Rene Ritchie.. the voice of reason

[Ethosik]

Why? What is wrong with not having a new processor every year but sustantial improvements every 3-4? In this case I may understand the better video decoding thing (ie M1.2), but otherwise, why do _we_ want a new processor every year or so just for the sake of it? I understand apple, as it is gonna sell the brand new fancy thing, but I am not sure about the usefulness of the general trend in computers of having a 2.5% increase in performance every year, taking into account also important environmental aspects of all this wasteful system. Though, to give credit where credit is due, apple computers seem to last quite long compared to others.

But in any case I do not understand the argument of "keeping the architecture going" from a customer's perspective. When the new version is not good enough for me to justify the price increase, I will just buy the older one anyway.

Intel is catching up fast. Apple needs to keep up the pressure. We don't want Apple to get stale while Intel passes them by (hello history with Intel and AMD). Continual improvements are needed. M1 is almost 2 years old and architecturally much older. We got some really nice improvements on the base M2 and looking forward to what they do with the Pro and Max on these.

I don't want to see Intel blow passed Apple ONCE. Due to how well the Apple Silicon transition is going for apps running natively (sarcasm BTW). The only thing that has made me appreciate Apple Silicon is the additional stuff they cram in the SOC like the encoders/decoders and neural engine and how quiet it is in a laptop. Without those video encoders/decoders, it would be the exact same as my older i9 Windows PC in terms of performance for video editing.

 

[thadoggfather]

My favorite parts of the Rene video, rewatching it for the cringe vibes are:

'Back in the day, major tech publications had dedicated labs for testing this stuff staffed with the best and brightest and they wrote custom code to interrogate, they really deeply understood components'

and later 'this has created benchmark LARP' - where's the LARP? Casually discussing specs on the Internet?

It plays like something that would be fired out of Buzzfeed's buttocks. Sensationalized ADD nonsense and big words coming up on the screen as he says them for emphasis as if we're watching a program for children.

The argument is unless you work for a 'major tech publication' and rip the thing apart and create custom code, even if you just rip it apart and take a look and use 'off-the-shelf' benchmarking apps, you are not at all qualified to make an assessment or judgment even in terms of whether or not you should buy the machine.

Hyper elitist take.

 

[Zest28]

def pro_review(computer): print("The "+computer+" , is the fastest computer we have ever measured based on our custom benchmarks!") pro_review("M2 MBA")

Now I am a major tech reviewer from the old days. This piece of code is better than all the "off-the-shelf" benchmarks. This code should only be run if Apple pays enough ofcourse.

 

[burgman]

You're missing the point completely. The old 2021 M1 Air with 256/8GB would have been fine. The new one, not so much.

More expensive, newer model... 50% slower than last year's to do certain tasks. And then you get responses like yours that just say to spend more money on RAM. What a joke.

Are there any threads here about real world noticing slowdowns or just the usual MacRumors/social media game of telephone about synthetic benchmarks? More entertaining ones sound like something from Fast and Furious🙂

 

[DavidChoux]

Are there any threads here about real world noticing slowdowns or just the usual MacRumors/social media game of telephone about synthetic benchmarks? More entertaining ones sound like something from Fast and Furious🙂

Yes there's a video of Max Tech comparing the new M2 Air to the old M1 Air, both 256/8GB.
He did a LR export and switched over to Chrome to browse while LR was exporting. Was something significant like 50% slower for the new M2 Air.

 

[turbineseaplane]

Yes there's a video of Max Tech comparing the new M2 Air to the old M1 Air, both 256/8GB.
He did a LR export and switched over to Chrome to browse while LR was exporting. Was something significant like 50% slower for the new M2 Air.

Haven't you heard David?

The MacBook Air is a $1200 machine that is to be used only for checking email or scanning Facebook

Or so I keep hearing..

 

[Analog Kid]

Doesn't have to be. The entire contention is ridiculous. Fast SSD's make a big difference in terms of swap, period. You're just theory crafting your numbers and your conclusion is wrong.

Someone doesn't need to defend their statements? The contention that more RAM has a bigger impact on performance in a RAM limited machine compared to a faster SSD is ridiculous?

Why? Seriously. I laid out my thoughts on the matter and all I've gotten back has been "don't ask me, search" and ad hominem attacks-- which can only be interpreted as someone not knowing what they're talking about and being uncomfortable about being asked.

 

[Ethosik]

Yes there's a video of Max Tech comparing the new M2 Air to the old M1 Air, both 256/8GB.
He did a LR export and switched over to Chrome to browse while LR was exporting. Was something significant like 50% slower for the new M2 Air.

How should we trust their review based on what this post outlines?

https://forums.macrumors.com/threads/rene-ritchie-the-voice-of-reason.2351574/post-31272218

An almost full SSD is very very VERY slow.

They (Max Tech) needs to outline their full testing strategy like GamersNexus does.

