15" M2 Air or 14" MBP M2 Pro (Refurbished) for Computer Science?

[Zest28]

Btw, Xbox Series X ”instant loading” capability only works with their fast SSD’s.

If you use a slow external SSD on the USB 3.1 port, it doesn’t work and Microsoft actually forbids Xbox Series X games to be installed on external slow SSD’s for this reason.

But since MacRumors says a slow SSD is as fast as a fast SSD, this shouldn’t be the case right?

For next-gen open world games, you will also start to see that loading screens will also disappear and everything will be directly streamed from the SSD. And this is only possible with very fast SSD’s.

 

[Mity]

You seem to be trying to apply the decently reasonable uproar about the M2 MacBook Air SSD speeds, to the M2 Pro speeds, which when you look at the actual numbers is just silly. The slowest M2 Pro SSD is faster than the M1 SSD ever was.

That's not what I'm saying at all. This is the gist of what I'm saying:

Based on the above, my opinion is that it the base M1 gen machines are a better buy than the base M2 gen machines (And keep in mind that I haven't even entered the prices for the 15" M2 MBA.). I came to this conclusion based on the lower prices of the M1 gen machines with modest improvements in CPU performance for M2 gen CPUs and better SSDs in the base M1 gen machines.

It's true that you will get a slightly better processor with the M2 gen machines. The downside is that you will get a worse SSD with the base M2 gen machines.

I accept the opinions (and even facts) that @Rnd-chars and @kschendel have stated that the SSD will not play an important part in coding. I would reply that the CPU improvements from M1 gen to M2 gen may not be drastically large, as mentioned here. And since RAM is more important for Python coding, I would up the RAM or up the SSD if I know I'm going to utilize swap. I use about 7-8GB of RAM on my 2019 16" MBP even at idle. That means if someone gets a 16GB RAM machine, they only have 8GB left for actual coding.

-----
Note:
The SSD read/write speeds are not mine. They were collected from the sources below and I don't know if there were differences in methods when testing.

https://www.theverge.com/23220299/apple-macbook-air-m2-slow-ssd-read-write-speeds-testing-benchmark
youtube.com/watch?v=j6F_4za7dJg&t=473s
youtube.com/watch?v=kaGGyXzBG2M&t=390s

 

[BigMcGuire]

Man, you are the one that is spreading misinformation.

A 1400 MB/s SSD is not as fast as a 7000MB/s SSD. You can’t argue against facts.

It's amazing to see the variety here - talking my personal experience only:

I recently talked to someone who had no clue that they were using their external WD Essentials drive for gaming vs their internal gaming PC SSD (NVME) drives on their gaming PC.

Me? I can tell the difference between a 500mb/sec Sata3 SSD and a T7 Shield, and an NVME 3-4k mb/sec. Makes a MASSIVE difference when compiling a large Visual Studio project or dealing with large databases with lots of small files.

Most people probably won't notice the difference, that's why these companies think they can get away with it because those of us who want/need the speed will usually shell out $ for the high end stuff. I very much appreciate the speed my 16' MBP SSD has. I've had this laptop for over a year and have written 56.4TB to it. lol.

Profit is king.

 

[Zest28]

It's amazing to see the variety here - talking my personal experience only:

I recently talked to someone who had no clue that they were using their external WD Essentials drive for gaming vs their internal gaming PC SSD (NVME) drives on their gaming PC.

Me? I can tell the difference between a 500mb/sec Sata3 SSD and a T7 Shield, and an NVME 3-4k mb/sec. Makes a MASSIVE difference when compiling a large Visual Studio project or dealing with large databases with lots of small files.

Most people probably won't notice the difference, that's why these companies think they can get away with it because those of us who want/need the speed will usually shell out $ for the high end stuff. I very much appreciate the speed my 16' MBP SSD has. I've had this laptop for over a year and have written 56.4TB to it. lol.

Profit is king.

But the thing is, 7000MB/s is not “high-end”, this is just normal speeds. It‘s just $150 for a 2TB SSD that has these speeds.

