iMac Pro 10 core vs 18 core

[SFjohn]

I'm very sure the processing power will be able to last 10 years. I'm using the 2013 Mac Pro and it's as fast as it is on day 1. But I'm not sure if the screen can last that long though. My monitors usually start to turn yellow in year 4-5 with intensive use, and with some colour calibration they still work, but the screen durability is the main thing to worry about on the iMac.

Good point! I’m sure we can add a second screen if the main will no longer calibrate correctly...
[doublepost=1513724457][/doublepost]

Well, in seven years Apple will announce the iMac Pro obsolete, and you get no repairs and replacement parts.

(RX Vega 64 with 300 Watt I'm looking at you)

Good luck! Especially for the ones with maxed out iMac Pro.

Agree the RX Vega 64 will likely be a limitation as time goes by, hoping Apple will improve their EGPU support in time.

 

[SFjohn]

Wow... what a great iMac Pro you will have! Here’s to hoping it provides you with that ten year scope!

I went for a 10-core, 64GB RAM and 16GB Vega for FCPX and 3 streams of 4K. For me, I went with the maximum of what I could afford, so I am really happy with that configuration and hope that it lasts me for six to seven years. I may update to 128GB RAM if I ever need that much.

As it stands though, the 10-core is plenty fast enough for FCPX and I’m really pleased as to how well optimised it seems to be for the new machine. This seems to be a much better purchase for FCPX than the Mac Pros were, given that FCPX now seems to be extremely optimised for the iMac Pro, and yet a maxed out 2017 iMac could transcode video faster than the Mac Pros.

Joema2 is such a helpful poster and he was able to guide me in my purchase before the iMac Pros came out. Massive thanks to Joema2!

RuffDraft, congratulations, that’s an awesome machine and likely the best bang for your buck anyone could get. I have no doubt the 10 core will match or beat the 18 in different situations. I’m still running a maxed out mid 2010 iMac, so I personally have no doubt about your ability to get 7 years from your machine if you want to.

 

[OBirder]

I could be wrong, but for LR and PS, I don't think you'll see much of a difference between 10 cores and 18 cores.

I would think that might be incorrect based on usage.

When I come from a photo shoot of 1000 to 2000 RAW images, I need standard views and 1:1 previews (based on monitor size) rendered. That with my current 6 core i7 PC is using all 12 threads and takes several hours.
So I would assume that additional cores will improve that quite a bit.

 

[RuffDraft]

RuffDraft, congratulations, that’s an awesome machine and likely the best bang for your buck anyone could get. I have no doubt the 10 core will match or beat the 18 in different situations. I’m still running a maxed out mid 2010 iMac, so I personally have no doubt about your ability to get 7 years from your machine if you want to.

Thanks SFjohn - that's great to hear! I'm also pleased that the 2017 iMac screens don't seem to be suffering from any problems so far... I'm hoping the screen will look just as great in seven years as the computer will run etc.!

It's great that your 2010 iMac is running so well too! My 2011 MacBook Pro has had its moments with the death of the GPU and a replaced motherboard / GPU in one go... but it seems to be running really well at the moment, though it shows its age at times and of course, performance wise, isn't that great in comparison to what's available now for 4K multi-cam edits.

Definitely hoping that the 10-core's turbo boost is legitimate and not there to make people go from 8 to 10 cores... it's strange that the 10 can turbo boost more than any of the others..!

 

[shadowmatt]

Now thats a reply the internet can be proud of.

 

[Trebuin]

I suspect the 10 and 8 core will probably match single core in performance when tested over time. The multi-core will still be much faster, 20%, so that's the gain.

 

[nickster777]

Wow... what a great iMac Pro you will have! Here’s to hoping it provides you with that ten year scope!

I went for a 10-core, 64GB RAM and 16GB Vega for FCPX and 3 streams of 4K. For me, I went with the maximum of what I could afford, so I am really happy with that configuration and hope that it lasts me for six to seven years. I may update to 128GB RAM if I ever need that much.

