2020 MBA - a light weight mini-beast for mobile software development and graphic design?

[patent10021]

My 2015 MBP15 is so heavy when traveling with other things stuffed in my backpack. I do a lot development internationally so maybe this new MBA is the ticket.

New MBA has the Intel Iris Plus and the 2015 MBP has the Intel Iris Pro. Going by this data there doesn't seem to be much difference in performance. I'm thinking the general performance should be similar.

I'm getting by with my 2015 MBP 15 16GB for mobile development. Xcode plus lightweight machine learning computation. Was considering the 16" MBP but do not want to deal with that crazy weight. If I need more power I'd be using desktops or external GPUs anyway.

Any other designers and devs out there thinking the same thing?

 

[Dhonk]

I'm of the same thinking. Moving into a new home in a few months and considering making an iMac my primary device, but getting an Air for mobile use. My current computer is a Mac pro 13", mid-2014 with 3 GHz dual core i7. I can't imagine the power would be a step down given the age. Happy to keep my MBP, but thought this device today sounded a great option starting at $899. Are there sites that let us compare the benchmarks from current and past devices?

 

[patent10021]

Well, in MBP System Info and Apple site:

2015 MBP 15: Quad-Core Intel Core i7, 2.8 GHz, 6 MB L3 Cache.

2020 MBA is configurable to 1.2GHz quad-core Intel Core i7, Turbo Boost up to 3.8GHz, with 8MB L3 cache.

Pretty impressive considering the price.

I'm not sure what the difference is between the older MBP's native 2.8 GHz and the 2020 MBA's Turbo Boost up to 3.8GHz. I do know that when running Xcode simulator the fans in my MBP are often full blast. I'm sure it wiould be the same for the MBA but I wouldn't mind at that footprint and weight.

 

[troop231]

I currently have a maxed out, late 2013 15" rMBP and was wondering if this new MBA could be a viable replacement?

Really wanted 32GB RAM though since I do Autodesk work and PCB design, but I've gotten by with 16GB thus far.

I have two 27" 4K monitors connected to my rMBP and you can definitely tell it taxes the 750M with the fans kicking in; the rMBP is always in clamshell mode so I don't use the internal display.

 

[austyn23]

I’m in the same boat, using the 15 inch 2015 MBP for graphic design.

But I’ll wait for the rumored 14 inch... I think a DGPU is necessary, let’s see if the new small MBP gets one...

 

[RegularGuy09]

I will be using this stock option to do some web development.

-quad core i5
-8GB RAM
-512GB SSD

I think the i5+8GB combo would ensure my notebook can keep up with my web dev workload. I store a lot of media so 512GB works for me, and its future proofed.

 

[patent10021]

Does anyone know what the difference is between the older MBP's native 2.8 GHz and the 2020 MBA's Turbo Boost up to 3.8GHz?

 

[imp3rator]

Simple - In short peak is Air 36% better...in longer load is two times better MBP...

 

[raymanh]

Wait till there's more reviews about thermal throttling. On paper it's fast, but IRL it may not be able to sustain that speed if the cooling is passive like last year's model

 

[llllllllll]

if the cooling is passive like last year's model

The 2019 model does have a fan inside.

 

[nexu]

The 2019 model does have a fan inside.

But it doesn't help cooling anything at all

 

[llllllllll]

it doesn't help cooling anything at all

That may be, but it is "active" ?

 

[raymanh]

Yes, active but no physical connection to the CPU is what it meant

 

[intelligence]

New machine seems to idle at 70c, doesn't matter if you crank the fan to 100% or let OS decide RPM. The machine is just not meant to handle prolonged loads. 70c while idle will also be a bit uncomfortable if you plan on using this in your actual lap.

Granted, I've only seen the one review covering this yet.

 

[raymanh]

So I did a comparison between my 2016 MBP nTB with the base dual core 2.0 GHz i5 CPU (with 8GB RAM) and the new 2020 MBA with the i5 quad core using figures from this review.

Results:

Cinebench R20 (CPU test):
MBP: 710 (no thermal throttling, 91 deg C, sustained: 2.9GHz @ 19W, fan 3400rpm, but core utilization was only 66%?)
MBA: 863 (thermal throttled after 1 minute, 100 deg C, sustained only 1.7 GHz @ 10W, even with fan at 100%)

Unigine Heaven on extreme preset (graphics test):
MBP: Score of 210 with average FPS of 8.3, min FPS of 4.6, max FPS of 17
MBA: Score of 209 with average FPS of 8.3, min FPS of 4.5, max FPS of 16.5

So the MBA is slightly better in terms of CPU, but has the potential to be much better if cooling was improved, and very marginally worse in terms of graphics? I don't understand that because it is supposed to have much better G7 integrated graphics. Maybe Unigine is too old of a test?

 

[magbarn]

So I did a comparison between my 2016 MBP nTB with the base dual core 2.0 GHz i5 CPU (with 8GB RAM) and the new 2020 MBA with the i5 quad core using figures from this review.

