Rene Ritchie.. the voice of reason

[turbineseaplane]

Rene:

Guys .. these benchmarks don't look good for Apple on the M2 -- performance comparisons and benchmarks create toxic narratives (about an Apple product, in this case)..
No more benchmarks and performance comparisons!!

Also Rene:

 

[Ethosik]

Finally it is Apple who is boasting that the MacBook Air can edit 8K video not the reviewers.

Is it the base model or the maxed out one? Where did Apple state this? 8GB of RAM is nearly impossible for 8K work.

 

[Ethosik]

I’ll just add an anecdote from my own personal experience.

I have an M1 Air, 16GB/512GB. I am able to run Parallels with Windows 11 ARM using Visual Studio 2022 doing .net 6 based development and my machine never gets hot nor slows down that I can tell.

I will usually have 4 tabs open in macOS Safari and 4 or 5 in Microsoft Edge Chromium in Windows, plus Visual Studio doing builds, etc.

When I’m doing 20 minute Visual Studio builds in the Windows VM, while browsing and listening to music in macOS, my M1 Air barely gets warm to the touch and I’ve never experienced any perceptible slowdown.

That’s what’s achievable with my M1 Air. It’s head and shoulders faster and cooler (and 100% silent) than my 2018 MacBook “Pro”. The same work on that machine literally made it too hot to hold. I had to get a lap stand, and the screaming fan made it unpleasant to work with, not to mention sucking dust into the chassis and onto components. Oh, and builds were dog slow.

I’ve read a lot about how the Air shouldn’t be used for “Pro” workloads. That’s subjective. It should be used for whatever purpose it is fit for. The reviews that have contrasted the M1 and M2 Air performance were very helpful. Objectively, the M2 throttles sooner than the M1 and in some cases workloads take longer than on the M1 due to that. Valuable information to have for many of us “Pros” who “want it all” - cool, fast and silent - as the M1 Air delivered.

If your mom (not to pick on moms, but this is a frequently quoted demographic) won’t be running the Pro workloads as shown in some reviewers’ videos, then you’re likely correct that the base model would suit them fine. To each their own. At least we have the benchmark and workload testing results and can use them to inform our purchasing decisions. 😊

There is a difference between the base model (which is what this SSD mess is ALL ABOUT, which is the one aimed at parents) and a maxed out Air. 8GB RAM can’t be used for 8k video editing. But a maxed out model does.

 

[dumastudetto]

Rene is the leading voice in the Apple community.

His word is authoritative for me.

The new MacBook Air is the best damn laptop Apple has ever made.

 

[BellSystem]

Rene:

Guys .. these benchmarks don't look good for Apple on the M2 -- performance comparisons and benchmarks create toxic narratives (about an Apple product, in this case)..
No more benchmarks and performance comparisons!!

Also Rene:

View attachment 2031836View attachment 2031837View attachment 2031838View attachment 2031839View attachment 2031840View attachment 2031842

Exactly...someone insulted the M2 MacBook and hurt his feelings and made him cry. Then he went on a Kool-Aid bender (paid for by Apple). He reminds me of every person I worked with at Apple that I didn't like....which is why I didn't stick around....which also convinces me he is a paid Apple marketing employee. Maybe he just didn't have anything else to talk about.

 

[atownguy]

Rene:

Guys .. these benchmarks don't look good for Apple on the M2 -- performance comparisons and benchmarks create toxic narratives (about an Apple product, in this case)..
No more benchmarks and performance comparisons!!

Also Rene:

View attachment 2031836View attachment 2031837View attachment 2031838View attachment 2031839View attachment 2031840View attachment 2031842

To be fair, part of the point of his video was him admitting he made videos like this in the past and that he was part of the problem, and he explained why he's not making videos like this anymore. He even flashed bunch of his thumbnails like this up there when he was talking about it.

I'm not saying I agree with his points, but he's at least aware of the irony of him making a video like this now.

 

[thadoggfather]

Is there any other benefit than 'benchmarks' when the 2nd gen re-design Air next year or the following inevitably is physically identical to its predecessor but will have different internals?

He's kinda positioned himself in a corner to gloss over Mac upgrades for the next 5-10 years if he is deciding to only addressing them when re-designed or some new features.
Same with 'S-cycle' phones, though those usually have another bell or whistle.