As it stands though, the 10-core is plenty fast enough for FCPX and I’m really pleased as to how well optimised it seems to be for the new machine. This seems to be a much better purchase for FCPX than the Mac Pros were, given that FCPX now seems to be extremely optimised for the iMac Pro, and yet a maxed out 2017 iMac could transcode video faster than the Mac Pros.

Joema2 is such a helpful poster and he was able to guide me in my purchase before the iMac Pros came out. Massive thanks to Joema2!

How much storage did you order?

 

[kschendel]

Does anyone have examples of workflows which would benefit from 14 or even 18 cores compared to 10 cores?

Analytic database processing on large (multi-billion row) datasets. However, to get the benefit of 18 cores you might need a faster I/O subsystem, generally meaning more storage devices to split the load across. I haven't run the numbers on the iMac Pro so I'm not sure if it would be I/O limited or not.

 

[RuffDraft]

How much storage did you order?

Standard 1TB SSD. I am not that bothered about internal storage - I'd have taken 512GB if that was the base option. 1TB just means that I can keep my music on there; otherwise, I only really use Adobe's apps and FCPX with iTunes and Safari.

 

[mcewangus]

I calculate the 18 core would be approx 20% faster than the 10. However that’s just multiplying numbers... Does anyone know how the larger L3 memory would factor into speed? Single treaded it looks like the 10 core is 5% faster than the 18 core... For arguments sake, assume cost is not a factor. But overall performance is. Uses are Final Cut X, Logic, Lightroom, some Photoshop (not much). Thoughts?

It would in fact appear that the 14-core is 34% (multicore) faster than the 8-core and 18% (multicore) faster than the 10-core. This being true, the 18-core is very likely to exceed 20% over the 10-core.

https://browser.geekbench.com/macs/427

 

[jerwin]

It will be interesting to learn whether the Spectre/Meltdown mitigation fixes will substantially depress Final Cut Pro performance.

Apple has been treating the SSD as an adjunct to main memory for a long while, and I would not be surprised if some performance optimization strategies are no longer teneble.

 

[bxs]

In general more cores will always mean greater performance overall. There are two exceptions with one being unrealistic

1. Some workloads will actual cause performance to decrease with an increase of cores. However, this normally will only happen when you start throwing 100s of cores at a parallel task that has a very small serial piece of code.
2. The unrealistic case for most casual users is that running a task that is single threaded will perform best on a processor offering the highest GHz. However, in macOS there are 100s of small processes always requesting CPU cycles so this is why its unrealistic.

Ignoring cost to a large degree, then always max out your computer. Doing this will stop any future angst about having not done so.

As another note related to the stated Turbo Boost value for any of the iMac Pro Processors. From my benchmarking using the Intel Power Gadget and the Stockfish Chess engine the only time you will see the Turbo Boost reach its maximum value on a single core will be by configuring all cores except one to be disabled which can be done with the Xcode Instruments application. This configuration is simply not practical (it may even hang the system at times) and the other cores when running in production will be enabled and no one core will ever be able to run at the max Turbo Boost frequency because macOS has 100s of processes executing at any given time to use available cores. The bottom line here is, the max Turbo Boost will never be achievable on any core running a production system.

 

[anticipate]

I went for a 10-core, 64GB RAM and 16GB Vega for FCPX and 3 streams of 4K.

Depends on the CODEC for sure

My 15" MBP could do three streams of 4K ProRes in better performance mode. Even my little 12" MacBook can do a couple streams of 4K media converted to proxies... (love how easy that is in FCPX)

For rough, from memory (it was late) tests:

My 8 Core / Vega 56 iMac Pro with 32GB of RAM, in FCPX, could run at full fps in "Quality" (full rez) - without rendering or transcoding:

2 Streams of DJI Mavic 4K with Filmconvert applied
3 Streams of Mavic without filters applied
6+ streams of ProRes 422 4K with LUTS & CC applied (with 2 of those clips having FilmConvert, which seems to be the limit as it maxes out the GPU)
6+ streams of XAVC-I 4K from a Sony FS7 w/LUTS and CC and 3 clips with FilmConvert applied
4-5 streams of XAVC-L 4K from a Sony FS5 w/LUTS and CC and 2 clips with FilmConvert applied.

