Intel's Skylake-X architecture offers less gaming performance than Kaby Lake

It looks like Intel's new Mesh Architecture is to blame

Intel's Skylake-X architecture offers less gaming performance than Kaby Lake

Intel's Skylake-X architecture offers less gaming performance than Kaby Lake

 
When it comes to selecting a new CPU for gaming there is no clear cut performance or value leader, with several factors becoming increasingly important as high levels of multi-threading become more prevalent within the market.
 
Traditionally, CPUs with the highest levels of single-threaded CPU performance performed best in games but that does not translate fully into modern multi-threaded games. With the release of Intel's X299 and AMD's Ryzen series CPUs one new factor comes to light that has a profound effect on CPU performance in games, inter-core/thread-to-thread latency. 
 
Techspot has released some new data that compares Intel's i7 7800X and i7 7700K CPUs, comparing Kaby Lake and Skylake-X in a large number of gaming scenarios with a Nvidia GTX 1080 Ti. Below we can see data from just a few of the games tested, where Intel's i9 7800X was found to offer much less performance than Intel's i7 7700K. 
 
This data will be starting for some, as Skylake and Kaby Lake are effectively the same architecture and traditional wisdom would not peg a CPU with more cores and similar clock speeds as having this much less gaming performance. 
 
Below we can see a 6-core i7 7800X at 4.7GHz be outperformed by an Intel 4-core i7 7700K with lower clock speeds, which is something that would be considered by many as a testing error unless you are aware of recent changes to Intel's latest CPU designs. 
 

  

Intel's Skylake-X architecture offers less gaming performance than Kaby Lake  Intel's Skylake-X architecture offers less gaming performance than Kaby Lake
  
Intel's Skylake-X architecture offers less gaming performance than Kaby Lake  Intel's Skylake-X architecture offers less gaming performance than Kaby Lake

(Data from Techspot)


One major factor that changed when moving from Kaby Lake to Skylake-X is the change from Intel's traditional "ring" bus architecture to Intel's new "mesh" architecture, which changes the way that Skylake-X's core communicates with each other. 

Intel themselves admits that Broadwell-E (last-generation, X99) CPUs can outperform Skylake-X (current-generation, X299) due to this change in thread-to-thread communication architecture. This increases core communication latency but offers improvements in other areas, especially for the super-high core counts seen in Skylake-SP.

Below is a comment that Intel has made to Toms Hardware on this matter,

 


we have noticed that there are a handful of applications where the Broadwell-E part is comparable or faster than the Skylake-X part. These inversions are a result of the “mesh” architecture on Skylake-X vs. the “ring” architecture of Broadwell-E.

Every new architecture implementation requires architects to make engineering tradeoffs with the goal of improving the overall performance of the platform. The “mesh” architecture on Skylake-X is no different. 

While these tradeoffs impact a handful of applications; overall, the new Skylake-X processors offer excellent IPC execution and significant performance gains across a variety of applications.


This design change has a huge impact on game performance in select titles, the same titles that suffer from issues when using AMD's latest Ryzen CPUs, which themselves can suffer from increased thread-to-thread latency when communication between CCX clusters. 

In the future, video game developers will need to create games with these architectural quirks in mind, limiting how much of a performance impact that they can have on future titles. This latency issue now affects both Intel and AMD hardware, so developers should certainly take note. 

 

You can join the discussion on Skylake-X's game performance on the OC3D Forums

 

«Prev 1 2 Next»

Most Recent Comments

12-07-2017, 23:48:19

TheF34RChannel
Yep. Hopefully the mesh won't make it to the mainstream platform favoured by gamers (until it becomes actually beneficial compared to the ring). Same goes for the cache restructuring, which is the main reason of you ask me.Quote

13-07-2017, 03:00:50

AlienALX
This mesh wouldn't happen to be glued together would it?

Quote

13-07-2017, 03:21:06

NeverBackDown
Quote:
Originally Posted by AlienALX View Post
This mesh wouldn't happen to be glued together would it?

No that's too expensive for Intel. It's obviously tapeQuote

13-07-2017, 04:20:30

Wraith
Quote:
Originally Posted by AlienALX View Post
This mesh wouldn't happen to be glued together would it?

I see what you did there. Quote

13-07-2017, 04:22:16

AlienALX
Quote:
Originally Posted by NeverBackDown View Post
No that's too expensive for Intel. It's obviously tape
hahaha ! I bet it's thick tape Quote
Reply
x

Register for the OC3D Newsletter

Subscribing to the OC3D newsletter will keep you up-to-date on the latest technology reviews, competitions and goings-on at Overclock3D. We won't share your email address with ANYONE, and we will only email you with updates on site news, reviews, and competitions and you can unsubscribe easily at any time.

Simply enter your name and email address into the box below and be sure to click on the links in the confirmation emails that will arrive in your e-mail shortly after to complete the registration.

If you run into any problems, just drop us a message on the forums.