Making sense of differing speeds on multiple macs

[bourbonbelly]

Hey all,

Recently started folding on an old Mac (so CPU only), then figured why not do so on a slightly newer Mac, then finally on one that's about a year old. They break down as such.

1. Mid-2011 Mac mini, 2.3GHz Intel Core i5, 8GB RAM
2. Late-2015, (6th Gen Skylake) 3.3GHz Intel Quad Core i5, 24GB RAM.
3. 2020 MacBook Pro, (10th Gen Ice Lake) 2.3GHz Intel Quad Core i7, 32MB RAM.

While I don't have average times figured out, I'd say I'd rank their speed at completing work units and rough times as:

1.iMac (6-8 hours)
2. MacBook Pro (9+ hours)
3. Mac Mini (14+ hours)

The only real surprise is the iMac vs the MacBook Pro as we're looking at the iMac being 4 generations behind + being i5 vs i7 at that. May be a stupid question, but does the difference of 1.1GHz despite the differing generations and processors make all the difference here, leading the iMac to "beat" the MacBook Pro?

Thanks!

[Joe_H]

Welcome to the folding support forum.

First, when comparing speeds of process WUs, you need to be comparing WUs from the same project. Different projects will take more or less time to complete on the same systems.

Even though the iMac has a several generation older processor, it has the advantage of a better cooling system compared to the MacBook Pro. So it will run at full clock speed more often. Your MacBook may run at some throttled speed to stay within the thermal limits set for its processor. The additional CPU threads gained using HT in the i7 first add to the heat load, and second only improve processing by about 10-20%. The same i7 processor in a different system with better cooling might show as faster.

[bourbonbelly]

Thanks for the reply - the cooling aspect makes sense and hadn’t really factored that in.

One follow up in that regard is that the MBP’s fans are really kicking in beyond a very very low, barely audible level, and heat seems to be well within limits based on monitoring software. Wondering if it’s possible that it could be throttling down unnecessarily, as it’s like a jet engine when rendering video etc, and if so there’s a way around it.

[JimboPalmer]

I am pretending Apple used the latest CPUs in each

2011 Mac mini looks to be
https://ark.intel.com/content/www/us/en ... tIds=52212
4 threads of AVX

2015 iMac
https://ark.intel.com/content/www/us/en ... tIds=88188
4 threads of AVX2

2020 Mac Book Pro
https://ark.intel.com/content/www/us/en ... Ids=196594
4 threads of Avx2

The AVX2 should be 1.6 times the speed of the AVX

[bruce]

Thermal throttling is a factor on more than just Macs. Most notably, laptops often run at reduced speed because of the speed they could run if the processing were not continuous. If it runs like typical laptops are used, I's not uncommon to buy a laptops that is clocked higher than necessary. If the processor runs faster than a "reasonable" rating for a couple of seconds after you move the mouse or type something, that's good ... it soon runs out of work until you move something else. (The thermal capacity keeps it from overheating except when given something that runs a long time.) Short bursts of fast work is better that limiting all processing to what it can do continuously.