I use VRAY(among other renderers) and it recently went native for Apple Silicon gaining about 30% performance, making the apple laptops even better (faster than all other for laptop cpu rendering it seems, beating even 5900h).
But this native version of VRAY is not available when used from Houdini since you can't mix and match archs as I understand it. Or is there a way?

Oh how I dream of a future where H is fully native on this beautifully effective architecture. The bliss of working in total silence with 15+ hours battery while still having decent power is so lovely.

I think that once Apple releases the new Mac Pros perhaps SideFX will consider optimizing their code. Hopefully they will, but until they do I will not be renewing my license.

You would literally stop using Houdini due to this?
I know I've discussed before, but Mac users make up a tiny percentage, and inside that percentage an even tinier amount
are on the newer hardware. It's seems neither of you are aware of software development overheads/complexities, I agree the
new hardware is fantastic, but would you yourself seriously spend your dev time right now, on something a tiny fraction of
users are on? It will for sure be addressed, but I don't think the position you are taking is being fair in regards to how
long that hardware has been in the wild, and the percentage of users adopting it.

I'm not trolling, I'm genuinely asking how you see the development cycle in complex software.

tinyhawkus
I'm not trolling, I'm genuinely asking how you see the development cycle in complex software.

Here's the thing Lewis -- I know you keep harping on how small the percentage of MacOS users vs. Windows users, but fundamentally it's a question of "either offer proper support on all the platforms, or don't -- but don't pretend that all users are treated equally."

Either SideFX's goals are to bring powerful tools in the hands of as many artists as possible, or they aren't. I also find this pro-corporate profits-above-all type of thinking that you seem to embrace to be incredibly sad and limiting, especially in the creative community -- but I digress.

I aware of complexities of coding. I'm not asking for SideFX to reroute the team who might be working on a new version of Vellum, or the Karma coding to all of a sudden stop what they're doing to focus exclusively on MacOS. What I'm asking is for SideFX to hire (even on a part time basis) "someone" to work on MacOS optimization. It seems reasonable even by your own economic logic -- allot the resources in relation to the need.

I am going to bring up Blender (despite all the hate that it gets) -- the user base relation is similar, and yet even with a fraction of the staff and financial resources that SideFX has, they got the job done! And they're not alone -- Blackmagic Fusion and Resolve, the Adobe suite, AVID, C4D, and countless 3rd party companies. Are you implying that all of their code-base is considerably simpler than Houdini's?

I will not allow fanboy-ism to justify why SideFX has been obviously dragging their feet on MacOS (and I'm not even necessarily talking about M1 optimization -- I'm talking about working out basic buggy viewport behavior that's been present for quite some time now).

Working in MacOS is as important to my personal workflow pipeline as working in Linux is to yours (FYI I have a Windows machine running too and it does nothing but add extra work and maintenance drain to my already busy schedule). I feel that the only option I have left is to vote with my wallet. My renewal is coming up in a month and it seems like an appropriate time to do just that, even if ultimately it is just a small drop in the massive ocean of SideFX's earnings.

Midphase
the Adobe suite

Lol, Linux users surely have their own opinions on that

You really are thinking in the wrong way about this. Using Blender as an example is not comparable at all.
Do you have any idea how many more moving parts Houdini has compared to Blender?

How sidefx has a real commercial responsibility to paying customers to address and support the needs of them.
The blender Devs outnumber the sidefx ones by a long way, so you're mistaken there, they also don't have any
real commercial responsibilities, so can throw as many people as they want, and if things break, there's no
flood of companies literally stopping being able to work.

Yes, I am implying that Houdini is much more complex than most of the Apps you have listed, which points back to
you not fully understanding the depth of the software. Just the DOPs section of houdini would eclipse the codebase for Fusion
mate, it really is a big issue.

I get your frustration, but if you step back, can you not agree that turning a big ship in order to serve a fraction of users
is not going to be top priority. Of course they will do it, makes no sense not to. But if the majority of your Mac users are not
on that platform of hardware upgrade yet, it makes it a lower priority. I don't get how you can't see that, that you as a single
user can go out an buy a new laptop, but expecting houdini to pivot quickly because you chose to buy a new machine?

Midphase
Working in MacOS is as important to my personal workflow pipeline as working in Linux is to yours

I don't really have a dog in the fight, and I've accepted decades ago (literally) that platform discussions on forums or mailing lists are not really productive.

