with 6600K.

 

[ShintaiDK]

With a 6700K I need to buy a cooler. Its been a long time since.

Thinking on this:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835118126
http://www.zalman.com/global/product/Product_Read.php?Idx=468

Since it needs to fit in ITX case too. And it looks very good, good cooling, very silent.

Unless, anyone else got better idea?

 

[inf64]

That Zalman seems like a nice choice. I assume you will run that chip @ stock clocks so it should be fine. My OCed 4690K is below 70C during full load (Realbench stress test) on a very silent (and very cheap) CM Seidon 120V.

 

[ShintaiDK]

Yep, stock and in a Silverstone SG08B chassis. But I am slowing looking for something that could be smaller. The cooler looks to be able to handle 130W or more.

Water isnt an option like yours unfortunately

Else nice temps and quiet.

 

[inf64]

Yeah, for your rig that Zalman should be perfect. The only issue (or better to say fear) I have with my AIO WC is that water pump *could* fail at some point in time. Luckily CPU has a measure that will just shut down the whole system if that happens , hitting 100C is never fun to see lol. Haven't heard many stories about failed pumps but every now and then I stumble at some forum post from unlucky user who got into this situation XD.