2019 iMac vs 2019 Mac Pro

[Peterjk]

I am in need of an upgrade and are once again puzzled by how ProTools uses multithreads. And I know that benchmarks will only disclose so much and not the entire real world reality, but for comparing PCs that's the closest thing I can find. A fully loaded 2019 iMac i9, will give me a higher single core performance than any of the new 2019 Mac Pro's and if I compare the iMac to the somewhat equally priced (if I include a number of external hardware with the iMac), but still more expensive 16-core Mac Pro the number are as follows:

2019 i9 3,6 Ghz iMac: Single Thread: 1250 Multi Thread: 8200

2019 16-core 3,2 Ghz Mac Pro: Single thread: 1124 Multi Thread:14502

While the iMac wins in Single Thread, the Mac Pro win in Multi thread.

And here's the 64.000 dollar question: For ProTools use - mainly mixing with not THAT many VI's, which one should I aim for? (NOT taking into account the added possibility and bonus to be able to upgrade the Mac Pro later).

(With the iMac I also need to use an external chassis for 2 HDX cards with all the hassle that brings to the table, but for this discussion better to leave those out.)

https://browser.geekbench.com/mac-benchmarks/

[JFreak]

When I changed my iTrashcan CPU from 4-core 3.7GHz to 10-core 3.0GHz there was a very noticeable performance increase with PT.

[Peterjk]

Thanks JFreak.

What 10-Core are you talking about? Would you happen to have the benchmarks for those two machines?

[JFreak]

I think it is the same CPU as in late 2017 iMac Pro. Apple never used that in the iTrashcan however. Totally custom configuration I have here.

This is the 10-core iMac Pro 2017 spec
This is the 12-core Mac Pro 2013 spec
This is the 8-core Mac Pro 2013 spec
This is the original spec of my originally 4-core 2013 Mac Pro

Bear in mind those GeekBench4 scores are not pure CPU benchmarks so obviously the newer iMac Pro 2017 spec shows bigger numbers. However 8-core and 12-core specs of the 2013 Mac Pro are comparable and you can estimate what the 10-core numbers might be.

In real life, this 10-core is slightly faster in single-core operation and only marginally slower in multi-core operation, compared to the 12-core iTrashcan. Point being that PT knows how to use multiple cores.

[Peterjk]

Thx Janne. I was only looking for the Single- and Multi thread numbers of the two machines you have been working with, and your opinion whether Single core ranks higher that multi, when it comes to general PT work.

[Marsdy]

Originally Posted by Peterjk

Thx Janne. I was only looking for the Single- and Multi thread numbers of the two machines you have been working with, and your opinion whether Single core ranks higher that multi, when it comes to general PT work.

You would probably already know if you were having a bottleneck with one core in Pro Tools. I’m only ever maxing out one core with heavy VI use. The new Mac Pro is not that much slower anyways.

The iMac only has one Thunderbolt 3 bus and as soon as you connect a Thunderbolt 2 device (HDX chassis for example) ALL the ports go down to TB2 speed. The new Mac Pro is a no brainer here IMHO.

[Peterjk]

Thx, Marsdy

Originally Posted by Marsdy

The iMac only has one Thunderbolt 3 bus and as soon as you connect a Thunderbolt 2 device (HDX chassis for example) ALL the ports go down to TB2 speed.

Isen't there a TB3 chassis?

https://www.sonnettech.com/product/e.../overview.html

Originally Posted by Marsdy

The new Mac Pro is a no brainer here IMHO.

Yeah, only it's much expensive than a 2019 5K iMac.

[Marsdy]

Originally Posted by Peterjk

Thx, Marsdy

Issn't there a TB3 chassis?

https://www.sonnettech.com/product/e.../overview.html

Avid doesn’t support TB3 yet. You need to use a TB3 to TB2 adapter according to Avid’s compatibility info.