Maxwell Render | Official public forum

i think this email was removed/stopped, because i didn't receive it, and normally i receive the mails from nexlimit.... anyone else ?

The point was to tell those who actually got the mail that
there won't be an RC6 very soon.

Kabe

There's a new email sent from NL now ?! hmpf i still haven't got the first one

/ Max

Maximus3D wrote: There's a new email sent from NL now ?!

no, there's no other email apart from the cooperative one, which I also got this morning.

I got mine this morning too...second tier email list I suppose.

I don't understand the message either and statements like this...

"time barriers are broken and are no longer an issue"

Whew! I am glad they solved that issue !

Unbelievable.
Paxton

However, 10 divided by 9 = 1.11 hours, not 2.5 hours, so its not 100% scalable.

Not exactly as the speed of the machines used is different.

So: (2.2ghz x 10) * 2 = 44 ghz hours
and: (2ghz x 2.5) * 9 = 45 ghz hours


Kandor

We, here at Jestico + Whiles were just about to purchase another licenese.
I told the IT group to hold back because I have second thoughts. In fact,
if the CPU license policy doesn't change dramatically I'm out. That would
be so sneaky not to change it but then wouldn't surprise me.

I wonder where this is all going....

J

I got it this morning too.

I guess that the point is that we'll have cooperative render which didn't have before. Nothing else. (Something we knew since Victor's video, right?)

The point was to tell those who actually got the mail that
there won't be an RC6 very soon.
Kabe

That's a good point too: "Hey guys, this is going to take long; we have to announce something to calm down the people"

b-kandor wrote:

However, 10 divided by 9 = 1.11 hours, not 2.5 hours, so its not 100% scalable.

Not exactly as the speed of the machines used is different.

So: (2.2ghz x 10) * 2 = 44 ghz hours
and: (2ghz x 2.5) * 9 = 45 ghz hours


Kandor

Thanks for pointing that out....
Would make more sense to use the same systems for their 'scientific' analysis. If they have 9 CPU's @ 2 GHZ each to process the scene, wouldn't it make sense to take two of those 2 GHZ cpu's and process the same scene? Why a different processor? makes no sense to me, where is the scientific rigor!

paxton

Ernest Burden wrote: However, 10 divided by 9 = 1.11 hours, not 2.5 hours, so its not 100% scalable.

Not an accurate calculation!

9 cpus divided by 2 cpus is 4.5 times the cpu power.

10 hours (being equal to 1 render unit) divided by 2.5 is 4

That's a .8 loss per unit.

Not bad at all!

And for the record: This is great news for me! M~R can utilize multiple cores well enough to speed up the process with little loss of cpu power. I've always planned on upgrading my farm at some point and it will help greatly that my chosen renderer will benifit during visulization.

As far as other renderers already doing this: This is a unique renderer and I'm just glad that it works in this one

b-kandor wrote: Not exactly as the speed of the machines used is different.

So: (2.2ghz x 10) * 2 = 44 ghz hours
and: (2ghz x 2.5) * 9 = 45 ghz hours


Kandor

Ahh I missed that...

Or does it mean that they will not optimise Maxwell for speed, but expect us to use more CPU's

No that would be a highly unlikely choice to do, i cannot see that happening at all. Optimization is one of the cornerstones in the developement, that been said many times before.

/ Max

martindejager wrote:Or does it mean that they will not optimise Maxwell for speed, but expect us to use more CPU's

Ehm, I pointed out earlier that only huge speed improvements (say
something like 3-5 x at least) would really change the picture. They might
do this but frankly I doubt that there is so much performance left on the
table right now, because then the basics would not be optimized well
enough.

If they can keep the speed while including new features it's not too bad.
Actually the alpha is the fastest renderer until now, while the RC5 has the
best materials so far but is slower in most of my scenes.

Kabe

Maybe somone could elaborate on how you take separate low quality images and merge them into a singular, superior image. ......definitely not clear on that.