Not a member yet? Why not Sign up today
Create an account  

  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
 
Some Benchmarks: The new threading is about as good as a little overclock?

#1
I had to turn off my overclock as a result of the summer weather, but I took the opportunity to run some FtD benchmarks.
I found the results surprising, just 200MHz above boost is almost as good as the improved threading introduced in devtest.


All trials with GTX 950 and 1866MHz DDR3 on Manjaro Linux in a fresh FtD benchmarker
Code:
i5-4690k stock clocks (3.5-3.9GHZ)
Pre-threading:      FPS  Game Speed
   Boost Graphics: 19.8 0.81
   Max Graphics:   15.8 0.76 (degraded mode)
Post Threading:
   Boost Graphics: 29.8 0.93
   Max Graphics:   21.8 0.85

Code:
i5-4690k OC to 4.1 GHz
Pre-threading:      FPS  Game Speed
   Boost Graphics: 27.6 0.92
   Max Graphics:   14.1 0.69
Post Threading:
   Boost Graphics: 31.3 0.97
   Max Graphics:   21.9 0.89
Gehm's corollary: Any technology distinguishable from magic is insufficiently advanced
Reply

#2
i have an increas of nearly 150% (with very big ships 65k block +)

but im also using a rig(pc) with a little more juice
( I7 5930k@4.5ghz 32Gb DDR4 and 2x GTX 1080Ti)

i think u are mostly limitet by ur gpu

ftd is using on of my OC't gtx 1080ti with nearly 60% on max settings
like's strange music Tongue
Reply

#3
I think your benchmark is suffering from bad noise--both a 40% improvement from a 5% overclock in "Boost Graphics" and a noticeable regression in "Max Graphics" seem unlikely.
Allr andask.
Reply

#4
Its a case of post-threading that the GPU becomes the bottleneck. I actually noticed zero performance increase with the benchmark scene, but that was because it was already GPU bottleneck'd (even with the watercooled overclocked 1080).
That said the performance increase was certainly noticeable in large complex vehicle battles rather than lots of smaller vehicles, that and the garbage cleanup streamline has dramatically decrease the stutter-pause.
*edit*
Oh 50-53fps across all tests (did 5 pre & post) at 0.87- 0.9 gamespeed with absolute max graphics (thats max LOD distance, water tessellation etc; the lot, not just the preset) in the benchmark scene st 20 seconds gametime.
I didn't have any of my old Thyr vs Thyr benchmarks scores on hand but ships of that size in battle; I'd say has improved by a good 40-50% post threading
Reply

#5
(2019-06-25, 08:39 PM)khkpck Wrote: i think u are mostly limitet by ur gpu
My GPU is definitely a limit, it runs at 99% on even moderate graphics. But the improvements from CPU clock still exist at boost graphics.

(2019-06-26, 01:07 AM)Blothorn Wrote: I think your benchmark is suffering from bad noise--both a 40% improvement from a 5% overclock in "Boost Graphics" and a noticeable regression in "Max Graphics" seem unlikely.
There's definitely some noise, I only did a single trial at each level.
However, there is no regression in max graphics, they are uncomparable because max graphics on base clocks dropped me into degraded mode.
I guess I could go back and disable degraded mode for the sake of completing the comparison.
Gehm's corollary: Any technology distinguishable from magic is insufficiently advanced
Reply

#6
(2019-06-27, 07:18 PM)Chrozayis Wrote:
(2019-06-26, 01:07 AM)Blothorn Wrote: I think your benchmark is suffering from bad noise--both a 40% improvement from a 5% overclock in "Boost Graphics" and a noticeable regression in "Max Graphics" seem unlikely.
There's definitely some noise, I only did a single trial at each level.
However, there is no regression in max graphics, they are uncomparable because max graphics on base clocks dropped me into degraded mode.
I guess I could go back and disable degraded mode for the sake of completing the comparison.
Right, I missed that.
Allr andask.
Reply



Forum Jump:


Users browsing this thread:
1 Guest(s)