My old 8 Core D700 Mac Pro could maybe (from memory) handle 1 Mavic stream and 3 ProRes streams with one film convert. All in all it's about a 2-3X speed increase over that Mac Pro.

The 8 Core base model is FAST, and this 10 core being the sweet spot for FCPX is a slight myth, except you will get 10-15% faster render times. But faster than fast...

RED RAW was even more amazing, as I was able to play (in best performance) 1 stream of 6-8K RedRAW footage with filmconvert applied on one stream in real time (!!). I didn't try two but I suspect it could handle two-three streams of 4K easily. That's stunning as I couldn't even get it to playback decent on my 15" MBP.

 

[RuffDraft]

Depends on the CODEC for sure

My 15" MBP could do three streams of 4K ProRes in better performance mode. Even my little 12" MacBook can do a couple streams of 4K media converted to proxies... (love how easy that is in FCPX)

For rough, from memory (it was late) tests:

My 8 Core / Vega 56 iMac Pro with 32GB of RAM, in FCPX, could run at full fps in "Quality" (full rez) - without rendering or transcoding:

2 Streams of DJI Mavic 4K with Filmconvert applied
3 Streams of Mavic without filters applied
6+ streams of ProRes 422 4K with LUTS & CC applied (with 2 of those clips having FilmConvert, which seems to be the limit as it maxes out the GPU)
6+ streams of XAVC-I 4K from a Sony FS7 w/LUTS and CC and 3 clips with FilmConvert applied
4-5 streams of XAVC-L 4K from a Sony FS5 w/LUTS and CC and 2 clips with FilmConvert applied.

My old 8 Core D700 Mac Pro could maybe (from memory) handle 1 Mavic stream and 3 ProRes streams with one film convert. All in all it's about a 2-3X speed increase over that Mac Pro.

The 8 Core base model is FAST, and this 10 core being the sweet spot for FCPX is a slight myth, except you will get 10-15% faster render times. But faster than fast...

RED RAW was even more amazing, as I was able to play (in best performance) 1 stream of 6-8K RedRAW footage with filmconvert applied on one stream in real time (!!). I didn't try two but I suspect it could handle two-three streams of 4K easily. That's stunning as I couldn't even get it to playback decent on my 15" MBP.

I was coming from a user-upgraded 2011 MacBook Pro, so the difference for me was always going to be exaggerated.

It's a beast for sure! Amazing performance. I used my GH5 on a wedding for the first time in November, and the click wheel went AWOL during preps with the ISO for a few shots, so I applied Neat Video to the files and it fixed them almost instantaneously... on my old MacBook Pro, you could forget about applying Neat video: crashes and hours of time passing by for exports... the iMac Pro treats Neat Video like a cross dissolve...

It's ridiculous the performance that we can see out of the iMac Pro. I have heard that the new MacBook Pros can do 4K multi-cams well, but I bought a machine for the future, which is why we've found such great performance - just as you have with 8K footage. I don't anticipate shooting 8K within the next 3 years, but I do expect to use this iMac Pro when I do. If anything, I'll just use an 8K camera as an unmanned camera and crop a wide shot in post to 4K.

To be honest, I'm more impressed with H264 speed than I am with files that are already in ProRes (4K or 8K). I never thought the iMac Pro would be the right choice for people using highly compressed codecs, but it looks like it's the best by quite some way for FCPX. Very happy.

 

[anticipate]

I was coming from a user-upgraded 2011 MacBook Pro, so the difference for me was always going to be exaggerated.

It's a beast for sure! Amazing performance. I used my GH5 on a wedding for the first time in November, and the click wheel went AWOL during preps with the ISO for a few shots, so I applied Neat Video to the files and it fixed them almost instantaneously... on my old MacBook Pro, you could forget about applying Neat video: crashes and hours of time passing by for exports... the iMac Pro treats Neat Video like a cross dissolve...

