Cache node from dbwTools not working on AMD ?

[pyo]

Is this possible that Cache node is not working on AMD processors ?
I have a scene where I used the cache node and my computer at work renders it in about 18 seconds (Intel Q6700 2.66GHz) but at home this scene renders in 55 seconds ! (AMD phenom x6 1090T 3.2GHz).
And it looks like the problem is the cache node, cause it doesn't matter if I set it to "pass through" mode or not the render time is the same.
On the Intel machine when I set "pass through mode" rendertime is 1m30s.

Could anyone test the case ? Maybe something's wrong with my system ? :help:

Here's scene.

[Lightwolf]

I'll have a look.

I should work on an AMD as well though, as it's been developed on an AMD dual core.

Cheers,
Mike

[Lightwolf]

Hi pyo,

there's a bug in the new version. I changed some code to make things more compatible with FPrime, but that also added a bug causing the cache never to be used. D'oh.

Shoot me a mail and I'll send you a fix (also tell me for which platform): michael(dot)wolf(at)db-w(dot)com

I'll also update the public release to V1.6a soon.

Cheers,
Mike

[Lightwolf]

This is actually a really nice use of the Cache node, the render time goes down from 3m64s to just 44s on my dual core Athlon.
A boost of more than 550% - this is exactly what I had in mind when I wrote it. :D

Cheers,
Mike

[pyo]

Thanks for fast response. I've just send You an email.
Thanks again.

[Hieron]

Holy cow.. that's both a great use of the node and one big nodal flow. Thanks alot for sharing this pyo. Will take me a while to plow through though.

I would have used HDI and mapped some big PS'ed image on top of it, but this sure works as well and allows some nice tweaking. But surely it must have taken a good deal of time to set up?