Even my $499 Xbox Series X can do 5000MB/s.

I can‘t wait for the day when Apple will put slow SSD’s in the base M3 Ultra Mac Pro.

 

[Corefile]

Here's a comparison done by some Max tech experts at Max Tech.

 

[dmccloud]

OP asked for opinions. I gave mine and provided justifications, which included professional reviews. Some people disagreed. No big deal. Their opinions will also help the OP choose according to his/her desire.

If you want to now focus on the medium to diminish the value of my justifications and not the actual content, that's a convo I don't want to have. But I am happy to discuss why anyone should pay more to get the same performance.

You "justifications" completely misrepresented the facts at play here. Apple switched to 256GB NAND modules when the M2 Airs were released, so the base M2 SoC with 256GB is running 1 256GB NAND module instead of 2 128GB NAND modules like the base M1 had. That's the only one where speeds are affected. Any configuration of 512GB or more uses two NAND modules, which brings throughput back up. Trying to extrapolate the base M2 setup to the M2 Pro and higher is disingeneous at best. As others have said, the single NAND configuration will really only make a difference in activities which do a lot of extended reading from or writing to the SSDs. For most use cases the difference is negligible at best. Regarding why Apple switched to 256GB NAND, its because 128GB modules are being phased out by the NAND manufacturers.

Before you start claiming the 512GB is a single NAND, I will point you to the iFixit teardown of the 14" MBP with the M2 Pro:

While Apple is busy splitting the RAM modules down into smaller packages, they’re moving in the opposite direction with the NAND modules, going from four smaller 128GB modules on the 14″ M1 MacBook Pro up to two larger 256GB modules on the 14″ M2 MacBook Pro.

Personally, I'm going to take the word of someone who has actually looked at an M2 Pro logic board over MaxTech (who admittedly crafted a test specifically to choke the aforementioned 14" Pro and cause throttling) or MKB. Your pricing chart is also broken, because it's using EDU pricing and somehow listing Best Buy pricing as identical to Costco. If you want the pricing in that chart to actually mean something, you should be pulling the standard pricing rather than EDU here, Costco there, and refurb somewhere else.

 

[NT1440]

Literally whatever you want. There is no bachelors level of CS that is going to task your machine. You’re learning to code, not building anything that’s going to take hours to compile.

I’d side more on the air for portability BUT if multiple monitors helps you stay organized (I know it does for me) then the pro is the way to go.

 

[Mity]

You "justifications" completely misrepresented the facts at play here. Apple switched to 256GB NAND modules when the M2 Airs were released, so the base M2 SoC with 256GB is running 1 256GB NAND module instead of 2 128GB NAND modules like the base M1 had. That's the only one where speeds are affected. Any configuration of 512GB or more uses two NAND modules, which brings throughput back up. Trying to extrapolate the base M2 setup to the M2 Pro and higher is disingeneous at best. As others have said, the single NAND configuration will really only make a difference in activities which do a lot of extended reading from or writing to the SSDs. For most use cases the difference is negligible at best. Regarding why Apple switched to 256GB NAND, its because 128GB modules are being phased out by the NAND manufacturers.

Before you start claiming the 512GB is a single NAND, I will point you to the iFixit teardown of the 14" MBP with the M2 Pro:



Personally, I'm going to take the word of someone who has actually looked at an M2 Pro logic board over MaxTech (who admittedly crafted a test specifically to choke the aforementioned 14" Pro and cause throttling) or MKB. Your pricing chart is also broken, because it's using EDU pricing and somehow listing Best Buy pricing as identical to Costco. If you want the pricing in that chart to actually mean something, you should be pulling the standard pricing rather than EDU here, Costco there, and refurb somewhere else.