It's ridiculous the performance that we can see out of the iMac Pro. I have heard that the new MacBook Pros can do 4K multi-cams well, but I bought a machine for the future, which is why we've found such great performance - just as you have with 8K footage. I don't anticipate shooting 8K within the next 3 years, but I do expect to use this iMac Pro when I do. If anything, I'll just use an 8K camera as an unmanned camera and crop a wide shot in post to 4K.

To be honest, I'm more impressed with H264 speed than I am with files that are already in ProRes (4K or 8K). I never thought the iMac Pro would be the right choice for people using highly compressed codecs, but it looks like it's the best by quite some way for FCPX. Very happy.

Same, especially as I am really hammering it now during an edit. The only non-realtime playback I have had is rendering 3D titles over multiple 4K clips. And the background render on that is only a few seconds. It's super impressive!

 

[FredT2]

A bunch of 18-core machines have shown up on Geekbench. They are averaging about 3500 for single-core score and 60,000 for multi-core.

 

[jav6454]

I calculate the 18 core would be approx 20% faster than the 10. However that’s just multiplying numbers... Does anyone know how the larger L3 memory would factor into speed? Single treaded it looks like the 10 core is 5% faster than the 18 core... For arguments sake, assume cost is not a factor. But overall performance is. Uses are Final Cut X, Logic, Lightroom, some Photoshop (not much). Thoughts?

This might have been mentioned already but...
Thread & core count are not scalar. Also, they only matter in specific tasks. Lightroom is more likely to benefit. Video apps like more threads. Photo apps like more memory.

 

[SFjohn]

A bunch of 18-core machines have shown up on Geekbench. They are averaging about 3500 for single-core score and 60,000 for multi-core.

Sounds amazing! Do you have any links to those? Thanks! John

 

[FredT2]

Sounds amazing! Do you have any links to those? Thanks! John

https://browser.geekbench.com/v4/cpu/search?utf8=✓&q=iMacPro1,1

I compiled about a dozen entries of 64 GB machines, and the averages are 5505 and 60811. What I don't understand is that the 8-core single-core score is averaging about 5000 and the 10-core 5200.

 

[SecuritySteve]

https://browser.geekbench.com/v4/cpu/search?utf8=✓&q=iMacPro1,1

I compiled about a dozen entries of 64 GB machines, and the averages are 5505 and 60811. What I don't understand is that the 8-core single-core score is averaging about 5000 and the 10-core 5200.

The 10 core has a higher turbo frequency than the 8 core. That is why it is getting a higher single core score. In fact, all of the other processors (14 and 18 too) will have a higher single core score than the 8 on paper.

Edit: The 18 core machines here are all i9s - likely Hackintosh builds. Still impressive though! I'm seeing lots of 5500+

 

[sputnikBA]

Yep, a whole bunch of false “iMacPro1,1”s running on a mix of i9-7980XE, 7920X (12 core), 7960 (16 core), 7900X 6950X (both 10 core).

Always have to be wary with Geekbench results if you want an accurate estimate of the actual model you are looking for. I actually wish they had a better way of filtering out false Macs or hardware that’s also being overclocked.

 

[SFjohn]

https://browser.geekbench.com/v4/cpu/search?utf8=✓&q=iMacPro1,1

I compiled about a dozen entries of 64 GB machines, and the averages are 5505 and 60811. What I don't understand is that the 8-core single-core score is averaging about 5000 and the 10-core 5200.

Wow, I’m impressed. I didn’t expect the single core to hold up so well, also the multi core is well above my expectations. I know these are just benchmarks and actual performance will depend on the software and workflow I’m using. I’m exited.

 

[FredT2]

Edit: The 18 core machines here are all i9s - likely Hackintosh builds. Still impressive though! I'm seeing lots of 5500+

Well I certainly feel embarrassed to have missed that!

 

[ThatSandWyrm]

I could be wrong, but for LR and PS, I don't think you'll see much of a difference between 10 cores and 18 cores.

I don't think he'll see much of a difference between 8 and 18 cores in LR and PS.

 

[Trebuin]

I don't think he'll see much of a difference between 8 and 18 cores in LR and PS.

Agreed. Memory helps, not CPU there.