But I wanted to provide you this link, for a bit of light reading (only if you're interested, naturally). The Visual Effects Society conducts these surveys to get a feel of what's going on in the VFX world.
The condensed results have been published here: https://drive.google.com/file/d/15b-4GMTSEE9tyqeQdBfy_LZnxQIdp38Y/view [drive.google.com]

Almost 60,000 artist workstations at the participating studios, 11% mac, that includes the old ones. That's a nice data point for VFX.

I guess in the C4D scene (which was mentioned by you, that's why I bring it up) the mac usage is MUCH higher. Could explain the priorities.

That chart doesn't break it down to actual Artist's machines though. A lot of non production staff use Mac for
their work, I know I would! So when you factor in producers, design, etc, the users not very likely to be pulling
a houdini lic, or a third party renderer lic, that 11% is actually quite a bit lower.

I agree with him that the new hardware is amazing, I'm eyeing a macpro myself, but I'm under no illusions as to the
commercial responsibilities a company has.

does Blender run native on a apple M1? what i know is that cycles has M1 support under development.
cinema4d had a huge rewrite at the core and viewport that has been started years ago.

Source:
http://www.cgchannel.com/2016/02/maxon-working-on-new-core-architecture-for-cinema-4d/ [www.cgchannel.com]

i am pretty sure sidefx will do a metal based karma render engine, but that that will come when the XPU will be production ready.
currently apple M1 is still new and it just takes time, in one year we will have more information.
if houdini will have native support for M1 that is another question.

tinyhawkus
You really are thinking in the wrong way about this.

Lewis...seriously just drop it and feel free to butt out of this thread if you're so bothered by it. I'm sure if there was a Houdini issue that pertained to something that was important to you, you would raise a stink and a half about it (as I've seen you do for 3DL), so don't even.

mandrake0
does Blender run native on a apple M1? what i know is that cycles has M1 support under development.
cinema4d had a huge rewrite at the core and viewport that has been started years ago.

Blender is fully ARM/M1 optimized at this point (and has been for the better part of a year). Cycles is now running on Metal with GPU acceleration, with an official Metal support release planned for Blender 3.1 (due in about a month, but already working quite well in the beta).

tiny, my question in this thread was that I wonder if a m1 native plugin like vray can be used in some way on an x86 host or if we have to wait for a m1 native houdini. 30% exta performance would be nice and since it is already available for c4d and sketchup one might be a little keen on this.

l’ll refrain from addressing the flame war outside of saying this: mac is a very nice working environment for many of us and it would be nice to have Houdini fully working as soon as possible. This is one way to show concern.

Dude, you can choose to ignore the simple logic I'm presenting to you, or not.
I won't be butting out of it, because it simply comes back to exactly what I've outlined above
multiple times. If you want to disagree with what I've stated, that is your choice entirely.

I have simply responded to your statements with accurate counter points, the only result of that should be
feeling more educated about the actual machines turning to do all this.
It seems the point might still be being missed. It has nothing to do with being for or against any OS, or piece
of hardware, it's simply a matter of resources, and commercial realities.
When you are presented with the stats, and the real amount of software complexities, including how many people are
actually around working on these things, doesn't that temper your level of being upset?

Filip, no one in their right mind would say the performance leap is not amazing, it was never even suggested.
All I wanted to illustrate, was that there are realities at play, nothing to do with defending anything, I don't
care what OS/hardware someone uses, I simply found it a very odd stance for Midphase to take.


Midphase, if there was/are houdini issues that were bugging me, I'd be logging a ticket, and being realistic about the
use case, how many people would actually be affected by it, and the difficulty involved in solving it.
You have chosen to not be rational about it, and that is what I meant by "you're thinking about this the wrong way."


L

since recently i own both types of machines, solved all problems for me. logging the bugs both for win and mac :P

So that was a fun read

I think if we can remove some of the "heat" from this discussion, here are a few tidbits that lead me to believe that there will soon be a universal binary version of Houdini for Apple Silicon.

1) If you've ever watched any presentations by the Sidefx crew, especially while on-the-road, they seem to be using Macintosh an awful lot. So enough about the "small fraction of users are blah, blah blah."

2) I think I remember reading somewhere (sidefx website?) a statement that went something like "Apple Silicon not YET (my emphasis) supported" or similar phrasing that led me to believe it is being worked on.

3) I had an exchange with one of Sidefx's incredible team members where I said something like "please, dear VFX gods, be working on an Apple Silicon native version of Houdini" and in reply they said "We will make a good offering to the VFX Gods on your behalf." Call me optimistic but I read that as: "it's coming."

