Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I suspect that the only difference, compared to current 13"MBPro specs, are that the 2020 MBPro model will have:
1. New keyboard
2. New CPU/GPU combo
3. Possible new screen
4. Approx $200 price reduction due to base SSD configs being 256/512 rather than 128/256.

So...given that, the new MBPro is still going to be a sizeable increase in cost compared to a MBAir, essentially just to get a better fan in order to get better sustained CPU performance.

I'm starting to think the i5/16GB/1TB MBAir is the way forward, and then replace/hand down sooner than I would otherwise do with the MBPro if it starts to feel "slow" in a few year's time.

It is mainly the touchbar that I really don't like on the MBPros.

but it’s not just the fan...doesn’t the MBP get better processors as well?
 
but it’s not just the fan...doesn’t the MBP get better processors as well?
Yes, the MBP should be getting CPU's that can draw at least 50% more power or more. The MBA has a very power limited CPU so that even when the CPU is kept cold, it still can't maintain high boost due to power constraints. The MBP CPU should not have as much limitation.
 
  • Like
Reactions: throAU
Yes, the MBP should be getting CPU's that can draw at least 50% more power or more. The MBA has a very power limited CPU so that even when the CPU is kept cold, it still can't maintain high boost due to power constraints. The MBP CPU should not have as much limitation.

... and if this is something which matters to a person, then they probably should be purchasing a MBP not an MBA
 
but it’s not just the fan...doesn’t the MBP get better processors as well?
Yes, the MBP should be getting CPU's that can draw at least 50% more power or more. The MBA has a very power limited CPU so that even when the CPU is kept cold, it still can't maintain high boost due to power constraints. The MBP CPU should not have as much limitation.

But that's not really true. All the current CPUs only get max performance at well above of their rated TDP wattage.

Essentially, a CPU in a MBPro only performs better because it has "guaranteed" cooling to maintain at least 25 Watts of power (at 2.4 Ghz) compared to a "guaranteed" 10W of power on the MBAir CPU (at 1.1 Ghz). If the MBAir CPU had a better cooler (and it was enabled to stay boosted rather than arbitrarily dropping to lower frequency to keep wattage low) the performance would be identical.

Equally, if we gave the MBPro the same cooling solution as the MBAir, the sustained performance of the MBPro CPU would drop right down and be similar to the MBAir (or maybe slightly worse if the CPU is a less efficient model). So ultimately, the major difference is the cooling.

It's like the misconception that putting a 65Watt CPU in the current Mac mini makes it inherently more powerful than if it had a "mobile" 45W CPU. This is false. Given the same cooling system, both would perform more or less identically. Indeed the Mac mini would likely perform better had it been given the same "65Watt" cooling solution that it has, but the "45Watt" processors shipped in the current 16" MBPro.
 
Last edited:
You need the power headroom and the ability to cool the processor to keep boost going longer.
We've already had youtuber keep the CPU well below 100C with liquid cooling and the cpu doesn't keep boost despite being cooled much more than the stock solution. You're being limited by power more than heat. Although heat does drop the performance by about 10%.
Here's an i7 Windows laptop with the exact same i7 CPU as the 2020 MBA but it's the 15 watt version
Screen Shot 2020-04-30 at 1.03.38 PM.png

Here's my 2020 MBA i7 when cooled and then hot. About 10% difference only.
The Acer has much more power to play with
Screen Shot 2020-04-30 at 1.10.45 PM.png

All of these are for Cinebench R20 multicore
 
You need the power headroom and the ability to cool the processor to keep boost going longer.
We've already had youtuber keep the CPU well below 100C with liquid cooling and the cpu doesn't keep boost despite being cooled much more than the stock solution. You're being limited by power more than heat. Although heat does drop the performance by about 10%.
Here's an i7 Windows laptop with the exact same i7 CPU as the 2020 MBA but it's the 15 watt version
View attachment 910944
Here's my 2020 MBA i7 when cooled and then hot. About 10% difference only.
The Acer has much more power to play with
View attachment 910945
All of these are for Cinebench R20 multicore
Yes, but I am pretty sure this is due to a parameter that is baked in by the motherboard designer (in this case Apple). Have a read of the Anandtech article:


It makes it clear that whilst Intel provide guidelines for how long boost times should last (at some arbitrary max wattage), the vendor can decide the maximum power delivery (and for how long). With the obvious necessity of equivalently capable cooling.