Results:

Cinebench R20 (CPU test):
MBP: 710 (no thermal throttling, 91 deg C, sustained: 2.9GHz @ 19W, fan 3400rpm, but core utilization was only 66%?)
MBA: 863 (thermal throttled after 1 minute, 100 deg C, sustained only 1.7 GHz @ 10W, even with fan at 100%)

Unigine Heaven on extreme preset (graphics test):
MBP: Score of 210 with average FPS of 8.3, min FPS of 4.6, max FPS of 17
MBA: Score of 209 with average FPS of 8.3, min FPS of 4.5, max FPS of 16.5

So the MBA is slightly better in terms of CPU, but has the potential to be much better if cooling was improved, and very marginally worse in terms of graphics? I don't understand that because it is supposed to have much better G7 integrated graphics. Maybe Unigine is too old of a test?

No unigine still separates the weak GPU's from the strong ones in 2020. I think what you're seeing here is that the MBA has a very power/heat constrained chip that has to cut back performance as the poor CPU/GPU has to rely on a small channel of air blowing over a very thin heatsink. Sad, the CPU/GPU combination is capable of so much more if properly cooled. I'm just about ready to cancel my loaded i7/16/1TB order as I was looking at this to replace my 2016 MB 12 that throttles so fast when just looking at websites. This new 2020 MBA while quad core in higher models looks like to be another throttle fest like my 2016 MB 12 (which is quiet as a mouse so I put up with it's throttling) even with the MBA's annoying fan noise...

 

[thadoggfather]

So I did a comparison between my 2016 MBP nTB with the base dual core 2.0 GHz i5 CPU (with 8GB RAM) and the new 2020 MBA with the i5 quad core using figures from this review.

Results:

Cinebench R20 (CPU test):
MBP: 710 (no thermal throttling, 91 deg C, sustained: 2.9GHz @ 19W, fan 3400rpm, but core utilization was only 66%?)
MBA: 863 (thermal throttled after 1 minute, 100 deg C, sustained only 1.7 GHz @ 10W, even with fan at 100%)

Unigine Heaven on extreme preset (graphics test):
MBP: Score of 210 with average FPS of 8.3, min FPS of 4.6, max FPS of 17
MBA: Score of 209 with average FPS of 8.3, min FPS of 4.5, max FPS of 16.5

So the MBA is slightly better in terms of CPU, but has the potential to be much better if cooling was improved, and very marginally worse in terms of graphics? I don't understand that because it is supposed to have much better G7 integrated graphics. Maybe Unigine is too old of a test?

How can they be that close? Thought I read otherwise

 

[imp3rator]

raymanh : very good comparsion because I had MBP 2016

The i5-1030G7 should have mayebe a 20% lower performance than base MBP 2019 - but it looks like that lower TDP and worse cooling throttle GPU as well.

If I load a GPU to 100% I get a much higher temperature than If I load a CPU to 100%...
[automerge]1584804802[/automerge]

EDIT : Oh I see - you compare numbers from test Max Tech, which testing a new open Macbook, which indexing, maybe constantly loaded and throttled. He has definetly wrong numbers (as you can see in GB 5)

 

[yanksrock100]

Looking at it for a similar use case as well. My 2015 Pro has been a beast for my last four years of university and personal development work and I would love to replace it with one of these new Airs. Awaiting feedback about development from others before I pull the trigger...

 

[raymanh]

How can they be that close? Thought I read otherwise

Yeah I was confused too. I ran it multiple times on my 2016 nTB MBP and got around 710 +/- 10 each time. Interestingly I did it with fans forced to 100% and got the same score so the MBP definitely isn't thermal throttled.

So yeah I don't understand how my MBP and the new MBA are close in Cinebench R20, yet Geekbench scores are completely different.

And the graphics result was a complete shock to me. I did not expect my base spec 4 year old dual core MBP to beat a 4 core 10th gen 2020 MBA. Guess it shows how important cooling is.

 

[PeterJP]

Yes, active but no physical connection to the CPU is what it meant

Precisely. It seems to be a passively cooled machine (so no CPU fan) that happens to have a case fan to extract hot air.

 

[magbarn]

Yeah I was confused too. I ran it multiple times on my 2016 nTB MBP and got around 710 +/- 10 each time. Interestingly I did it with fans forced to 100% and got the same score so the MBP definitely isn't thermal throttled.

So yeah I don't understand how my MBP and the new MBA are close in Cinebench R20, yet Geekbench scores are completely different.

And the graphics result was a complete shock to me. I did not expect my base spec 4 year old dual core MBP to beat a 4 core 10th gen 2020 MBA. Guess it shows how important cooling is.

Just shows you how really useless Geekbench is for sustained workloads. It makes me cringe when the latest device comes out and everyone goes drooling over the latest high GB score and claims it as gospel of superior performance. Unfortunately as Apple didn’t want to redesign the chassis we’re stuck with a mini passive cooler on a chip that is really capable of so much more. It’s really to save costs and to also differentiate between the MBP/MBA lines...

 

[patent10021]

No unigine still separates the weak GPU's from the strong ones in 2020. I think what you're seeing here is that the MBA has a very power/heat constrained chip that has to cut back performance as the poor CPU/GPU has to rely on a small channel of air blowing over a very thin heatsink. Sad, the CPU/GPU combination is capable of so much more if properly cooled. I'm just about ready to cancel my loaded i7/16/1TB order as I was looking at this to replace my 2016 MB 12 that throttles so fast when just looking at websites. This new 2020 MBA while quad core in higher models looks like to be another throttle fest like my 2016 MB 12 (which is quiet as a mouse so I put up with it's throttling) even with the MBA's annoying fan noise...

So what you're saying then is your waiting for the 2020 14" MBP like I have decided to do ?