Jump to content

Mac Silicon OS and VW


Recommended Posts

@Learning Vectorworks what’s the spec for the iMac you have? - if it the mid range iMac with Radeon pro 555 then the gpu may be ok for some light 3D modelling (if it’s the base model then it’s intel iris gpu and is not much good for vectorworks- so def time to cut your losses and move on)

 

 I’d suggest (like m5d said try an ssd drive to speed up the boot process and stop the wheel) you can purchase  a decent sized ssd drive ~£50 for 500gb (ideally needs to be at least same size as whatever amount you’ve used on the internal one so you can fit everything on the drive and allow for some extra breathing space ) buy a thunderbolt/usb3 drive external enclosure (they’re really cheap) - copy the whole drive over to it using the trial version (no need to purchase) ‘carbon copy cloner’- and try booting off the ssd drive by holding down Alt key at boot- see what a difference it makes - may get you through next year or so?? If it’s no better you can still use the drive and enclosure for another Mac/pc so it’s won’t go to waste.

Edited by neal-2002
Link to comment

Hello everyone.

 

Will VW have 2 installer, one for intel Mac basics and one for M1? Off course I don't mean using Rosetta 2, but natively for M1.

 

In my opinion, if i were to buy a new mac i would choose a mac mini with M1 and wait patiently for the first iMac with M1.

 

Sometimes I think of those in the world who bought a mac pro to discover Silicon a few months later and I don't know whether to laugh or feel sorry for him.

 

No doubt for the CPU, but many for the GPU, especially for the developments planned for Redshift integration etc.

 

I'm afraid that the goal for 2022 will become to move VW to M1 and not carry out integrations on GPU-based rendering, which is essential. 

 

Of course this will also depend on Maxon. But we all saw what was the only professional software shown during the M1 presentation ...

 

 

  • Like 1
Link to comment
4 hours ago, Zeno said:

Will VW have 2 installer, one for intel Mac basics and one for M1? Off course I don't mean using Rosetta 2, but natively for M1.

 

If you have your Code clean/ready for Apple Silcon, you can compile a Universal Binary.

This is an App Package that contains all versions and macOS will install what is

appropriate for the computer., No matter if Intel or ASi based.

  • Like 3
Link to comment

So we know that the M1 Mac Mini beats the current top line iMac 27" in CPU speed, but it would be interesting to see real world graphics results. Here's my guess from looking at GPU comparison sites: Given that it is about on par with the RX 560X, it looks the Mac Mini could beat the graphics performance of the baseline 27" iMac with the Radeon 5300 but the topped out 27" iMac with the Radeon 5700 XT will be about twice as fast.

 

Of course the Mac mini is going to run out of RAM much quicker than a maxed out iMac 27", which will have a big impact for a large BIM model, but its 1/4 of the price of the iMac with the Radeon 5700 XT.

 

Once we know how well vectorworks performs on the M1 processor, the 16Gb M1 Mac Mini could well be a good buy for 2d work and it might still work pretty well for domestic size projects in 3D. For the 200-300 m² projects that are typical for me I have single file BIM models which are about 100-200mb on disk and Vectorworks is using 2-3gb of RAM when I have one open, although this goes up to about 10gb when I update section viewports.

 

I'm really looking forward to the Apple Silicon iMacs and high end MacBook pros now!

  • Like 1
Link to comment

It's a case of "yes and". Yes, these are excellent machines and you will have buyer's remorse a year from now when the next gen is released. So, it's a win-win!

 

Thankfully, the improvements to v2021 like multicore processing and graphics caching are helping me ride out the transition on this 2014 MBP that would get trounced by the low end M1 machines that were just released.

 

ecf.gif

Edited by Mark Aceto
  • Like 2
  • Laugh 2
Link to comment

Hmmh,

the first generation of M1 SoC for ultra portables,

may not exactly be what VW users should look at at all ....

And first generation Apple devices may be already obsolete

after 2 years (my Mac Pro 2.1, iPad 1, Trash Can, ...)

 

But these are (native code) still faster in Cinebench R23

than my 6 core Trash Can I am currently typing on.

Already forgot the numbers but AFAIK it was more than 2x

in single core and around 40% faster in multicore (?)

OK, there may be a 25% Rosetta2 handicap.

I am pretty sure that GPU may also not be much slower than

my dual D700s.

 

So I think it would not be such a mistake to play with one

of these now.

On the other hand I more and more think the whole ARM

transition will not really take 2 years and the next iteration

could hit us already in Q1 2021.

  • Like 2
Link to comment

Yes- I keep saying it’s not for me....just wait for M1x.....

...but then keep reconsidering as it’s all looking pretty good- MUST STOP LOOKING AT THEM!

 

just know I’ll have buyers remorse within days of getting one...just like Mark says...🤣🤣

 

and it doesn’t help when Vectorworks have nothing to say on the matter either....

Edited by neal-2002
  • Like 2
Link to comment
31 minutes ago, neal-2002 said:

MUST STOP LOOKING AT THEM!!🤣🤣

 

On the other hand,

if your iPad Air 2 starts becoming battery problems,

you saw a video how to replace the battery by opening glued connections

by hot air and and tiny tiny plugs ... and you old eyes want something bigger

anyway.