Based on the video, it looks like the CPU implementation in the MBAir is hard coded to have a limited boost time regardless of temp. But this has little to do with it being a "10W" CPU.
 
Last edited:
Here's an i7 Windows laptop with the exact same i7 CPU as the 2020 MBA but it's the 15 watt version
View attachment 910944
Here's my 2020 MBA i7 when cooled and then hot. About 10% difference only.
The Acer has much more power to play with
View attachment 910945
All of these are for Cinebench R20 multicore

I'm no expert at such things, but I'm not sure I'd use the term "exact same i7 CPU" when one is an i7-1065G7 and the other an i7-1060NG7?

From https://www.cpu-monkey.com/en/compare_cpu-intel_core_i7_1060ng7-1327-vs-intel_core_i7_1065g7-933

1588278820392.png

1588278967099.png
 
I'm no expert at such things, but I'm not sure I'd use the term "exact same i7 CPU" when one is an i7-1065G7 and the other an i7-1060NG7?

From https://www.cpu-monkey.com/en/compare_cpu-intel_core_i7_1060ng7-1327-vs-intel_core_i7_1065g7-933

View attachment 910949
View attachment 910952
They probably are identical silicon designs. Just with different power and frequency limits, then given a different name, and sold at a different price. It is a way to make money when yields vary, and when customers (and therefore vendors) like segmentation.
 
They probably are identical silicon designs. Just with different power and frequency limits, then given a different name, and sold at a different price. It is a way to make money when yields vary, and when customers (and therefore vendors) like segmentation.

You are likely 100% correct. I suspect some take that and then believe they should perform the same - when they won’t.
 
Yes, but I am pretty sure this is due to a parameter that is baked in by the motherboard designer (in this case Apple). Have a read of the Anandtech article:


It makes it clear that whilst Intel provide guidelines for how long boost times should last (at some arbitrary max wattage), the vendor can decide the maximum power delivery (and for how long). With the obvious necessity of equivalently capable cooling.

Based on the video, it looks like the CPU implementation in the MBAir is hard coded to have a limited boost time regardless of temp. But this has little to do with it being a "10W" CPU.
Yes, but the point is that chips are binned based on the TDPs provided. Apple lets the Pros run at up to 25W, which is the upward cTDP of the 15W chips. The 10W chips have an upward cTDP of only 12W. So even if Apple put in the cooler for the Mac Pro, Intel wouldn’t certify these to run at above 12W for any sustained time. No surprise, then, that the 2019 13” MacBook Pro is about twice as fast as the 2020 MacBook Air for sustained CPU tasks.
[automerge]1588284464[/automerge]
They probably are identical silicon designs. Just with different power and frequency limits, then given a different name, and sold at a different price. It is a way to make money when yields vary, and when customers (and therefore vendors) like segmentation.
That’s the whole concept of binning. Intel won’t certify that the lower TDP chips can operate at the higher frequencies for extended periods of time. So the “10W” 1060G7 maxes out at 12W while the 15W 1065G7 maxes out at 25W. They do additional testing, or prepare the various wattage chips using different processes to maximize yields. That’s also how they distinguish between the i5 and i7 within a product line, as well. The two are “identical” but the i7 is certified to a higher frequency.
 
Last edited:
I think the main goal of this thread was to not get into details such as wattage or other stuff. I'm not saying we shouldn't debate such issues but in this thread, we are expecting user reviews. There are tons of things that affect cpu performence, it is not just the wattage or benchmarks. Mine should arrive soon. I will do some logic pro and coding, will let you know about my experience. I'm also considering the base model pro if I'm not satisfied with it. Note that single cpu performance of the air is very good, but nobody talks about it.
 
I think the main goal of this thread was to not get into details such as wattage or other stuff. I'm not saying we shouldn't debate such issues but in this thread, we are expecting user reviews. There are tons of things that affect cpu performence, it is not just the wattage or benchmarks. Mine should arrive soon. I will do some logic pro and coding, will let you know about my experience. I'm also considering the base model pro if I'm not satisfied with it. Note that single cpu performance of the air is very good, but nobody talks about it.
Apologies - I take your point completely - this thread got rather derailed!
It is certainly very useful to read USER reviews!👍
 
I had a loud fan spinning moment today when doing a zoom call (1x1) + Safari (Slack, Google Docs, Okta, Concur, Gmail, Calendar .. probably a couple of more web apps).

It lasted for about 5-10 minutes, and subsided even though the call was still going on (and all the same apps were still open), but I can see how this sort of machine behavior may make people weary.