 

[turbineseaplane]

To be fair, part of the point of his video was him admitting he made videos like this in the past and that he was part of the problem, and he explained why he's not making videos like this anymore.

I’ll be shocked if he isn’t back at it eventually…next time Apple has “torched” the competition on some metric

 

[techno-Zen]

Rene:

Guys .. these benchmarks don't look good for Apple on the M2 -- performance comparisons and benchmarks create toxic narratives (about an Apple product, in this case)..
No more benchmarks and performance comparisons!!

Also Rene:

View attachment 2031836View attachment 2031837View attachment 2031838View attachment 2031839View attachment 2031840View attachment 2031842

OMG, I am absolutely DEAD

 

[thadoggfather]

He actually looks deeply troubled in this thumbnail that people would dare doubt the awesomeness of M2 machines.

Rene Ritchie is the voice of Apple, not reason.

 

[clevins]

Benchmark, however flawed it might be, by far is THE most scientific view of a machine’s performance in different metrics and can be compared relatively objectively. Now, whether user takes benchmark seriously or not matters little of other users decisions, but none should dissect the importance of benchmarks recording the progression of our hardware design and evolution.

Benchmarks have to reflect a) the intended use the machine being tested and b) correlate to actual real world use.

 

[kp98077]

Rene:

Guys .. these benchmarks don't look good for Apple on the M2 -- performance comparisons and benchmarks create toxic narratives (about an Apple product, in this case)..
No more benchmarks and performance comparisons!!

Also Rene:

View attachment 2031836View attachment 2031837View attachment 2031838View attachment 2031839View attachment 2031840View attachment 2031842

omg... too funny...i am kinda tiered of all these "benchmark" experts, what makes them an expert? Why cant people just try and use the computer for their normal flow and if it doesn't work just return it! LOL

 

[clevins]

If you disagree with benchmarks, I guess you also don’t give scientific research a ton of weight because they don’t really reflect real use either. However, what benchmark can do is giving users a clear idea of relative performance between different hardware and different configs of same classes of hardware. Yes, most people don’t intentionally stress their machine 7*24 at 100% load, but if 100% load performance has noticeable differences, one can reasonably assume medium to low load would have some differences as well, albeit less pronounced.

No, you cannot assume that. And this is why the narratives around benchmarks is stupid.

An Air will absolutely throttle if you peg the CPU/GPU for hours. It will throttle after a few minutes. But that says nothing about what it will do in a bursty use scenario where someone is normally cruising along well below 50% but occasionally bursts to 90% for several minutes. Or how it will work if the CPU is being used, consistently, at some moderate percentage (30-60% say) but never jumps above that.

If we wanted to know that... we'd need benchmarks that emulate those scenarios. All current "stress this system as much as possible" marks will tell you is... what happens when you stress that subsystem.

That CAN be useful. If someone routinely does run stuff that stresses, say, the GPU then they can compare model to model (Air to Pro to Max, say) and see where performance works for their use case. That's ALL the benchmarks are good for.

 

[Zest28]

Apple: “Shows benchmark where M1 is faster than Intel”

Rene: “OMG!!!!!!! WOW!!!!!!!!”

Max Tech: “Shows benchmark where M2 SSD is 50% slower than the M1 SSD”

Rene: “TOXIC!!! TOXIC!!!!”

 

[Asiatic Black Hebrew]

omg... too funny...i am kinda tiered of all these "benchmark" experts, what makes them an expert? Why cant people just try and use the computer for their normal flow and if it doesn't work just return it! LOL

It sounds so simple doesn't it? But the sad truth is that they want YOU to be unhappy, when a product that YOU love, doesn't work for them.

 

[kp98077]

It sounds so simple doesn't it? But the sad truth is that they want YOU to be unhappy, when a product that YOU love, doesn't work for them.

totally! they want us to be sad, or evoking emotion.... kinda like media.. oy!

 

[camotwen]

Benchmarks are a standardised way to get some data on different machiness' performance. All this drama is stupid and pointless. One can factor in the benchmark results in the decisions on what to buy themselves, there is no single intended use or one single real world usage of these machines, it is up to each person to know what their own intended use is and how benchmark results relate to it. All the rest is just soap opera the tech youtube tries to sell.

 

[svenmany]

I know nothing about this guy, so I can consider his video in isolation from his other stuff (his supposed Apple bias that many are mentioning). I found the video to be impressive and it made many good points.

