iMac Pro / iMac i9.... Can't decide

[mattmower]

First off, if you're not in a hurry, pay attention to this thread.
https://forums.macrumors.com/threads/i-now-have-an-imac-pro-and-maxed-out-2019-imac-ama.2180758/

Not how things work. Really, they don't. Tests without real world tasks are meaningless. If an app uses multiple cores and the task is large enough, no single core example will be faster.

Scroll down and look at the performance differences of some apps among various Macs.
https://www.apple.com/imac-pro/




The T2 issue that plagued audio interface users (USB 2 interfaces and only certain brands at that) appears to have been solved by Mojave 10.14.4. Anyone notice how few BTO iMac Pros are in the Refurb Store since that happened (ok, today there are a bunch but that's rare)? The $4,249 base model is Apple's way of lowering the price—if they were actual refurbs, Apple would run out now and then (they never do).

Yes, I am following that thread with interest, thanks.

With respect to performance, I meant that the 10-core Xeon more closely approaches the single-core speed of the i9 than the 8-core Xeon. I'd also be hopeful that the better thermals and subsystems would allow higher performance longer. Single core performance while not the be-all and end-all is important to me.

Good to hear that the T2 audio issues may be fixed. As someone who does a lot of audio stuff that was a concern. If nothing interesting appears at WWDC I guess I will roll the dice on an iMP and hope I am not one of the people suffering T2 bridge crashes.

Thanks.

Matt

 

[messiry]

I am using universal Audio Arrow TB3 audio interface and it work just perfect, i have it for 2 months now and no issues at all with MBP and iMP

 

[mikehalloran]

Single core performance while not the be-all and end-all is important to me.

No reason for that.

Single core test scores exist to take space in reviews, make YouTube armchair experts sound important and confuse end users. Application performance is the only thing that counts.

 

[mattmower]

No reason for that.

Single core test scores exist to take space in reviews, make YouTube armchair experts sound important and confuse end users. Application performance is the only thing that counts.

I don't remember mentioning test scores. I am interested in single core performance because I use applications (my own for example) which are dependent on this.

Matt

 

[mikehalloran]

I don't remember mentioning test scores. I am interested in single core performance because I use applications (my own for example) which are dependent on this.

Matt

So, you write apps that perform worse on an 18 core iMac Pro than a 2017 4-core?

Oh dear... I hope that you are doing this for your own amusement only. As a business decision, it looks like a really, really bad idea.

 

[mattmower]

So, you write apps that perform worse on an 18 core iMac Pro than a 2017 4-core?

Oh dear... I hope that you are doing this for your own amusement only. As a business decision, it looks like a really, really bad idea.

Is it necessary to try to belittle me to argue your point?

Some processes are more amenable to being rewritten in parallel form than others. For those that are not, single core performance will tend to have an outsized influence on overall performance.

Matt

 

[GoodDoc]

I too have 2011 27" iMac that I'm itching to replace.

I was planning to replace it when the iMacs were recently updated, but as the 2019 models feels like they might be an end of life spec bump I'm hoping I can hang on for the next update to see if the iMac gets iMac Pro style updated internals.
Plus, my current iMac is still doing the job and given that I'm hoping to hang on to my next Mac for another 8 or so years, I'm willing to wait for another update.

The only thing that gave me pause for thought was a refurbished iMac Pro. When they pop up in the refurb store there is just a £155 price difference between an iMac with options, and the base iMac Pro.

2019 iMac Std - 8 Core Core i9/ 32GB RAM / 1TB SSD / Vega 48 = £4,004
2017 iMac Pro - 8 Core Xeon W / 32GB RAM / 1TB SSD / Vega 56 = £4,159​

Granted, if you buy third party memory you can save another £340 in the iMac, but for less than £4,160 you get the Mac with the more more modern internal design.

 

[jday7757]

So, you write apps that perform worse on an 18 core iMac Pro than a 2017 4-core?

Oh dear... I hope that you are doing this for your own amusement only. As a business decision, it looks like a really, really bad idea.

Many of the Adobe Creative Cloud apps, such as PremierePro don’t take advantage of multiple cores so I guess they’re not for business or professional.

 

[mikehalloran]

Many of the Adobe Creative Cloud apps, such as PremierePro don’t take advantage of multiple cores so I guess they’re not for business or professional.

Whoever told you that was severely misinformed. You really need to do some research.

There is a big debate going on in Adobe world about this. The crux being that 14 cores is considered the current breakeven point for PremierePro for most users — after that, many don't believe that the minor increase in performance is worth the major increase in cost. But that was last year.

As you do more reading, you'll find that the i9 isn't up to the X where PremierePro is concerned either. This goes back to the original question.