Coworkers have been having Zoom issues on Macs where it overheats and shuts itself down (all sorts, not just MBA), so I will guess it may have something to do with Zoom and/or Slack (which is also known to be a resource hog), and I honestly don't know what caused it or what stopped it..

Still do not think it is going to make the machine "under powered" for my use, but just wanted to come clean that I've now had this happen on my MBA 2020, i5/16 with latest OS, as I had said in previous posts that I had not had it happen to me yet (which, of course, was true then).

FWIW though, it does happen on my work issued 2018 MBP 13 i7/16 also under similar use.
 
I had a loud fan spinning moment today when doing a zoom call (1x1) + Safari (Slack, Google Docs, Okta, Concur, Gmail, Calendar .. probably a couple of more web apps).

It lasted for about 5-10 minutes, and subsided even though the call was still going on (and all the same apps were still open), but I can see how this sort of machine behavior may make people weary.

Coworkers have been having Zoom issues on Macs where it overheats and shuts itself down (all sorts, not just MBA), so I will guess it may have something to do with Zoom and/or Slack (which is also known to be a resource hog), and I honestly don't know what caused it or what stopped it..

Still do not think it is going to make the machine "under powered" for my use, but just wanted to come clean that I've now had this happen on my MBA 2020, i5/16 with latest OS, as I had said in previous posts that I had not had it happen to me yet (which, of course, was true then).

FWIW though, it does happen on my work issued 2018 MBP 13 i7/16 also under similar use.

My 2019 MacBook Pro did the same with zoom. Zoom is horrible. I will only use on iPad now. My company won’t let us install zoom on our work PCs.
 
  • Like
Reactions: ChinaRye and throAU
FWIW though, it does happen on my work issued 2018 MBP 13 i7/16 also under similar use.

Yup. A lot of the problem is software, and especially video conferencing/editing/playing software using codecs apple do not support.

I think in general, irrespective of what apple laptop you buy, you want to be trying to limit your use of third party, poorly optimised and/or non-system codec using software as much as possible. This sort of thing impacts all Macs, but if you're using one that is both thermally and power constrained to the degree the MacBook Air is, it will be more noticeable.

It's not just video software though either. I just found a bug in VMware workstation (maybe mentioned earlier in thread) that results in massive excessive CPU consumption due to a bug between VMware's video driver and metal. Disable VMware Fusion's metal support via a config file edit for it and performance is night and day better with much less CPU use, noise, better performance, etc. I suspect Parallels is a lot better in this respect.

TLDR: in a small low power device like this, software choice can make a HUGE difference because resources are limited - there's less to throw away wastefully.

If you aren't willing to adjust your software use to suit the hardware, a Pro (or even better, a desktop Mac) is probably a better choice. But even with that, you'll still pay the price in reduced (vs. potential) performance, worse battery life, more heat, etc. It will just handle the load better because there's more power available.

If you ARE willing to restrict your use to well behaved software, the 2020 Air is mostly silent, gets great battery life, etc.

If you make those compromises though you'll improve the battery life, noise and performance on any other Mac as well.
 
My 2019 MacBook Pro did the same with zoom. Zoom is horrible. I will only use on iPad now. My company won’t let us install zoom on our work PCs.
Yup, unfortunately my work currently requires the use of Zoom, although I think we are also exploring google hang outs(?) and other video-conferencing software.. which may or may not be better.

So, in the meantime, my workaround has been to use a fan-less machine, specifically an iPad like you, where/when possible pursuant to corporate security policies etc.

I guess I could go in a "look for the highest spec" trip, but honestly .. can't control software so not going to try.
 
I'm curious if people think 256 gb is enough storage? I figure with an icloud folder in Finder and also my dropbox folder in Finder, 256 should be more than enough considering I won't be installing any games or doing any movie editing projects. I also don't store photos locally, those are all in dropbox. So really 256 will just be used for system files, programs, updates, I can't think of anything else I'd need to store locally.
 
I'm curious if people think 256 gb is enough storage? I figure with an icloud folder in Finder and also my dropbox folder in Finder, 256 should be more than enough considering I won't be installing any games or doing any movie editing projects. I also don't store photos locally, those are all in dropbox. So really 256 will just be used for system files, programs, updates, I can't think of anything else I'd need to store locally.

keep in mind that 256GB is base 10, it really corresponds to a total storage of aprox 238GB.
waiting for the MBP... is the amount of storage that I will choose, there is no money for more, I prefer to invest in RAM
 
  • Like
Reactions: ChinaRye
I'm curious if people think 256 gb is enough storage?