What apps are installed?
How much of the SSD is used up?
Is it running the same OS and software versions?
When was it last booted?
When was it last restored to base macOS image?
Is it running in isolation or could a rogue app be running in the background skewing the results?
Is there an SSD bottleneck that was the cause of this percentage drop?
WHAT was the cause of the drop? Just processing images in lightroom should not be saturating a 1.5 and 3 GB/s drives.
Is it swap that caused this percentage decrease? Well guess what, that proves you just need more RAM. NEVER depend on swap performance for work systems.

Also, with ads these days one website can be nice or bad depending on which ads choose to display. "browsing" is a very very vague test. Is the same website with the same content with the same ads being presented? With the same javascript running?

 

[Asiatic Black Hebrew]

Someone doesn't need to defend their statements? The contention that more RAM has a bigger impact on performance in a RAM limited machine compared to a faster SSD is ridiculous?

Why? Seriously. I laid out my thoughts on the matter and all I've gotten back has been "don't ask me, search" and ad hominem attacks-- which can only be interpreted as someone not knowing what they're talking about and being uncomfortable about being asked.

Honestly, are you surprised? This has been happening for days. It seems like a few of these posters here just want to exist in an echo chamber & probably don't even have a good enough understanding of the points they're trying to make, to actually have an actual technical discussion about it. That has been proven countless times in this very thread, as thought out technical responses/questions often go unanswered or dismissed, with either conjecture, or willful ignorance.

The majority of this thread now, is just regurgitated talking points from the same few posters, or ad hominems against the YouTuber. I would say expecting to have any meaningful discussion at this point, is not only a fool's errand, but a tremendous, if not complete, waste of your time.

 

[hans1972]

Yes there's a video of Max Tech comparing the new M2 Air to the old M1 Air, both 256/8GB.
He did a LR export and switched over to Chrome to browse while LR was exporting. Was something significant like 50% slower for the new M2 Air.

Yes, but that was using a professional software package like Lightroom and exporting 48 MP images. If you're a photographer and using Lightroom, would you really buy the 256GB SSD version and use the internal SSD for your photos?

How about an example using Photos exporting 10 JPEG photos taken with an iPhone?

That would be a much more realistic test for the users of the MacBook Air 256Gb SSD version.

 

[m1maverick]

Yes, but that was using a professional software package like Lightroom and exporting 48 MP images. If you're a photographer and using Lightroom, would you really buy the 256GB SSD version and use the internal SSD for your photos?

How about an example using Photos exporting 10 JPEG photos taken with an iPhone?

That would be a much more realistic test for the users of the MacBook Air 256Gb SSD version.

If true then the M1 would be a better choice for the MBA target marker as it costs less than the M2.

 

[rkuo]

Someone doesn't need to defend their statements? The contention that more RAM has a bigger impact on performance in a RAM limited machine compared to a faster SSD is ridiculous?

Why? Seriously. I laid out my thoughts on the matter and all I've gotten back has been "don't ask me, search" and ad hominem attacks-- which can only be interpreted as someone not knowing what they're talking about and being uncomfortable about being asked.

BUT DID YOU DO ANY RESEARCH OR NOT? You can only interpret it as an "ad hominem"? For what possible reason is that your only interpretation?

You quoted RAM vs SSD bandwidth. It's extremely rare for apps to actually saturate RAM throughput. So your contention in the other post that hitting swap is simply hitting the performance wall regardless of the SSD's speed is very very wrong. SSD's resulted in a huge swap performance improvement relative to HDD's, and doubling the throughput again in this day and age is still very significant.

 

[Analog Kid]

Skipping to the interesting part of your comment:

You quoted RAM vs SSD bandwidth. It's extremely rare for apps to actually saturate RAM throughput. So your contention in the other post that hitting swap is simply hitting the performance wall regardless of the SSD's speed is very very wrong.

It's a pretty simple assertion that's being tested here: is not swapping faster than swapping. With more RAM, you don't swap, with less RAM you swap. Not swapping will always be better unless you somehow imagine the SSD is actually more performant than RAM.

All of the attention is on the throughput specs because that's what one set of circus clowns decided to publish and everyone else keeps repeating. The throughput to RAM is much, much higher than the throughput to the SSD.

But as I've pointed out multiple times in this thread, including in the comment that you eventually replied into, throughput isn't really that important to swap performance. Latency is the name of the game, and there is a massive difference in latency between access to the unified memory system and swap space and likely no difference in latency between the 512GB and 256GB SSDs in the M2s (and possibly an advantage toward the non-RAID 256GB drive).

SSD's resulted in a huge swap performance improvement relative to HDD's, and doubling the throughput again in this day and age is still very significant.

As I pointed out, SSDs have about 100-1000 times higher latency than non-cache RAM based on the ballpark numbers I can find for SSD access and the Anandtech memory access latencies for the M1-- microseconds versus nanoseconds. Rotating media has about 1000 times higher latency than flash, milliseconds versus microseconds.

