• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

skylake processor, which one???

Im honestly lost now to which for a gamer is a better cpu. I'm not in the market for one but when I bought my 4790k it was the obvious choice at the time.
 
Im honestly lost now to which for a gamer is a better cpu. I'm not in the market for one but when I bought my 4790k it was the obvious choice at the time.

Any recent i5 or i7 is more than sufficient for gaming unless you are running high end multi GPU or benchmarking 720p. Though I'd contend that many recent games are smoother with more cores even when the raw FPS numbers are the same on an i5 (though its more noticeable at 120+fps/Hz gaming and not really so relevant if you vsync at 60Hz).

Which is why I think the 6600K pricing is ridiculous when for a little more you can get a proper 4c/8t or 6c/12t CPU which will give better overall performance unless you spend a lot of time doing stuff that exclusively depends on strong single thread performance.
 
So what your saying there's really little difference in performance. So ideally from a cost point of view, go with the cheapest one?

Don't forget to consider which platform, for your usage, may have more longevity unless your in the habit of upgrading motherboard and cpu every gen....
 
So what your saying there's really little difference in performance. So ideally from a cost point of view, go with the cheapest one?

Its not so much little difference in performance its where the best balance of performance lies (which might differ from person to person depending on their usage).

You also have to figure in RAM and motherboard costs as well - a system with a more expensive CPU might end up overall cheaper, the upgrade route might also be worth considering depending on the type of processing you expect to be doing.
 
Except reviews don't show the 6700K winning every time do they...

http://www.kitguru.net/components/cpu/luke-hill/intel-core-i7-6700k-i5-6600k-skylake-cpu-review/8/

GTAV
=======

[email protected] 1920*1080 avg FPS 74.8
[email protected] 1920*1080 avg FPS 75.4

Shadow or Mordor
==============

[email protected] 1920*1080 avg FPS 121.2
[email protected] 1920*1080 avg FPS 122.1


Tomb Raider
===========

[email protected] 1920*1080 avg FPS 90.3
[email protected] 1920*1080 avg FPS 87.9

So what they actually show is a 5820k and 6700k trading blows with very little between them. If you up the resolution to 1440p or 4k these games would become even more GPU bound so that's not going to be helped by a four core 6700k either.

Why don't you try backing up your assertions without posting nonsense cnet articles that compared pre built PC's with totally different GPU and O/S configurations.

Also what are these mysterious advantages that you believe Skylake's architecture gives you over X99? Ill trade your improvement in IPC and higher stock and OC core speeds for more cores/threads, more PCI-E lanes and more L3 cache (2mb per core on 6700k vs 2.5mb on 5820k).

As expected the article I have linked to shows that in gaming there's rarely more than 1fps between a 6700k @ 4.7ghz and a 5820k @ 4.5ghz and the winner is not always the same chip. Where the 6700k excels most it trades blows with a 5820k where the 5820k excels it totally trashes the 6700K

ie

Cinebench (multicore)
==============

[email protected] - 1034
[email protected] - 1307

Handbrake Conversion
================

[email protected] - 44.9
[email protected] - 54.6

7 ZIP benchmark
=============

[email protected] - 30091
[email protected] - 40410

these are not trivial differences, in heavily multi threaded apps the 5820k always beats the 6700k and by some margin.

Given that a 6700k and a 5820k system currently cost about the same its a bit of a no brainer which to choose especially when you consider that X99 will almost certainly have more longevity than Z170

why will X99 have more longevity you ask?

1) more PCI-e lanes allows for more cards to be added giving access to upcoming I/O, multi GPU setups and multiple fast PCI-E SSD'd

2) Z170 will be getting one more CPU lineup - 'kabylake' still 14nm with most of the effort directed towards improving the iGPU which is of little interest to anyone

X99 will get a die shrink down to 14nm 'Broadwell-E' this will almost certainly (as a percentage) yield more CPU based improvement over Haswell-E then Skylake to Kabylake will as Kabylake is neither a die shrink nor a proper redesign of the CPU (ala the old tick/tock) its a minor refresh

