Windows 10 Idle versus Load Performance on FX and Ryzen

Peanut253 · July 7, 2017, 6:24am

This posting is part of a series of posts meant to explore the following topics:

Testing designed to compare FX and Ryzen scaling with various workloads.
Testing designed to compare GTX 660 and GTX 1050 Ti scaling with various CPUs.
Testing designed to compare Windows 7 and Windows 10 under real-world idle conditions.
Testing designed to compare gaming/encoding performance while encoding (under CPU load) in the background.

Level1Tech Threads:

External Topic Index:

The following benchmarks were performed with the following hardware configurations:

Windows 7 Sp1 Updated, Windows 10 1703 Updated
GeForce GTX 660 and 1050Ti both at stock frequences.
Tests focus on real-world configurations and actual usage variations, not solely hardware component isolation. For that, check out gamersnexus.
1% lows, 0.1% lows and standard deviation calculations (for accurate error bars) not performed due to data analysis and time limitations.
For full disclaimers, detailed configuration information, and results data please see the raw results Google doc. Tabs exist.
Regarding MetroLL and Ryzen's SMT.

Synthetic CPU/Memory Benchmarks

My FX system has lousy memory writes.
Oddly enough, putting a load on the system consistently increases memory write performance. Unclear if this is a testing quirk or indicative of real world memory write performance.

Encoding time does not quite double, implying two encodes started simultaneously would take less time overall than stagnating them. That said, it would be quite a while before either encode finished.

Encoding at 0.52 fps typical under load. At least Ryzen doubles that to 1.15 fps. Or even 1.82 for those rare moments when there is only one encode in the background, instead of multiple ones. Welcome to the world of encoding.

Firestrike seems like it tries very hard to ignore whatever you are doing in the background when issuing a score. Idle apps will not affect your Firestrike score.

So loads affect minimums disproportionately compared to averages.
That said, Unigine does not properly factor in the minimum framerates into how it is reporting averages.

Tomb Raider does not care about your CPU. It doesn't care if it is a dual core, or you are encoding in the background. It just doesn't.
But sometimes, while under load, FPS can dip due to OCing. This would not be perceptible under Windows 10 but it is measurable. Unclear why this happens.

Regarding MetroLL and Ryzen's SMT.
From the averages, it looks like FX dips much harder than a Ryzen system. The Ryzen system would be smoother.
Given how little minimum frame rates are indicative of actual performance in MLL, it is difficult to draw any conclusions as to the extent of CPU loads while running this game beyond the above.

Another Tomb Raider. Feel free to encode in the background while playing Shadow of Mordor under Win 10. It will not affect performance significantly.

Synthetic and CPU bound workloads suffer tremendously if there is an existing CPU load on the system. Remarkably unremarkable.
Games that already have a GPU bottleneck will not really be affected, unless the game is CPU bound, at least for some of the frames some of the time. I am looking at you Metro Last Light.
The 2-core 4850e system was completely unusable under load (not shown) hence not bothering with any load testing. That CPU is at 100% running Firefox.
8-core CPUs, regardless of architecture, are awesome at multi-tasking and running multiple workloads of independent types simultaneously with very little performance penalty. This reasoning does not apply to same workload types (CPU/GPU).