2019 i9 iMac - Geekbench scores

[mmackinven]

Hey guys,

Just wanted to say that this model iMac is a beast! Here's my config:

3.6GHz 8-core i9
8GB of 2666MHz DDR4 memory
Radeon Pro 580X with 8GB of GDDR5 memory
1TB of SSD storage

Here's the stock standard benchmark with 8gb RAM: https://browser.geekbench.com/v4/cpu/14262673
Removed the 8GB and added 32GB: https://browser.geekbench.com/v4/cpu/14263188
Added the 8GB back in, total 40GB: https://browser.geekbench.com/v4/cpu/14263250

Now this is interesting, the single core score was higher with only 32gb, any reason for this??

Now compare that to our iMac Pro at work, with 32gb RAM: https://browser.geekbench.com/v4/cpu/13212687

Have to say that this is the best mac I've owned to date, just flies through everything. And now that the new Lightroom has GPU acceleration, even better! However if I'd known Adobe was going to do that then I would have upgraded to the Vega GPU, but I don;t think it's necessary still.

Should I buy another 32gb RAM and go 64gb??

 

[casperes1996]

Now this is interesting, the single core score was higher with only 32gb, any reason for this??

Yeah; If the right slots are populated anyway, two sticks is enough to run in dual channel. Adding more on top of that, like two DIMMs per channel, may spread the information such that some of the simultaneous access is lost, but intelligent memory management placement means this isn't generally a concern unless you use the extra memory at which stage it's better to have it than not to. That's one possible contributing factor, but likely not a major influence but worth mentioning.
A more likely reason, perhaps mixed with a bit of the aforementioned, is that the configuration and extra memory also puts more pressure on the memory controller of the chip, perhaps causing it to take a cycle extra for an operation here or there.
Thirdly it could also just be natural run-to-run variance. Doesn't seem a very big difference.