Buy a 5820k now and you may have a meaningful upgrade a few years down the line to an eight core Broadwell-E processor or a six core version both on a smaller 14nm process. Buy a 6700K now and you wont likely have much of a meaningful upgrade to the '7700k' Kabylake CPU unless a better iGPU is your thing


3) Your stuck with 4 cores/ 8 threads with Z170, not so with X99. If games do become more multi threaded (likely given DX12 and the inability to scale processor speeds up much of late) then your Z170 based computer will lag a long way behind a hex core or Octo core X99 setup

4) X99 has more cache per CPU core and more potential memory bandwidth than Z170 has

Lookup a 6700k review where they are running 3000Mhz memory. It is faster than a 5820k in 99% of games. Memory speed makes a big difference for Skylake, as the chip officially supports upto 4133Mhz DDR4 (http://www.intel.com/content/dam/www/public/us/en/documents/datasheets/xmp-memory-for-intel-core-processors-datasheet.pdf), it's memory controller is amazing. X99/Z97 were very hit and miss with some chips not wanting to do above 2400Mhz DDR3.

Also the 5820k has only 28 PCIE-V3 lanes - it cannot do SLI 16x/16x, so it's in exactly the same boat as Skylake running two way SLI/Crossfire at 8X/8X.

6700K has another 20 PCI-E V3 lanes from the chipset for use with PCI-E SSD's - so one could argue it's better equipped than a 5820k for 2 way SLI + 2x M.2 SSD's with 4 lanes each.

You have to go up to the 5930k if you want the full 40 PCI-E lanes from the CPU - a much more expensive CPU than the 5820k.

Obviously the 5820k does better in video encoding, streaming, any productivity program that supports more than 4 cores, though the discussion was about gaming, something Skylake excels at.
 
Last edited:
Can you link these reviews using 3000 MHz DDR4? I've only seen one review where they compared memory speeds and it made no difference in games (which is what I'd expect except for testing pointlessly low resolutions).
 
Also the 5820k has only 28 PCIE-V3 lanes - it cannot do SLI 16x/16x,
so it's in exactly the same boat as Skylake running two way SLI/Crossfire at 8X/8X.

6700K has another 20 PCI-E V3 lanes from the chipset for use with PCI-E SSD's -
so one could argue it's better equipped than a 5820k for 2 way SLI + 2x M.2 SSD's with 4 lanes each.

You have to go up to the 5930k if you want the full 40 PCI-E lanes from the CPU -
a much more expensive CPU than the 5820k.

Doesn't the 6700K only have upto 20 lanes available if you use onboard gpu - unless I've understood it wrong?

Pf5UoXul.gif.png
 
Last edited:
Used Z87 with memory at 2600mhz, pointless exercise for gaming though as it made no difference over XMP of 2400mhz. Heck even at bios default of 1333mhz you'd be very hard pressed to find any difference in games. Something I'm not finding either with 3000mhz ram on X99.

Benchmarks, sure. But I don't play those much. Bit too repetitive.:D
 
Lookup a 6700k review where they are running 3000Mhz memory. It is faster than a 5820k in 99% of games. Memory speed makes a big difference for Skylake, as the chip officially supports upto 4133Mhz DDR4 (http://www.intel.com/content/dam/www/public/us/en/documents/datasheets/xmp-memory-for-intel-core-processors-datasheet.pdf), it's memory controller is amazing. X99/Z97 were very hit and miss with some chips not wanting to do above 2400Mhz DDR3.

Also the 5820k has only 28 PCIE-V3 lanes - it cannot do SLI 16x/16x, so it's in exactly the same boat as Skylake running two way SLI/Crossfire at 8X/8X.

6700K has another 20 PCI-E V3 lanes from the chipset for use with PCI-E SSD's - so one could argue it's better equipped than a 5820k for 2 way SLI + 2x M.2 SSD's with 4 lanes each.

You have to go up to the 5930k if you want the full 40 PCI-E lanes from the CPU - a much more expensive CPU than the 5820k.

Obviously the 5820k does better in video encoding, streaming, any productivity program that supports more than 4 cores, though the discussion was about gaming, something Skylake excels at.