Please go back to the first page and CAREFULLY read what I wrote before jumping on the misinfo bandwagon:

https://forums.macrumors.com/thread...ed-for-computer-science.2392891/post-32251452

"The base M2 has a horrible SSD that is slower than the base M1. Please do not buy any base M2 Machine, be it an Air or Pro. Either get a base M1 or M2 with an upgraded SSD, which would be 512GB for the MBA and 1TB for the M2 MBP."

What does that quote say, especially the last sentence?​

I NEVER conflated M2 with M2 Pro. NEVER. I have TWO SEPARATE posts that compare M1 MBA and M2 MBA:

https://forums.macrumors.com/threads/15-m2-air-or-14-mbp-m2-pro-refurbished-for-computer-science.2392891/post-32251434

and another that compares the M1 MBP and M2 MBP:
https://forums.macrumors.com/thread...ed-for-computer-science.2392891/post-32251452

Did you actually take the time to read what was written and click on the sources? Or did you decide to just jump on the bandwagon that @teh_hunterer started because (s)he can't read?​

You are, however, right about this chart that I had originally uploaded. The M2 MBP has half the number of NAND chips than the M1 MBP. I accidentally wrote 1 instead of 2. You are right on this. However, the speeds that I screenshot are accurate. And my opinion is still valid. Edit: I updated my original post and cited you for the correction.

https://forums.macrumors.com/thread...ed-for-computer-science.2392891/post-32252434

The pricing chart is not broken. Did you read OP's original post? (S)He is as CS student and qualifies for Apple's EDU pricing.

Best Buy:

Costco (in-store only):

Microcenter (in-store only):

 

[Bodhitree]

Between the 14” M2 Pro MacBook Pro and the 15” M2 MacBook Air, I think its a no brainer to go for the Pro. Better soc, more ports outweighs the slightly heavier and slightly less screen area in my mind.

 

[Corefile]

Before you start claiming the 512GB is a single NAND, I will point you to the iFixit teardown of the 14" MBP with the M2 Pro:

I take it you don't work in hardware engineering? The product can change over time as other vendor pricing fluctuates, for cost reduction, or for scarcity of parts.

 

[kschendel]

Man, you are the one that is spreading misinformation.

A 1400 MB/s SSD is not as fast as a 7000MB/s SSD. You can’t argue against facts.

LOL I didn't say it wasn't faster. I said you probably can't notice the difference unless you are doing some very specific large-file thing.

Sports car A can do 0-60 in 2.3 seconds, while "dog slow" car B can only do 0-60 in 4.8 seconds. Which one is faster in city traffic? Wrong answer is car A. Correct answer is that 0-60 is an irrelevant metric in that situation. Same for sequential SSD speeds.

 

[teh_hunterer]

My gosh, this so pathetic. Why did you leave out the rest of my quote? To make it sound like you can actually read? Note the last sentence. The "M1 or M2" refers to the generation. Again for the nth time.

https://forums.macrumors.com/threads/15-m2-air-or-14-mbp-m2-pro-refurbished-for-computer-science.2392891/post-32251452

"The base M2 has a horrible SSD that is slower than the base M1. Please do not buy any base M2 Machine, be it an Air or Pro. Either get a base M1 or M2 with an upgraded SSD, which would be 512GB for the MBA and 1TB for the M2 MBP."

Do you see how it says "be it an Air or Pro?" And how it breaks out 512GB for the MBA and 1TB for the MBP?

I've already explained that you misconstrued M2 machine with the M2 chip, as I explained here but you won't accept your mistake. Now you're just deliberately and desperately stating the same thing over again. https://forums.macrumors.com/thread...ed-for-computer-science.2392891/post-32306526

I salute your reading comprehension skills.

The fact you are referring to the entire M1 or M2 generation is exactly the issue. While the SSD speeds are a valid concern for base models of devices with the M2 chip like the MacBook Air, they are not a valid concern for base models of devices with the M2 Pro chip, because even though the 512GB M2 Pro configurations have SSDs that are half the speed of the 1TB+ configs, they are still fast enough.

The very slowest SSD that can go in an M2 Pro device, ie a 14" M2 Pro MacBook Pro with 512GB, while half the speed of 1TB models, is still much faster than any M1 MacBook Air.

