Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Hunter5117

macrumors 6502a
Original poster
Mar 17, 2010
569
401
First I will admit I have not kept up with the details of how the M1 chips handle multi-core/multi-threading of applications.

Used to be that an app had to be specifically written to take advantage, and generally only high-end apps like Final Cut, Adobe apps, etc really made use of it.

Is this still the case with the M1 chips? Or has Apple done some silicone magic and allowed the SoC to manage this regardless if it is written into the software?

Just wondering because it certainly would set a limit on how practical these multi-core beasts are for the average everyday user.
 
The app has to be written to use multiple cores, but it is fairly trivial. Both of my company's apps (EarthDesk and Cinematica) are multi-core capable. It's a good way to do things when tasks can be broken up in parallel without dependencies. That simply means that the results of one operation are not needed to perform the next operation. In our case, rendering images or importing video are easy to break down into tasks for multiple cores.

I would think a whole lot of general apps that perform real calculation-intensive work are written like this.
 
  • Like
Reactions: severe
The app has to be written to use multiple cores, but it is fairly trivial. Both of my company's apps (EarthDesk and Cinematica) are multi-core capable. It's a good way to do things when tasks can be broken up in parallel without dependencies. That simply means that the results of one operation are not needed to perform the next operation. In our case, rendering images or importing video are easy to break down into tasks for multiple cores.

I would think a whole lot of general apps that perform real calculation-intensive work are written like this.
Did you run tests of Cinematica on M1 Macs? How does it perform? Do you have any information if (and when) a native M1 version of Cinematica is on the horizon? Cinematica is the only app (that I can't live without) that prevents me from switching to a M1 Mac. I would pay (almost) any price for a native M1 Cinematica version.
 
Did you run tests of Cinematica on M1 Macs? How does it perform? Do you have any information if (and when) a native M1 version of Cinematica is on the horizon? Cinematica is the only app (that I can't live without) that prevents me from switching to a M1 Mac. I would pay (almost) any price for a native M1 Cinematica version.

Cinematica v3.6 is native on M1 (and contains native Intel code as well in a fat binary). It does run faster than Intel in general but imports are quite dependent on the IO speed to the device that videos are stored on.
 
  • Like
Reactions: umue
Cinematica v3.6 is native on M1 (and contains native Intel code as well in a fat binary). It does run faster than Intel in general but imports are quite dependent on the IO speed to the device that videos are stored on.
Wow, that is good news! You really should put this information on your website (publicly available release notes would be best). On my Mac Mini 2018, Cinematica has little lockups (spinning beachball, 1 to 4 secs) on a regular basis. But other than that, when it runs, it runs pretty fast (7000 files, ca. 900 smart collections). Now I'm looking forward to jump on the Apple Silicon train. :)) IMHO, there is no substitute for Cinematica regardless of the OS platform. Keep up the good work!
 
You should have automatically received release notes via the in-app updater. We are working to redesign the site to hold release notes as well.

900 Smart Collections is more than we have tested it with... that may be a record. The number of collections should not affect performance, only the total number of files.

We are currently working on EarthDesk 8, then Cinematica 4. Thanks for the kind words. Saving a very large Library can result in the beach ball as it does a safe save and all the video thumbnails, etc have to be relinked by the OS which can be a bit of a slow process.
 
Last edited:
First I will admit I have not kept up with the details of how the M1 chips handle multi-core/multi-threading of applications.

Used to be that an app had to be specifically written to take advantage, and generally only high-end apps like Final Cut, Adobe apps, etc really made use of it.

Is this still the case with the M1 chips? Or has Apple done some silicone magic and allowed the SoC to manage this regardless if it is written into the software?

Just wondering because it certainly would set a limit on how practical these multi-core beasts are for the average everyday user.
Looking at chess engines, used by more than 200.000.000 active chess people everyday, you can easily see that the M1 chips cpu and gpu are extrem slow. Integer math is very slow too. Even more than 10 year old cpus are faster.
Apple wasn’t even able to sell a MacBook Pro with M1 Ultra chip and 16 TB SSD.
 
Looking at chess engines, used by more than 200.000.000 active chess people everyday, you can easily see that the M1 chips cpu and gpu are extrem slow. Integer math is very slow too. Even more than 10 year old cpus are faster.
Apple wasn’t even able to sell a MacBook Pro with M1 Ultra chip and 16 TB SSD.
Looking at chess engines, tells us that these run slow on M1 Macs, but not if M1 Macs are "underpowered" or "slow" in general.

I'm a Stockfish user myself and follow everything about current and future M1 support (I do not own an M1 yet). My impression from all the talk online (including many devs) is, that Stockfish is poorly optimized for the M1 platform at this time, and nobody really gives a f**k about M1 Macs (Maybe I'm wrong here, but that's my impression).

"Even more than 10 year old cpus are faster." ??? Sorry but that's utter BS. Stockfish was as fast (not faster) on a 10 year old i7 3930k in a test by one guy in this forum a year ago. That's quite different from the general statement that "M1 chips cpu and gpu are extrem slow".
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.