U can also do 3 way sli with x99 8/8/8 with 4 lanes left over for pcie ssd
 
Doesn't the 6700K only have upto 20 lanes available if you use onboard gpu - unless I've understood it wrong?

No the CPU only provides its standard 16 lanes but the Z170 chipset provides 20 lanes regardless. Those lanes are for things like the SATA interfaces, ethernet, audio and the M.2 slots.

X99 chipset only provides 8x PCI-E 2.0 lanes, but since the CPUs have more than 16x 3.0 lanes and the M.2 slots use the CPU's lanes it's not too much of a problem.

But Z170's extra lanes does help with the 'limitation' of 16x lanes on the CPU itself, particularly when you're using an M.2 slot or two.

Different approaches for what's more or less the same outcome to be honest. Unless you're using 3 or 4 way SLI and then X99 is the only sensible choice.
 
Last edited:
Lookup a 6700k review where they are running 3000Mhz memory. It is faster than a 5820k in 99% of games. Memory speed makes a big difference for Skylake, as the chip officially supports upto 4133Mhz DDR4 (http://www.intel.com/content/dam/www/public/us/en/documents/datasheets/xmp-memory-for-intel-core-processors-datasheet.pdf), it's memory controller is amazing. X99/Z97 were very hit and miss with some chips not wanting to do above 2400Mhz DDR3.

Also the 5820k has only 28 PCIE-V3 lanes - it cannot do SLI 16x/16x, so it's in exactly the same boat as Skylake running two way SLI/Crossfire at 8X/8X.

6700K has another 20 PCI-E V3 lanes from the chipset for use with PCI-E SSD's - so one could argue it's better equipped than a 5820k for 2 way SLI + 2x M.2 SSD's with 4 lanes each.

You have to go up to the 5930k if you want the full 40 PCI-E lanes from the CPU - a much more expensive CPU than the 5820k.

Obviously the 5820k does better in video encoding, streaming, any productivity program that supports more than 4 cores, though the discussion was about gaming, something Skylake excels at.

I call total BS unless you can post proof otherwise. Even the quickest of internet searches shows there's practically no gains past 2400 DDR4 on either X99 or Z170



'After spending dozens of hours benchmarking ten different memory speeds on the Intel Z170 + Skylake platform we must admit that we are too shocked by the findings. Our benchmarks show that the memory bandwidth increased, but there wasn’t a tangible improvement in system performance with real applications. We ran other applications and game titles when we tested this memory kit and you mostly ended up with flat performance charts'

http://www.legitreviews.com/ddr4-me...nding-the-best-ddr4-memory-kit-speed_170340/6

test after test shows that shoving fast ram on z170 (or x99 for that matter) makes very little difference whilst gaming.

Going from 2400 to super fast 3733mhz ram nets you about 1-2fps at games already running at 100fps+

Above 2400mhz it seems that the bottleneck almost entirely rests elsewhere and faster memory makes little to no difference

This article posted on Linus suggested improvements but they were comparing DDR4 1600 to DDR4 3000 on Skylake!!
To add insult to injury you will notice that the 6700K can barely beat a 4790K with DDR3 memory when the 4790 was clocked at the same speed in BF4!

http://linustechtips.com/main/topic...nificantly-with-ramspeed-in-dedicated-gaming/

X99 has more PCI-E lanes available for graphics cards whichever way you try to work it (cant do Tri NVidia SLI or 16X with 8X on Z170 without PLX chips)and has capacity to house CPU's with 6+ cores and forty lanes coming direct from the CPU


Like I said, you're really just preaching to the choir at this point, one side has facts and the other has marketing guff.

Seems to be a running theme with the Z170 crowd.....

Why don't you show us some of these gaming benchmarks where an overclocked 6700k Skylake 'excels' over an overclocked 5820k

I say it again where a 6700K excels it trades blows with a 5820K where the 5820K excels it trashes the 6700k in a system that will typically cost around the same or less
 
Last edited:

Your datasheet makes no mention of the 6700k which officially supports up to DDR4 2133 http://ark.intel.com/products/88195/Intel-Core-i7-6700K-Processor-8M-Cache-up-to-4_20-GHz

All intel chips have stupidly low officially supported jdec speeds for no clear reason. Yes, they can run much faster as those XMP profiles indicate with the 6600k though given that's limited to 8GB total for the 4GHz kit I'm hardly viewing it as particularly useful if talking about 'future-proofing' - from the same datasheet the 6600k using 16GB kits it's getting much the same speeds as X99 can do with 32GB but in dual channel not quad so hardly a win or evidence it's memory controller is amazing.