A basic fanless Macbook Air M1 will be more affordable than an iPad Pro 13",

better specced and you can type on it (?)

 

 

OK, nothing to touch and no real Camera and LIDAR

Edited by zoomer
  • Confused 1
Link to comment

^ 🙂

 

As I said, these current devices are not really thought for CAD and 3D usage

and it is obvious that there will better options coming (soon?).

 

But I am sure that there are lots of VWs outside in real world, that run on

older, even less capable Mac/PC hardware than that.

So I am very looking forward about your testing results.

(especially if Apple Silicon SoCs with ARM e.g. need a bit less memory than

standard PC RAM, for any reasons and current RAM limits will not matter

that much. If you compare the same projects on both architectures)

Edited by zoomer
  • Like 1
Link to comment

I need a new computer, but not desperately.  I've decided to wait until mid-2021 before buying.  The recent reviews of the MacBook Pro M1 seem too good to be true, so I would better wait and then decide!

 

Also noticed that the Big Sur update has marginally improved the speed of my 21.5 inch iMac with 1tb HD.  Its not great, but I'm just learning patience.

  • Like 2
Link to comment

VW has a lot of potential balls in play that could dramatically move performance forward with the next version. Shifting to Metal should help everyone on a Mac (existing machines and M1 emulation). Shifting to the new version of the C4D engine could be amazing for rendering on an M1 machine. I would imagine that a native M1 version of VW is not likely until VW2022.... but at least there are interim steps.

 

Kevin

  • Like 2
Link to comment

@neal-2002 "Comparison is the thief of joy."

 

On the one hand, a strong case could be made for buying a $1700 Mini now (the one with ports and a fan), and gifting / reselling it a year later. On the other hand, I'm still on Mojave, waiting for Catalina supplemental update 3 to upgrade to 10.15, so I'm not eager to beta test Bug Sir. I currently have my bare 2014 MBP operating in clamshell mode elevated (for maximum airflow) on the desk next to me.

 

Although I totally get Apple's strategy to upgrade from the ground up, I was questioning the rumors, "Would they really release machines at the low end that will outperform / cannibalize sales at the mid to high end?" Adding insult to injury, professionals like us who've been ignored by Apple while they continue to update everything they can except pro machines have to continue waiting another 6 months to a year... How many years have we been playing that game? Every year.

 

We need to form a support group: Remaining Mac Together.

  • Like 2
  • Laugh 1
Link to comment

I wonder if it is possible,

in the meantime,

beside just compiling a working Rosetta2 compatible App,

(because you also need to wait until all your 3rd party Libraries are ASi ready too)

to do some optimizations in the own code that make the code better for Rosetta2's

translation to ASi. Just for the most important and demanding tasks.

 

Like adding a bit of Metal API calls into graphics engine or something making

use better use of shared Memory or even changes in data handling, in a way

that Apples API can send it to the ML cores or such things.

Something needed later anyway, if the final target is Apple Silicon optimized

Software for the Mac Version.

 

Is there a potential to bring the Rosetta2 penalty of about 25% a bit down

that way ?

  • Like 1
Link to comment
19 hours ago, zoomer said:

Is there a potential to bring the Rosetta2 penalty of about 25% a bit down

that way ?

 

Translating Intel machine code into ARM machine code will always be costly, whatever the technology used for it.

 

(I do not think it is worth to try to do any optimization on the Intel version of VW, if in the long-run VW will be compiled for ARM directly.)

Link to comment
3 hours ago, Nicolas Goutte said:

(I do not think it is worth to try to do any optimization on the Intel version of VW, if in the long-run VW will be compiled for ARM directly.)

 

 

AFAIK,

 

If you developed an App for Intel Mac, and kept your code modern and optimized

for Apple over time, which means, every now and then updating code and using

Apple's latest APIs and Tools as recommended,

you wouldn't have to change or prepare your code for Apple Silicon much at all.

Just compile it.

Which worked that way for many Apps where developers talked about a needed

adaption work time from 5 minutes up to half a day.

 

It is just when e.g. you still use OpenGL, ignored the deprecation of it and Metal API,

old Python and QT versions, special custom optimizations for Intel, exotic 3rd party

libraries or such things,

you will have a much harder time adapting to Apple Silicon.

In some cases even problems to get you App running via Rosetta2.

 

Unfortunately this is usually the case for many complex Cross Platform 3D or CAD Apps.

 

 

So I think,

everything you do now to optimize the code for Apple Silicon,

will also help the x86_64 App version.

Edited by zoomer
  • Like 1
Link to comment
3 hours ago, zoomer said:

Unfortunately this is usually the case for many complex Cross Platform 3D or CAD Apps.

 

 

Not to be misunderstood,

there are valid reasons why this is the case.

 

And VW now stated they will support Apple Silicon natively.

That makes sense as they have about Windows/Mac 50/50 user base.

There a lots of cross platform Apps offering a macOS version that have

far less macOS users: and you have to see effort vs yield,

that may just have to abandon Apple support.