Here's my unsubstantiated gut feel...

Benchmarks are great if they are understood. But, probably most are taken out of context and being consumed by people who don't critically evaluate them. Probably people draw the wrong conclusions from them. People focus on things they understand (I do). And, reviewers focus on things people want to read (which are the things they understand and are inflammatory). We'll find lots of videos about disk speeds, far fewer about bus speeds. And the videos about bus speeds will be of an educational nature and not designed to enrage the viewer.

In software we write repeatable tests at various levels. We test small pieces of code. We test collections of those pieces as integrated pieces. We also do full end to end tests that simulate what a user would experience. I'm sure the same kinds of things exist in benchmarking. The end to end kind should be the ones users care about. But they're harder to craft and the results are harder to understand. The video seems to be complaining about the kind of benchmarking that would be analogous to testing small pieces of code. Those kinds of benchmarks are easy to craft, easy to understand, and easy to misinterpret.

The size of the universe is 5. The details are left to the reader.

 

[IllinoisCorn]

Rene makes many good points

Solid perspective...

Shilly McShillstein

 

[BootsWalking]

Did you guys actually watch the video? He makes very good points. I find a lot of tech in the last 5 years or so has been way to heavy with benchmarks with the way he stated. Download Benchmark app, click Go and produce results and say how HORRIBLE the new i9 is compared to AMD just because of one benchmark number.

People like Linus, GamersNexus and JayzTwoCents are at least a little better where they test dozens of things. GamersNexus especially since he goes so far in depth in his testing strategy. But its still a test in isolation. What we need is more of a "productivity review after 1-2 weeks" reporting instead of "Product has been out for 1 minute here are all the benchmarks and my thoughts" content.

I much prefer those type of content than ANY benchmark, even if they are in favor of the product I want.

Benchmarks only tell a story that a system is as good/bad FOR THAT BENCHMARK. Benchmarks of a NVIDIA GPU is all good, but my GPU can be poor or great depending on the game due to program optimization. Relying on benchmarks for a purchase is not a good way to approach this.

This is ultimately leading me to being less and less interested in tech news/reviews and new products in general. Especially with Apple. It gets 100 points less in a Geekbench score and the entire sky is falling. Or testing a base Macbook Air with a workflow that is not intended for it and saying "Yeah it sucks". Its like me doing a review for a 3060 and saying "Yeah it sucks since it can't do 4K gaming well".

I watched the video and think he's completely off base. He talks about how complex modern designs like the M1/M2 SOCs and Intel's performance/efficiency cores are and how that makes the benchmarks less reliable. He's wrong. Many benchmarks are designed with real-world workloads and mimic how applications implement logic. Applications aren't supposed to have to concern themselves with the intricacies of these modern CPU models - that's the job of the Operating System and its scheduler. If a benchmark runs poorly using common algorithms implemented in common ways then so will the real-world applications they're designed to mimic.

The fact he thinks you need to tailor your benchmarks to fully take advantage of the peculiarities of a platform actually makes his results less useful than the turnkey benchmarks he's panning.

 

[Danfango]

Rene is a YT influencer with a background in web design and media, not a computer scientist, engineer or anyone with any experience on programming.

I've got 30 years of experience behind me in these areas spanning EE (RF, pushing FETs around in cadence), embedded systems engineering and commercial software engineering and I'm 100% not going to start a YT channel with speculative and misinformed crap because that's all I will turn out as well without YEARS of research on top of what I already know.

His videos and these incessant channels full of garbage need to just go away. They are merely contributing to the heat death of the universe for the sake of getting attention and advertising revenue.

 

[Zest28]

I watched the video and think he's completely off base. He talks about how complex modern designs like the M1/M1 SOCs and Intel's performance/efficiency cores are and how that makes the benchmarks less reliable. He's wrong. Many benchmarks are designed with real-world workloads and mimic how applications implement logic. Applications aren't supposed to have to concern themselves with the intricacies of these modern CPU models - that's the job of the Operating System. If a benchmark runs poorly using common algorithms implemented in common ways then so will the applications.

The fact he thinks you need to tailor your benchmarks to fully take advantage of the peculiarities of a platform actually makes his results less comparable to actual end-user experience than the turnkey benchmarks he's panning.