It's quite possible it does have a better memory controller, but that source did not indicate it nor do reviews seem to back up your assertion that faster memory makes a particularly big difference for gaming on skylake (or other platforms).
 
Last edited:
Your datasheet makes no mention of the 6700k which officially supports up to DDR4 2133 http://ark.intel.com/products/88195/Intel-Core-i7-6700K-Processor-8M-Cache-up-to-4_20-GHz

All intel chips have stupidly low officially supported jdec speeds for no clear reason. Yes, they can run much faster as those XMP profiles indicate with the 6600k though given that's limited to 8GB total for the 4GHz kit I'm hardly viewing it as particularly useful if talking about 'future-proofing' - from the same datasheet the 6600k using 16GB kits it's getting much the same speeds as X99 can do with 32GB but in dual channel not quad so hardly a win or evidence it's memory controller is amazing.

It's quite possible it does have a better memory controller, but that source did not indicate it nor do reviews seem to back up your assertion that faster memory makes a particularly big difference for gaming on skylake (or other platforms).

I suggest wearing glasses when viewing the link - the i5 6600k is the same silicon as the 6700k - both support those ram kits. As you can see, Intel has officially validated 4000Mhz DDR4 kits on Skylake - and just look around at what people are overclocking their memory to. Of course the ARK page for each CPU only list the fastest JEDEC speeds available - which has been the case for many, many years.

This is significant improvement from Haswell, where the 4790k struggled to run XMP 3000Mhz kits, unless you got the luck of the gods and won the silicon lottery. I've seen X99 users who couldn't get their boards/CPU to work properly with 3000Mhz kits also. Skylake just supports it out of the box, thanks to it's memory controller being significantly upgraded.
 
I've seen X99 users who couldn't get their boards/CPU to work properly with 3000Mhz kits also. Skylake just supports it out of the box, thanks to it's memory controller being significantly upgraded.

Shock horror simpler dual memory controller can run memory quicker than more complex quad channel memory controller none of which actually matters much because the quad memory setup has more bandwith and running ddr4 above 2400mhz on either z170 or x99 makes next to no difference to performance.....

It would also seem that the bandwidth of quad channel memory is largely wasted...

http://www.pcworld.com/article/2982...e-shocking-truth-about-their-performance.html

Bottom line neither x99 nor z170 are platforms that are short of either memory speed or bandwidth even when using 'regular' 2400mhz or 2666mhz memory in a dual channel configuration
 
Last edited:
Well, got X99 with 3000mhz ram, works straight out of the pack. Its what id call boringly reliable. Whack it in, hit XMP in bios, 125 strap then start to overclock. Which seems to be an easy thing to do on both X99 and Z170, although the former being a soldered chip tends to run a fair bit cooler when stress testing. Intel do need to get away from pasting the mainstream chips, get rid of the igpu which is as useful as an ashtray on a motorbike. The money saved could go into getting the thermals right. No one in their right mind wants an onboard gpu, buy a cheap 20 quid secondhand card if your main high end card fails. Bung it in a drawer to keep for future use.
 
Well, got X99 with 3000mhz ram, works straight out of the pack. Its what id call boringly reliable. Whack it in, hit XMP in bios, 125 strap then start to overclock. Which seems to be an easy thing to do on both X99 and Z170, although the former being a soldered chip tends to run a fair bit cooler when stress testing. Intel do need to get away from pasting the mainstream chips, get rid of the igpu which is as useful as an ashtray on a motorbike. The money saved could go into getting the thermals right. No one in their right mind wants an onboard gpu, buy a cheap 20 quid secondhand card if your main high end card fails. Bung it in a drawer to keep for future use.

Agree there any of the K cpus should not have igpu and should be soldered heat spreaders as standard.
 
Back
Top Bottom