Upgrade to MBP 13" 2020 8th Gen i5 - is it a slower CPU than 2018 MBP 15" i7?

[unoporfavor]

Hi guys,

I was about to pull the trigger on the base model 2020 MacBook Pro 13" and need some advice!
This is to replace my 2018 MBP 15" i7 2.2Ghz.

Requirements:
13" - non-negotiable (please don't suggest the 16" )
Non-Butterfly keyboard non-negotiable (please don't suggest an earlier refurb )
Not be slower than my 2018 MBP 15" i7 (I already have a 2020 MacBook Air which is fantastic machine but also doesn't meet use case)
16GB RAM (I occasionally run out of memory on 8GB)
--

My question is:

Will the 2020 MBP 13" with 8th gen CPU (base model i5) be faster than a 2018 MBP i7? ( I believe it's called Coffee Lake?)
Or will I need to go to the 2020 MBP 13" with 10th Gen CPU to see a noticeable speed increase?
The portability and size is key, along with the new keyboard, so using the MBP 15" is not an option. However, if I can save some money not going to the 10 Gen on the 13" MBP that would be great!

Thanks very muchly ✌ ☺

 

[sdedalus1]

No, the 8th gen 13" MBP will be slower all around than your 15" i7. The 10th gen 13" chips can beat the 15" in short, bursty single core loads, and the 10th gen i7 might get close to it in bursty multicore work. But no quadcore 13" MBP can beat a more powerful hexacore 15" with heavy workloads that take advantage of all six chips.

single core / multicore Geekbench
i5-8257U (13” MBP 2020, quadcore 8th gen 15W) = 925 / 3820
i5-1038NG7 (13” MBP 2020, quadcore 10th gen 28W) = 1145 / 4246
i7-1068NG7 (13” MBP 2020, quadcore 10th gen 28W) = 1178 / 4394
i7-8750H (15” MBP 2018, hexacore 8th gen 45W) = 962 / 4785

multicore Cinebench R20
i5-8257U = ~1590
i5-1038NG7 = ~1900
i7-1068NG7 = ~1900
i7-8750H = 2500

 

[Sterkenburg]

As said above, for multi-core workloads there is no way to technically match the performance of a hexa-core 15" with the 13". However, if you get a 10th gen the difference would be relatively limited. Based on your use case, would a ~10% slowdown in multi-core performance be something you'd notice?

 

[unoporfavor]

No, the 8th gen 13" MBP will be slower all around than your 15" i7. The 10th gen 13" chips can beat the 15" in short, bursty single core loads, and the 10th gen i7 might get close to it in bursty multicore work. But no quadcore 13" MBP can beat a more powerful hexacore 15" with heavy workloads that take advantage of all six chips.

single core / multicore Geekbench
i5-8257U (13” MBP 2020, quadcore 8th gen 15W) = 925 / 3820
i5-1038NG7 (13” MBP 2020, quadcore 10th gen 28W) = 1145 / 4246
i7-1068NG7 (13” MBP 2020, quadcore 10th gen 28W) = 1178 / 4394
i7-8750H (15” MBP 2018, hexacore 8th gen 45W) = 962 / 4785

multicore Cinebench R20
i5-8257U = ~1590
i5-1038NG7 = ~1900
i7-1068NG7 = ~1900
i7-8750H = 2500

Thanks mate that's absolutely what I needed to see! I couldn't figure out what the Intel generations really meant materially.
I actually find the 2018 MBP is more than fast enough for the exporting and spreadsheets. I am slightly disappointed in the single-core speed of it in some tasks, so I certainly wouldn't want to go slower.
I can deal with lower multi-core as usually that's for longer tasks that I accept. Just more concerned if it goes from 3 minutes to 7 minutes.

Thanks again - it looks like I'm going to have to go 10th Gen in the 13" But I will stick to the i5 as so little gains with the i7. Damn was looking to free up some capital too! It's about a direct swap price-wise if I sell the 15" and buy a new 13" so guess it's not tooooo bad.
Cheers ✌
[automerge]1589339057[/automerge]

As said above, for multi-core workloads there is no way to technically match the performance of a hexa-core 15" with the 13". However, if you get a 10th gen the difference would be relatively limited. Based on your use case, would a ~10% slowdown in multi-core performance be something you'd notice?

It looks like it it might be acceptable. I think it might have to be a try and see type deal ! I am keen to get to the smaller form-factor though so if it takes an extra minute that's an acceptable trade-off.
If it was going to take 250% longer (like the air sometimes does), then I might have to re-think.
Looking at the benchmarks, it does look like I will have to go to the 10th Gen to get similar performance.

DAMMIT !

 

[sdedalus1]

And don’t forget the discrete GPU on the 15” is 2x better than the integrated GPUs on the 13”. Not that it matters for Excel . . .

 

[unoporfavor]

And don’t forget the discrete GPU on the 15” is 2x better than the integrated GPUs on the 13”. Not that it matters for Excel . . .

Good call - I have thought about that, but I typically don't do a lot of video editing - less than weekly. So I thought that office work and photo exporting the GPU should be good enough - especially since they can now drive those hi-res external monitors.
Thanks for your points ✌