E-Cycles - one week left to get up 4x faster rendering + upcoming AI techs with 100€ off!

Hi Wolfie,
send me your file per Email if you want and I’ll have a look to it.

it’s massive, about 180mb w/ the packed files etc etc, however i’ve just done a quick test - i deleted everything, added a cube and camera, and the crash still happens, so could it be something in the file itself rather than an object or material? Rendered View still works fine, but hitting F12 gives this:


then it locks up.

i was going to put this cut-down file in pasteall but it’s giving some weird “User armadillica already has more than ‘max_user_connections’ active connections” error :-/ if you want it, PM me your email ~ it’s only 3.5mb

cheers

It still renders fine in 2.79.

You can send an email by responding to one of the gumroad emails. If you send it now, I’ll have a look tonight.
I think with 3.5mb, you can even attach it here or in a pm.

Fixed. Was due to the old option “save buffers”. If anyone needs it, I can bring it back, but then auto tile size has to be deactivated when using it.

The new version is already available, with all the latest improvements.

1 Like

thanks for the fast work! I’ll give it a try when i get home tonight.

Got the RTX version last night, was a bit apprehensive w/ the Read Me saying 9xx cards might be slower, but i found it pretty similar. I’ve done a couple of test renders of a frame of my animation, 500 samples at 1280x720 at 100%, machine is Win7, 16gb, using a GTX980;

Blender 2.79 ~ 3m:09s
Ecycles 9xx ~ 1m:42s
Excycles RTX ~ 1m:47s

needless to say, i’m impressed!

i want to try something a bit more complex, using my fire scenes but apparently 2.8 needs to have all that stuff baked first so i’ll have to look into that, might have to wait until the weekend ~ i’ve done nothing in 2.8 so far but load it up so i could use E-Cycles.

Good to know the RTX version is now working properly on older cards. @Komposthaufen can you confirm with your GTX 980? If it’s the case for you also, I’ll move everyone officially to the _RTX version.

The machine i’m doing my rendering on is a Win10 using a GTX970, i’ll try this as well. The difference for the 980 was only a few seconds, but given i need to render over 1400 frames i’ll probably stick w/ the 9xx version of E-Cycles after i’ve tested, it might seem trifling but all those extra seconds will add up. Might be something for you to consider about code optimisation for bot RTX and 9xx, as I would expect many others will be wanting E-Cycles for animation rendering rather than one-off images.
cheers

Of course, I’m also happy about each second I can get off rendering time. But with what is coming next month (one new feature every month), many people will get 20% faster rendering on top of the existing speedup. So by having to make only one build for all, I can invest the time in much bigger improvements. And in the meantime as you said, you can always use the 9xx build anyway.

:open_mouth: :sunglasses:

i’ve just seen 2.8 can apparently do CPU+GPU combined, does E-Cycles take advantage of this feature? Don’t think i’d try it for my animation just in case i overcooked my CPU lol but i might try a few tests to see what the speed’s like.

yes, it supports CPU+GPU, but your CPU must be really fast compared to your gpu to see any benefit. I’m always interested in comparison between GPU+auto tile size vs GPU+CPU with whatever tile size you like. (specify the cpu and gpu type and clock if changed)

tried a quick test before i read this post
Win 10, I5 Skylake, GTX970
500 samples @ 1280*720
GPU: 2m:01
GPU+CPU: 1m:49s
didn’t mess w/ the tile size, it’s whatever’s defaulted on your builds.

so yeah, the RTX version works fine on this GTX970 too.

Great. The speedup with CPU is in the range I got from other users until now, around 10%. Compared to a doubling in power usage, it’s low, but if you are on a deadline, it helps.

A thing I noticed is that the frequency of the CPU core managing the card has an important impact on the GPU perf. As most cpu overclock (boost) when only 1 thread is used, it may also be a factor. So if you are to buy a new computer, it may make sens to buy a CPU with a low core count but a high frequency. It would also help when using Blender itself, as most operators and modifiers don’t use multi-threading much yet. If the pre-processing phase is the biggest part (building BVH, etc.), a high core count is then more important.

Unfortunately I don’t get any speedups with the new RTX build in my test scene.
Even worse, this build is very unstable. I tried to render this scene 5 times, and only two times he was able to finish the render, the other times blender just closed itself at about 50% progress.
The render time was about 31:02.35 for both successful renders.

ok, it’s not very logical at first that the 970 are ok and the 980 get such an impact from newer Cuda SDK. But then I’ll continue to provide cuda 9 kernels for 9xx cards. :slight_smile:

i’ll chip in here and say i had a couple of closures on my 970 too. i’m rendering a different scene now, one w/ fire/smoke sims in so maybe that’s a place to start looking ~ i’ve rendered about 200 frames on the scenes previous to this, both w/out fire sims, so this may logically be the culprit. however, i’ve asked elsewhere on this site about 2.8 fire sims and my baked data was misplaced;


so maybe this is a contributing factor?

Didn’t have time to look into it, but i’ll try and give it a run on the 980 tonight and run it via CMD so i can see any error messages.
cheers

I did some more tests and got some interesting results.
Since the first measurements on Dec 31, I’ve updated my graphics card drivers and now the render time increased for the Scene. It went up from 25:13.23 to now 28:54.86. the new RTX build is at 31:26.29.
I always do multiple measurements to ensure that the results are correct, but now I cannot recreate the 25:13.23 anymore.

GPU+GPU is faster in theory because some tiles are rendered (slower) by the CPU and so, GPU will have less tiles to render in order to complete the task. But most of the times the last tiles to be rendered are taken by CPU, so they render slower, and it can happen that these last tiles take such time to finish that all the advantage is lost, and goes negative, resulting in an overall slower render

Interesting. If you can give me the driver version that gave the best times, that may help other user of the 9xx series. Luckily on Windows, it’s pretty easy to roll-back to the previous version in the hardware manager.

Sorry, I was wrong it is still possible.
If you open Blender and start rendering it is always slower than if you render it the second time.
The render time with the RTX version is indeed only a tiny bit slower than the old build (~1min).
I’ve also fixed the Crashes for the RTX version with your registry trick.
Sorry for the confusion :confounded: