All Posts by Date or last 15, 30, 90 or 180 days.
also by Lloyd: diglloyd.com photography and WindInMyFace.com

Thank you for buying via links and ads on this site,
which earn me advertising fees or commissions.
As an Amazon Associate I earn from qualifying purchases.

Other World Computing...
B&H Photo...
Amazon
As an Amazon Associate I earn from qualifying purchases.
Upgrade the memory of your 2020 iMac up to 128GB
Upgrade the memory of your 2020 iMac up to 128GB

2019 Mac Pro: the 8/12/16 Core CPUs Can Use 1TB Memory (24/28 Cores Not required)

Consult with Lloyd to determine the ideal setup for your particular workflow including CPU cores, GPU, memory, SSD, internal and external storage, backup and fault tolerance, color management, whether for video or still photography, including what is best for Photoshop, Lightroom, Capture One, focus stacking, panoramas, etc.

Read 2019 Mac Pro: Tips for Upgrading Memory as background for this post.

Apple limits the memory upgrade options for the 8/12/16 core CPUs of the 2019 Mac Pro to 768GB as 6 X 128GB or 12 X 64GB modules.

But OWC has an interesting find in OWC Confirms 1TB of Memory in the 8, 12, and 16-Core Mac Pro! as follows:

If you were hoping to hit the 1TB level of RAM with the new Mac Pro, you were likely disappointed to see on Apple’s Mac Pro page that you couldn’t purchase the 8-core base model. Instead, you needed to upgrade three levels to the 24-core machine, costing you an additional $6,000 – twice the base price of the computer itself!

But not all is lost! We have tested and verified that with OWC memory, the 8-core Mac Pro supports up to 1TB. This is excellent news because Apple doesn’t offer that particular configuration with their memory modules. The highest they offer is 768GB, and the next jump from there is to 1.5TB for an additional $15,000. Add in the $6,000 for the necessary upgrade to the 24-core processor, and you are spending $21,000 just to hit the 1TB level of RAM with Apple.

So, what about OWC memory?

1TB of OWC memory is under $14,000. Apple’s factory-installed 1.5TB is $25,000, so you’ll save yourself another $11,000 if 1TB of RAM is enough for you. And don’t forget the $6K you don’t need to shell out for the processor upgrade. That’s a total of $17,000 in savings if you don’t need 24-cores and 1TB of memory is enough for you. Not too bad!

Plan ahead—save a ton of money via buy your Mac Pro with just 32GB, then buy 128GB LRDIMM modules to upgrade (if you buy 64GB modules to start, or RDIMM modules, you’ll have to remove those to go to 768GB or 1TB).

OWC 12-module memory kit for 2019 Mac Pro

“Enough is enough” when it comes to memory

The Mac Pro memory performs best in 6 channel mode, e.g., 6 modules in 6 corresponding slots, or 12 modules provide peak memory bandwidth. However, if memory runs short, more memory is always faster.

So what happens with eight modules (8 X 128GB) needed for 1TB memory? Either:

  • Memory bandwidth will be 6-channel speed for 768GB and 2-channel speed for 256GB; OR
  • Memory bandwidth will be 4-channel speed for all 1TB.

What macOS allows for the two configurations above, or whether it optimizes for memory usage of the fastest memory first is unclear.

What is clear in past testing is that memory bandwidth differences often don’t matter in any measurable way (2 or 3% at most). OTOH, if memory is insufficient, then additional memory provides massive benefits far exceeding any bandwidth limitations.

So the situation here is clear: if 768GB is not enough, then going to 1TB is going to be a big win. OTOH, if 768GB is ample, going to 1TB without good cause will reduce memory bandwidth and might (or might not) affect peak performance. So it’s a win: start with 768GB and if observation shows that more memory is needed, go to 1TB.

 

View all handpicked deals...

Nanuk 905 Hard Utility Case with Foam Insert (Black)
$64 $48
SAVE $16

diglloyd.com | Terms of Use | PRIVACY POLICY
Contact | About Lloyd Chambers | Consulting | Photo Tours
Mailing Lists | RSS Feeds | Twitter
Copyright © 2020 diglloyd Inc, all rights reserved.
Display info: __RETINA_INFO_STATUS__