Have you been able to grasp that the M1 Pro and M2 Pro SSDs are far faster than the M1 and M2 SSDs?

The 256GB MacBook Air SSD is actually pretty slow, and I agree that makes it a bit of a weird buy. But the 512GB M2 MacBook Pro SSD is plenty fast.

Did you actually take the time to read what was written and click on the sources? Or did you decide to just jump on the bandwagon that @teh_hunterer started because (s)he can't read?​

You're so combative at this point that you aren't reading anyone's posts properly.

 

[Mity]

The fact you are referring to the entire M1 or M2 generation is exactly the issue. While the SSD speeds are a valid concern for base models of devices with the M2 chip like the MacBook Air, they are not a valid concern for base models of devices with the M2 Pro chip, because even though the 512GB M2 Pro configurations have SSDs that are half the speed of the 1TB+ configs, they are still fast enough.

The very slowest SSD that can go in an M2 Pro device, ie a 14" M2 Pro MacBook Pro with 512GB, while half the speed of 1TB models, is still much faster than any M1 MacBook Air.

Have you been able to grasp that the M1 Pro and M2 Pro SSDs are far faster than the M1 and M2 SSDs?

I know I can't compare Air speeds to the Pro. How many times do I have to show you what I previously wrote? How many times? I am comparing M1 Air with M2 Air and M1 MBP with M2 MBP. I am not comparing Airs to Pros! Why can you not comprehend this?

The OP asked for a recommendation between Air and Pro. I said regardless of which (s)he chooses, go with the M1 gen models or upgrade the SSD on the M2 gen models. And I provided my justification, which was price + SSD performance. And you keep claiming that I'm spreading misinfo, which is just false.

It is PERFECTLY valid to state that the older gen base machines have better SSDs than the newer gens. YOU seem to be explaining away Apple's anti-consumer policies by saying the base M2 MBP "speeds are good enough." That's a ridiculous statement because the they're literally half of the previous gen! Why should a newer model of the same product line have a worse SSD than the previous year? It makes no sense, especially when I can buy the M1 MBP for $1600 and brand new, not a refurb M2 MBP for $1700.

 

[teh_hunterer]

I know I can't compare Air speeds to the Pro. How many times do I have to show you what I previously wrote? How many times? I am comparing M1 Air with M2 Air and M1 Pro with M2 Pro. I am not comparing Airs to Pros! Why can you not comprehend this?

You said please don't buy any base config M2 gen devices. It's a statement that I don't believe really applies to the M2 Pro SoCs because the SSDs in the base config are fast enough.

It is PERFECTLY valid to state that the older gen base machines have better SSDs than the newer gens.

Yes, it is.

YOU seem to be explaining away Apple's anti-consumer policies by saying the base M2 Pro "speeds are good enough." That's a ridiculous statement because the they're literally half of the previous gen!

I'm glad you're actually engaging with this point now. As has been pointed out by others, SSD speeds do eventually hit a point of diminishing returns for real world usage. While I agree that the poor SSD speeds of the 256GB Air are a real issue, because they are slow in the grand scheme of things, I don't agree that the M2 Pro SSD speeds are an issue because they are genuinely fast enough not to really impact real world usage.

They are half the speed - but when you take real world usage into account, it's a number you are taking out of context and misusing to reach a very hyperbolic and I think flawed conclusion about the value proposition of that product.

Why should a newer model of the same product line have a worse SSD than the previous year? It makes no sense, especially when I can buy the M1 MBP for $1600 and brand new, not a refurb M2 MBP for $1700.

It's not great, I get it. But I don't think it's a dealbreaker like it is with the M2 Air. I would hands down take the extra performance of the M2, the extra GPU cores, and the battery life improvements from the extra efficiency cores, and eat the downgraded SSD, for an extra $100. Because the upgrades of the M2 Pro will impact you, and the half speed SSD won't.