 

 

Link to comment

As I look into my crystal ball, and try to predict the future upgrade path (where there is no roadmap), here are some stray obvservations about the current 2020 MBP 16":

  • I was expecting the usual GPU bump (or price drop) but it seems like that may have already happened this summer
    • Assumption here is that this could be the last Intel MBP 16"
    • However, I'm still 50/50 on whether this is actually the penultimate MBP 16", and they eek out one more spec bump next year
  • I was also hoping for a 10th gen CPU update that would run cooler
  • This is the last MBP that will definitely natively run Windows in Bootcamp (that we know of for now)
  • This is the last MBP that is compatible with an eGPU (that we know of for now)
  • This is the last MBP that will install Catalina
    • It's finally stable after a year of updates (as per usual with Apple)
    • For those of that choose a more conservative / delayed upgrade path, this machine that runs a .7 OS is available now as compared to running a .01 OS on a new machine
  • It has 4x Thunderbolt 4 ports (and we know how Apple loves to kill their darlings)

A little bit more context:

  • All of the bullet points above will hold a strong resale value for a small set of users that need those features (whether we do or not)
    • The M-whatever out in the wild will make this machine invaluable (irreplaceable)
    • Think how in-demand the 2015 MBP became (the Butterfly Effect)
  • All of the current M1 comparisons thus far have been against low end Mac's or "best selling Windows laptop" (WTF that means) which leads me to suspect it wouldn't destroy this thing
    • In boxing, these matches are called "tomato cans" to build a fighter's confidence
    • What the longterm future holds for Apple ARM is exciting but the next 6 months to a year are uncertain
  • The imminent M-whatever will probably beat this machine but... 
    • By how much? Enough that we will feel buyer's remorse for longer than a week?
    • At what cost? Will they remove ports? eGPU compatibility?

It sucks that we don't know if Apple will give this thing a spec bump next month but the more I come down from the fever pitch, the more I go back to my initial gut feeling: this is the machine to buy right now, and then upgrade sometime in 2022. Then leapfrog over the next gen industrial design refresh in 2023 when they replace the physical keyboard with a Touch Board, remove all the ports... 

 

really resent Apple for not providing a roadmap for their professional user base that depend on their machines for work. And not just us but also our clients, artists, teams, colleagues, crew, vendors... but until they move to a more predictable upgrade path with the M-series, we're still playing Mac Roulette. Although, I absolutely would recommend that "Enthusiasts" hold out for the M-whatever sometime next whenever... Professionals that are using 6+ year old machines with 2GB of VRAM, and holding out for the "smarter purchase", maybe that isn't the smarter business decision for the next year or two.

 

Really curious what other folks in our Mac support group see when they read their tea leaves, so I can bet on the right horse this time (see my signature for the losers I've bet on in the past 5 years). If you're the adventurous type, an ARM Mac Mini or MacBook Pro running Bug Sir could be a smart move (hopefully your 3Dconnexion mice still work). If you're less adventurous, the 2020 MBP 16" running Catalina could be a smart move (if it ain't broke, don't fix it). However, riding out the next year (or more) on what will shortly become a 7-year old machine as we take on more work during a pandemic seems like it could be the dumb move (notwithstanding budgetary restrictions; gotta put food on the table).

 

1461544242_ScreenShot2020-11-21at11_00_34AM.thumb.png.70215361c0ee0acba064371f580f8e6a.png

Edited by Mark Aceto
Link to comment

It is mainly just about if you still need Bootcamp or Parallels VM Windows or not

and if your important Software will already run on Apple Silicon, reasonably or not.

 

If you need a mobile Mac right now,

and you need the highest CPU and GPU power available, go with a maxed out

Macbook Pro 16".

 

If you can wait for another 6-12 months and need a mobile Mac,

need no Windows at all and your important Software announced ASi support,

wait of course.

 

If Windows isn't necessary, you don't use special complicated Software not

ASi ready and need a mid level Macbook (Pro) or Mini now.

Buy a M1 device from Apple imedeately.

 

M1 devices are far more powerful than the Intel devices that they replace,

at the same price with more battery - a no brainer.

 

 

I got experiences from a user of my other Software.

His M1 Macbook Pro 16 GB is noticeably faster in CPU and GPU than his

15" (2017?) MBP, in any situations, also for our other CAD.

But there are issues with that CAD under Rosetta, like opening a new blank file

or just too complex files will crash for now. So if that's important, there are

still Intel Macs, without uncertainties or problems.

 

And we still wait for VW Apple Silicon testing feedback.

If everything would look bright, we may have had a report already.

Edited by zoomer
  • Like 1
Link to comment

For context, it's stuff like this that keeps me from updating to the latest macOS release every year (and buying a Mac that will only run that brand new release):

 

https://www.fujitsu.com/global/support/products/computing/peripheral/scanners/scansnap/faq/bigsur-ss.html

 

6 of 8 apps in their bundle are not compatible yet.

 

It usually takes about 3-6 months before all software / drivers are compatible. Sometimes it take another 6+ months to fix bugs in the same apps / drivers (3Dconnexion).

 

Hopefully, removing kernels in Bug Sir will at least help with drivers from now on... 

  • Like 1
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...