OTOY Forums

Load time: 6-7sec
Compilation time: 53-54 sec

Windows 8.1
Octane v3 alpha 4
intel i7 5960X overclocked at 4Ghz (8c/16ht)
SSD / 64GB DDR4
1x TitanBlack + 2x 980Ti

CPU Monitoring: https://db.tt/jrVAf2uG

ok, then my gues about HT migh not be working..however it's still hilarious considerign the fact my 250$ CPU bought like 4-5Years ago.. do better compiling the scene than Tutors Xeaons =DDD.. I mean WHAT???!!! hehe.. that single core performance rally matters here =)

think it's time to go to liquid nitrogen Guys..check this out: current world record for a single-threaded Intel Core i7-6700K is 7.02566 GHz, which is achieved with a voltage of 4.032V (For comparison, the i7-6700K is typically around 1.3V at load) Guess.. it would do compilation in what.. half a minute? =DDD

This is fucking crazy.

Please, OTOY, can you put some light on this issue??

This is effecting my new build spec decisions so much

Tutor,

Is this only compilation time (no opening the file, no rendering)? This is so sick (in negative way) that it takes so long on your machine...

Let's hope OTOY will give us thorough explanation of what's going on.

aiven wrote:Load time: 6-7sec
Compilation time: 53-54 sec

Windows 8.1
Octane v3 alpha 4
intel i7 5960X overclocked at 4Ghz (8c/16ht)
SSD / 64GB DDR4
1x TitanBlack + 2x 980Ti

CPU Monitoring: https://db.tt/jrVAf2uG

And again we see that the cpu is not stressed fully to 100% in the 25-40 seconds period, unlike mine.

So my ques is that the compiler is not MT (MultiTthreading) optimized for gpus over 4 physical cores. If the code is corrected and will make use of all possible cores we shall see significant improvements for Xeons.

=) I would not hold a breath for this being solved anytime soon, but ..I would lieke to be wrong..

Yam/Sebastian..if I could say something.. - go for sweetspot (high clockspeeds/sensible amount of cores - packaged with good price) to allow You build this machine.. & then..let's hope OTOY is going to optimise those things in the end..at least Your rendertimes will be as good as possible.

glimpse wrote:=) I would not hold a breath for this being solved anytime soon, but ..I would lieke to be wrong..

Yam/Sebastian..if I could say something.. - go for sweetspot (high clockspeeds/sensible amount of cores - packaged with good price) to allow You build this machine.. & then..let's hope OTOY is going to optimise those things in the end..at least Your rendertimes will be as good as possible.

Yeah i guess it's the best way to approach it. Thanks mate.

smicha wrote: to grimm - 3.75x faster (probably due to 4GB of vram on 980 - the scene takes around 3.6GB, cpu is 5820k- 6 core @3.8GHZ) - do you have OOC texture on ?

Nope, I didn't change any settings. I'm wondering if there might be a difference in operating systems as well? I will try some other tests and see if there is any difference, I was running other processes when I did the first test (chrome, 3d-coat, etc).

Update: my 32-core system compiles the test scene in Octane V2.25 in 1 min. 28 secs.

Tutor wrote:Update: my 32-core system compiles the test scene in Octane V2.25 in 1 min. 28 secs.

I don't know why I start laughing..again after reading this.. =DDD..this is insane..