 

[Zest28]

LOL I didn't say it wasn't faster. I said you probably can't notice the difference unless you are doing some very specific large-file thing.

Sports car A can do 0-60 in 2.3 seconds, while "dog slow" car B can only do 0-60 in 4.8 seconds. Which one is faster in city traffic? Wrong answer is car A. Correct answer is that 0-60 is an irrelevant metric in that situation. Same for sequential SSD speeds.

You can't notice a difference? That is a bunch of non-sense.

The Xbox Series X is able to launch games in an instant only with a fast SSD's. With slow SSD's it doesn't work. It's a huge difference.

Fast SSD > Slow SSD.

 

[dmccloud]

I take it you don't work in hardware engineering? The product can change over time as other vendor pricing fluctuates, for cost reduction, or for scarcity of parts.

So you're relying on theoretical changes over observed facts. Got it...

 

[dmccloud]

"The base M2 has a horrible SSD that is slower than the base M1. Please do not buy any base M2 Machine, be it an Air or Pro. Either get a base M1 or M2 with an upgraded SSD, which would be 512GB for the MBA and 1TB for the M2 MBP."

What does that quote say, especially the last sentence?​

The problem with this claim is that any M2 Pro or Max MBP starts with a 512 GB SSD, which is TWO 256GB NANDs not one. However, you lump all M2 systems into the same group and make these sweeping generalizations which are not based in reality. Furthermore, even the 512 GB 13" MBP would have dual NANDs as well, which further contradicts your "advice". The fact of the matter is that the SSD speeds would only matter in a handful of scenarios which rely on heavy reads and writes to/from the SSD. For the other 99.5% of use cases, the difference in SSD speeds would be negligible at best, unnoticeable for the vast majority of users who use their computers to check email, pay bills, watch YouTube, etc. Even using office apps (whether from Microsoft or Apple) wouldn't see any measurable performance hit from the "slower" SSD.

 

[LinkRS]

Howdy Everyone,

Something that I have not seen discussed here is screen real-estate. For a CompSci person, or anybody really, an important consideration is screen real-estate. This goes beyond the physical dimensions of the screen, and is a combination of the physical size, the number of available pixels (resolution), aspect ratio, and retina scaling. Typically the Pro series has more real-estate compared to the Air-air line of similar size. Unfortunately this is not something that is commonly mentioned in reviews. It is important as it is what determines how much information you can have on the screen at once, without scrolling. macOS does let you tweak this a bit in the "looks like" description on the display properties, but it can only go so far. You can see the raw pixel counts at the bottom of this page: https://www.apple.com/macbook-air-1...Air-M2-15,MacBook-Pro-14-M2,MacBook-Pro-16-M2 (comparing 15" Air, 15" MBP, and 16" MBP). Good luck!

 

[bill-p]

macOS does let you tweak this a bit in the "looks like" description on the display properties, but it can only go so far.

If you're an advanced user, there's always been a way to expose all resolutions, including native 1:1 for the display.

In Ventura, it is now possible to explicitly make the Display panel in Settings app display all possible resolutions so now even regular users can just set their display to any resolution, as long as their eyes can still read texts at those resolutions.

 

[kschendel]

You can't notice a difference? That is a bunch of non-sense.

The Xbox Series X is able to launch games in an instant only with a fast SSD's. With slow SSD's it doesn't work. It's a huge difference.

Fast SSD > Slow SSD.

It most certainly isn't nonsense. I use the machine I mentioned previously every single day, often 8-10 hours a day for work. Depending on what I need to be working on, I might be using files on the SATA drive, or I might be using files on the fastest NVMe drive in the box. Unless I'm copying gigabytes around (sequentially, with no other processing), there's basically no difference for what I do, and much of that involves large queries against large (50-100 gigabyte) databases.

The Xbox is a very different beast. It's optimized to do sequential I/O as much as possible and that includes file layouts. General purpose OS's aren't. I would expect to see a subjective difference on an Xbox, although not an enormous one in most games.