You're asking if a two foot piece of string is long enough.

Answer depends entirely on what you need to do with it.

From what you describe that seems like it'd be fine, but the real thing to do is look at your storage needs on your current computer. How much is in use, what's the capacity, and how much is free.
[automerge]1588332363[/automerge]
If you ARE willing to restrict your use to well behaved software, the 2020 Air is mostly silent, gets great battery life, etc.

If you make those compromises though you'll improve the battery life, noise and performance on any other Mac as well.

I propose that that eschewing crappy software isn't making a compromise as much as it's having standards. Granted we sometimes have no choice if the software is dictated to us for some reason, but in general I'd rather use well-behaved efficient software wherever possible.

As to performance / noise -- I spent yesterday evening starting to go through 70GB of RAW format travel photos from last year tagging/ranking/categorizing using PhotoMechanic6, little bit of image tinkering in Luminar4, interspersed with bits of Mail and Safari and Messages. My MBA behaved superbly, cool on my lap, snappy & quick, completely silent, with projected battery life of 9+ hours.
 
Last edited:
We stopped using Zoom for work, but it absolutely roasts any laptop I have seen it run on, including high end Windows 10 workstations. It's not a great testing platform for performance...definitely built and optimized for mobile OS.
 
I'm curious if people think 256 gb is enough storage? I figure with an icloud folder in Finder and also my dropbox folder in Finder, 256 should be more than enough considering I won't be installing any games or doing any movie editing projects. I also don't store photos locally, those are all in dropbox. So really 256 will just be used for system files, programs, updates, I can't think of anything else I'd need to store locally.

Depends what you do. 256 isn't a lot these days, if you work with video at all a few tens of gb here or there for some uncompressed input fills up the disk quick. the step up to 512 or 1tb isn't massive compared to the cost of the entire machine.

Personally I think in 2020 512 gb should be the base storage for any computer you want to do more than browse the internet or run email/basic web apps. You might only need say 200-300 (which I think is reasonable for a typical non-power user to be comfortable) or so but it gives you some working space to say, back up your phone, copy a bunch of data or shuffle things around if you need to. You also need free space for local time machine backups to work. These are useful if you want to roll back to old copies of files in between time machine backups to external storage.

Needing 200 (which isn't unreasonable for light use with a number of apps and some media) is tight on a 256 gig machine. 300 is impossible.

I went for 1TB for this reason, wasting time shuffling data around because you're short on space kinda bites into a lot of real world time and means that your fast storage just isn't so fast any more when you take that time into account.

Instead of looking at the cost of the storage upgrade (and thinking "damn that's expensive for XX GB!"), look at the difference in cost of the entire machine either with or without it. You're talking what... 10% cost increase to make the machine far less painful to deal with due to storage space anxiety or constant data management. Will 10% more money make the machine >10% nicer to use, for a significant portion of the time you own it? That's the question you should ask yourself IMHO. For me, bumping all the way to 1TB for my usage was a no brainer.
 
But if you only need this storage for your media you can get an 2TB external SSD for 250$ with almost the same speeds. Getting 1TB of internal storage costs you 500$ so its up to you. External is just not that convenient for daily usage, but not everyone is a video editor and needs all of his media library instantly available. I think 256 is perfectly fine for programmers, and students.
 
  • Like
Reactions: adrianlondon
It's not just zoom. I had few zoom and few google meet VCs in last few days. For zoom, cpu was averaging around 80-85 degrees where as for Google meet it was around 96-97 degrees and fan was on full blast. FaceTime works without any issues though. As someone said, its could be because of apple not supporting codecs.
 
  • Like
Reactions: larryleveen
We stopped using Zoom for work, but it absolutely roasts any laptop I have seen it run on, including high end Windows 10 workstations. It's not a great testing platform for performance...definitely built and optimized for mobile OS.
Not true in my case, an old XPS handles it just fine at 60*C. Maybe after 3 hours getting to 70-80*C. Short of people using screenshots, I've resigned to respect but not trust opinions - nor should you trust mine as I did not provide screenshots because I'm not currently in a zoom call. There is one simple trick that would help all of you in your goal of lower temps and increasing performence: undervolting. My machine has a -100mV on the CPU and GPU. Same performance, 20*C lower temps.

At this point I'm not even sure I want a MBP (let alone a MBA) if I can't undervolt it, knowing that I could easily drop temps by 20*C with it.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.