So, try and relax a little. I wouldn't be surprised if a universal binary Houdini will be rolled out when an actual Mac Pro with Apple Silicon finally ships or maybe even an iMac Pro this year. It would be a great time to showcase the power of the machine running one of the most powerful VFX tools on the market.

Hi. Flipping the debate around. How much further would blender be if they had taken the resources of a mac port and put that time into the core program. We are going to move back to maya for direct modelling having taught 100+ students blender. Blender was good, but two years later yes we have more os support, but very not enough direct modelling improvements to justify using it. Best

This assumption that there is some "finite budget" of man-hours and money for a company choosing to retool code for a universal binary version of software that somehow wastes resources to be used elsewhere seems very weak to me.

There is an existing tool (Xcode 12 and later) that will do the heavy lifting up front, so it's not like developers are starting from scratch. From what I understand reading over Apple's developer page, a universal binary is built automatically by Xcode when making a clean build and choosing arm64 from Apple's standard list of build architectures. It gets more complicated as programs deviate with custom makefiles and build scripts and probably a bunch of other stuff I sure don't know about, but that's why it takes far more complex programs longer to be ported. I wouldn't be surprised to learn that Sidefx has been working on a universal binary since day one.

Adobe gave us Illustrator, Photoshop, Media Encoder and Premiere in universal binary rather quickly. We're still waiting for After Effects (it's been in beta as universal binary since Oct 2021) but I assume AE is their most complex program to retool. Granted, Houdini is even more complex than that.

Universal binary ports on programs like Houdini and Cinema 4D and Blender, I'm sure, are not easy. But even if Apple users represent around 10% of their entire user base, any company will happily take a 10% increase in revenue if the resource investment is some smaller percentage of that. Especially given that people who buy Apple products tend to spend more money. As an example on the iOS side, Apple's app store generates nearly double the revenue of Google Play while the worldwide user base of iPhone users is dwarfed by phones running Android. Ignore Apple users at your own financial risk They're vocal, they're loyal and they evangelize.

Anyhow back to your assumption (or perhaps you are just being devil's advocate) that a company is somehow "robbing Peter to pay Paul" doesn't have much merit, in my opinion. If a company felt it was a waste of time and resources that would not pay off, they wouldn't bother. Of course I could be wrong, but personally, I'm confident Sidefx appreciates that a universal binary of Houdini will be worth it and that they have been working on it for some time.

Hi. Yes I am slightly playing Devils advocate. It will be interesting long term to see what happens with vulcan. We use windows at work rarher than Linux purely because of Adobe, Macs for us are out of our price range. We are also in no rush to move to windows 11 for years. If it ant broke, don't fix it. For our courses that do run on macs it is going to be at least 5 years before anything is upgraded. Maybe by then more software will be up and running. Best

rlvelocity
2) I think I remember reading somewhere (sidefx website?) a statement that went something like "Apple Silicon not YET (my emphasis) supported" or similar phrasing that led me to believe it is being worked on.
working on an Apple Silicon native version of Houdini" and in reply they said "We will make a good offering to the VFX Gods on your behalf." Call me optimistic but I read that as: "it's coming."

From the system requirements page:
macOS:

Requires 64-bit Intel-based or Apple Silicon Mac with macOS 10.13 and higher
Note: on an M1 mac it is NOT supported natively yet. You can run via Rosetta and therefore the performance won't be as good as it will be once native support is added.

1) If you've ever watched any presentations by the Sidefx crew, especially while on-the-road, they seem to be using Macintosh an awful lot. So enough about the "small fraction of users are blah, blah blah."

There's no stress here, just straight up logic and stats about the current reality.
Using Mac on the road for a presentation means nothing, not sure why you even point to that. They will port it, no-one is
saying they won't, but I'm still sitting here scratching my head trying to work out why people can't see the reasoning
for the delay.

The small fraction of users directly correlates to how much Dev you put immediately into something, and yes, there are finite
resources. Again, not sure how you are constructing this logic. The only reason any heat came into this, is purely down to
people wanting sidefx to pivot on demand, and not relaxing a bit when the realities are pointed out.
You can run it on the m1, just won't be as big an improvement. But it will come.

The last bit, those m1's are fantastic, but what type of simulations are you doing on a laptop where you are hitting walls
that getting 20%+ speed improvement would be life or death?