6.4.23.562625 - Second Life Project Performance Floater

[NiranV Dean]

1 hour ago, Kathrine Jansma said:

Nice as well, but, well I'm trained in physics, so not scared of graphs and statistics.

Thats quite some significant FPS drop in your intro text:

Does that scale linear? Like 2x Maitreya users => 81 FPS, 3x => 69 etc.?

Otherwise it looks pretty conclusive from the numbers and graphs.

 

Framerate and impact cannot scale linearly due to how framerate is calculated.

Say a full frame takes 10ms to render this corresponds to 100 FPS (1000ms / 10ms). If any part of the rendering has a certain impact, say the avatar has a total impact of 5ms impact it will have drastically different FPS impacts depending on how little the baseline render time is. 5ms on top of 10ms would mean dropping from 100 FPS to 66.6 FPS (1000 / 15), whereas adding 5ms on top of 100ms would drop you from 10 FPS barely under 10 FPS, it wouldn't make a noticeable difference. This means small millisecond impacts don't make much of a difference when we're talking about low framerates, whereas if you are getting into the range of 30-60FPS they start to make a difference.

[Kathrine Jansma]

3 hours ago, NiranV Dean said:

Framerate and impact cannot scale linearly due to how framerate is calculated.

Makes obvious sense, when thinking about it. The number of draw calls might scale linear, but the visible effect doesn't.

[Beq Janus]

4 hours ago, Kathrine Jansma said:

Does that scale linear? Like 2x Maitreya users => 81 FPS, 3x => 69 etc.?

pretty much yes. That was, to some extent, the surprise to me. Based on what we are all used to hearing I expected to see something less than linear because I was expecting to see a larger impact from the triangles and thus as the tris per drawcall reduced you'd see a "trade off". The only time I see any indication of that is on the laptops without the dedicated GPUs and even then it only held true for the first few. 

[Beq Janus]

15 minutes ago, Kathrine Jansma said:

Makes obvious sense, when thinking about it. The number of draw calls might scale linear, but the visible effect doesn't.

Yes, indeed. I was referring to the linear change in frametime. where FPS is logarithmic with respect to frametime.