(Suppose a game file is 5 gigabytes, which is larger than many, but not unreasonable. Assuming a perfectly optimized sequential read, and ignoring CPU effects, the 1.5 GB/s drive will do it in just over 3 seconds, while the fastest 7 GB/s drive will do it in under a second. That's noticeable but hardly world shaking.)

Faster SSD's stop being subjectively faster when the slow SSD is fast enough.

 

[Zest28]

It most certainly isn't nonsense. I use the machine I mentioned previously every single day, often 8-10 hours a day for work. Depending on what I need to be working on, I might be using files on the SATA drive, or I might be using files on the fastest NVMe drive in the box. Unless I'm copying gigabytes around (sequentially, with no other processing), there's basically no difference for what I do, and much of that involves large queries against large (50-100 gigabyte) databases.

The Xbox is a very different beast. It's optimized to do sequential I/O as much as possible and that includes file layouts. General purpose OS's aren't. I would expect to see a subjective difference on an Xbox, although not an enormous one in most games.

(Suppose a game file is 5 gigabytes, which is larger than many, but not unreasonable. Assuming a perfectly optimized sequential read, and ignoring CPU effects, the 1.5 GB/s drive will do it in just over 3 seconds, while the fastest 7 GB/s drive will do it in under a second. That's noticeable but hardly world shaking.)

Faster SSD's stop being subjectively faster when the slow SSD is fast enough.

It is not a subjective difference on the Xbox Series X, as slow SSD’s are literally not usable on the Xbox Series X. It doesn’t work.

Microsoft couldn’t get their instant loading technology working with slow SSD’s.

 

[kschendel]

I expect that has more to do with SSD architecture than SSD speed. I believe the XBox might be using some more recent protocol features that aren't supported on SATA at all, and maybe not even on older NVMe drives. (and definitely not over USB.) You could hook up a mercury delay line storage and have it work on the XBox, very very slowly, as long as you used the internal interface and implemented the necessary protocols.

 

[theorist9]

The fact of the matter is that the SSD speeds would only matter in a handful of scenarios which rely on heavy reads and writes to/from the SSD. For the other 99.5% of use cases, the difference in SSD speeds would be negligible at best, unnoticeable for the vast majority of users who use their computers to check email, pay bills, watch YouTube, etc. Even using office apps (whether from Microsoft or Apple) wouldn't see any measurable performance hit from the "slower" SSD.

What about the times to open apps or large office documents? It seems those would be affected by the number of NAND chips only when the document or app was stored across both chips, and I don't know how common that is.

MaxTech did a video on this (

), but I think he looked specifically for tasks where the difference would be large. He compared these:

And reported results like this. But these are 8 GB machines, and thus might be using SWAP, which would exacerbate the difference in SSD throughput:

 

[solq]

SSD speeds are hardly the only important thing here.

An M2 machine has better longevity (in terms of OS support from Apple) and a better resale value compared to a similarly specced M1.

 

[LinkRS]

If you're an advanced user, there's always been a way to expose all resolutions, including native 1:1 for the display.

In Ventura, it is now possible to explicitly make the Display panel in Settings app display all possible resolutions so now even regular users can just set their display to any resolution, as long as their eyes can still read texts at those resolutions.

Even exposing more resolutions, there is still a practical limit based on the pixel count of the display. The 14" MacBook Pro has more pixels, which could be either used for more information on the screen, or more detail, at the sacrifice of having to scroll around more, or as you said make everything tiny. My last non-Mac laptop had a native 4K display at 17", and I hated it. Everything was so tiny in apps that weren't High-DPI aware (Microsoft speak for retina), on that display. I much prefer the way Apple handles it, but I do get less usable screen space than in Windows. Even my 16" M1 MBP has less native pixels than my old 17" 4K display, but the trade-off was I have still not found (knock-on-wood) a non-Retina aware app for macOS, and there are still many in Windows-land. FWIW, the way Apple handles retina is why "5K" displays seem to be the sweet-spot.