So moving from rotating to solid state was a 1000x improvement. Adding RAM is on the order of 1000x improvement. The doubling you're talking about is throughput, not latency, but even if it was I don't see 2x really moving the needle in that context.

 

[Zest28]

Honestly, are you surprised? This has been happening for days. It seems like a few of these posters here just want to exist in an echo chamber & probably don't even have a good enough understanding of the points they're trying to make, to actually have an actual technical discussion about it. That has been proven countless times in this very thread, as thought out technical responses/questions often go unanswered or dismissed, with either conjecture, or willful ignorance.

The majority of this thread now, is just regurgitated talking points from the same few posters, or ad hominems against the YouTuber. I would say expecting to have any meaningful discussion at this point, is not only a fool's errand, but a tremendous, if not complete, waste of your time.

Dude, I have a 16" M1 Max MacBook Pro with 32 GB RAM. If I need the best performance, I already have it.

You and the other guy can read all about it here on MacRumors how memory swapping was sufficient on the base M1 MBA. Plenty of real world experiences from people using the base M1 MBA on MacRumors. This has nothing to do with "ignorance". It is called referencing.

The M2 MBA is not supposed to replace my 16" M1 Max MacBook Pro with 32GB RAM, but a cheap travel device. And memory swapping is fine with a fast SSD for this purpose, only if the device is cheap.

 

[DavidChoux]

-deleted-

 

[DavidChoux]

How should we trust their review based on what this post outlines?

https://forums.macrumors.com/threads/rene-ritchie-the-voice-of-reason.2351574/post-31272218

An almost full SSD is very very VERY slow.

They (Max Tech) needs to outline their full testing strategy like GamersNexus does.

What apps are installed?
How much of the SSD is used up?
Is it running the same OS and software versions?
When was it last booted?
When was it last restored to base macOS image?
Is it running in isolation or could a rogue app be running in the background skewing the results?
Is there an SSD bottleneck that was the cause of this percentage drop?
WHAT was the cause of the drop? Just processing images in lightroom should not be saturating a 1.5 and 3 GB/s drives.
Is it swap that caused this percentage decrease? Well guess what, that proves you just need more RAM. NEVER depend on swap performance for work systems.

Also, with ads these days one website can be nice or bad depending on which ads choose to display. "browsing" is a very very vague test. Is the same website with the same content with the same ads being presented? With the same javascript running?

I completely forgot about that post.

After reading it, and rewatching that Max Tech video to confirm what was said, I have no choice but to completely disregard that Max Tech video. A little embarrassed that I skipped all that talk at the start about him justifying why he believed it fair to test the 256GB drive while it was close to 100% capacity. For a supposedly scientific test, that Max Tech chap's logic is all over the place.

Thanks @Analog Kid for that detailed analysis.

 

[idktbh]

"toxic benchmark culture" are ****ing kidding me? my god this dude is the worst

 

[chengengaun]

@Analog Kid Thank you very much for the detailed critique. Frankly, I have no appetite to watch the two videos on both sides of the debate or have much expectations about the ensuing discussions, but you have really summarised the issues succinctly. Bravo.

 

[thadoggfather]

Doesn’t want a studio display because not mini led

Yet he loves it and thinks reviews are completely out of touch

Makes sense!

 

[temende]

Doesn’t want a studio display because not mini led

Yet he loves it and thinks reviews are completely out of touch

Makes sense!

I still find it funny that he works for Google after years of making videos about their data harvesting malpractices.

 

[turbineseaplane]

I still find it funny that he works for Google after years of making videos about their data harvesting malpractices.

A good reminder of how anyone, any voice, any brand .... can be co-opted into some "new and updated viewpoints" by money.

Related... It's also a reminder how sponsorship and access to information/products absolutely impact how folks cover Apple and how critical (or not) they are.

 

[thadoggfather]

Where’s Rene to oooo and ahh over 48MP and A16 and AOD and peak 2000 nit brightness specs?

 

[canadianpj]

"toxic benchmark culture" are ****ing kidding me? my god this dude is the worst

What's wrong with that statement? I see post after post after post going on about benchmarks which means nothing to 99.999% of people actually using the computer. I do this really weird thing when I get a new computer, I just use it. I don't test how hot it gets, I don't test benchmarks.

I install my programs and I just use it.

I've had it since launch, with zero problems, we all really should move on from this thread.

 

[ericwn]

What's wrong with that statement? I see post after post after post going on about benchmarks which means nothing to 99.999% of people actually using the computer. I do this really weird thing when I get a new computer, I just use it. I don't test how hot it gets, I don't test benchmarks.

I install my problems and I use it.

I've had it since launch, zero problems, we all really should move on from this thread.

I’ll refer you to post 9 for a sane counter statement.

 

[sevoneone]

I install my problems and I use it.

This is the best typo ever, I am so using this statement all the time now.