And the SSD speeds were compared as Apple Silicon vs Apple Silicon.

If it was because of an ARM limitation why the SSD speed is so low, the M1 should not be doing double the speed.

With the same test, my 16” M1 Max MacBook Pro does 7000MB/s, which is 5 times faster.

So you cannot blame it at ARM being terrible at these SSD benchmarks.

 

[thadoggfather]

With the same test, my 16” M1 Max MacBook Pro does 7000MB/s, which is 5 times faster.

So you cannot blame it at ARM being terrible at these SSD benchmarks.

theres people out there in third world countries living in mud huts and shoeboxes starving for SSD speeds, and just like that, you drop that info about the blazing speeds of the 16".

:O impolite

 

[thadoggfather]

One thing to think about, maybe, with Apple stuff if you're really into 'the platform', etc. and still set on it regardless for the long haul despite the minor speed bumps along the way, and aren't really strongly considering switching to another platform any time soon, is not to compare benchmarks from Apple-to-Android or Apple-to-Windows since that literally makes no difference to you who's staying put and maybe just not upgrading as frequently and sort of trivial to know, but if anything it is more than fair, reasonable, and in fact logical its to compare the new product to its predecessor it is in essence replacing as the following generation, even if still sold along side the previous, especially when more money is being asked for it in exchange. How can that possibly be cast as some obscure train of thought?

The bottom line is the SSD is slower M2 Air than previous base model M1. You can say 'it doesn't matter', 'its marginal', 'its imperceptible' and there's a debate to be had and all sorts of things, and even if it does not matter, the fact remains that it is slower. That is a statement of fact, what you do with it is your call its just a consumer electronic device it's not really going to matter. It's as if an iPhone gets worse battery life than its predecessor, that is a regression you can say 'its not that bad' or 'I charge my phone about just as frequently as before' or 'its worth it' but the fact remains there too in those situations and what else can you say but thats disappointing? Is it the end of the world? Nah, but we hope that they pick the ball back up after dropping it, with the next one.

If the mini 7 screen is kinda meh, I'll probably continue to hold onto my mini 5. I want a new mini, I want to give them money for it, I Just dont want a regression in screen quality, jelly gate aside, for a wad of cash-- it is a perceptible downgrade to my eyes. If I were to put up with that, I'd only consider it for really cheap I'm not gonna spend top dollar on something that is worse than what I have in a key way.


I've never really cared about how one platforms stacks up against another, as a concession of dabbling with Mac during the IBM processor days, I accepted a slower processor than Intel is what it is. I think A(x) series chips and M are now leagues beyond Qualcomm and Intel for their respective classes but battery life is important too and other things. I dont care to compare megapixels of the iPhone vs an android, doesnt matter to me, nor do I think that 8MP of one generation iPhone is identical to its successor 8MP camera because the device on paper has same MP, the optical sensors matter and software... does the device produce good pictures? Are they better and clearer than previously? is my benchmark there.

...but I certainly think its fair to compare 2020 Air to 2022 one and expect at least parity on everything and hopefully surpassing it in some/many/most ways.

--

It does kinda stink that they miss the mark on things that should be obvious to not have to deal with.

Like make the display controller the other orientation, figure it out, and use less crappy panels for Mini 6.
You're already charging $100 more and the display has never had these worries with the previous one.

And for M2 Air, have two NANDs. You're already charging $200 more. Make it right, offer for those wanting it - to replace single nand 256 with 2x128 or 2x256gb if out of those and if really feeling generous and repurpose those 1x256gb for other 2x256 Pro's, etc. as a supply chain guru running the show figure out how to navigate it without losing your heiney on profit.
And ship new units moving forward with two NANDs and apologize and move on.

I would love to see them do something that's a stand up act at this point. They have the resources to. Instead the statement is 'it looks like its slower' but overall its faster because its M2. No it is slower, the drive at least , can't speak to the overall experience but independent of that.

I was excited about the prospects of a 16" air, or another throwback to the 12" or something ultra compact like that, Both extremities, and still am slightly, but now I'll keep expectations more tempered on some tomfoolery with prices, features, specs, or some combo.

The M1 Air @$999 for what it offered and continues to is still such a grand slam. They're quickly going off the beaten path and 'Pro-ifying' the Air price.

 

[exoticSpice]

One thing to think about, maybe, with Apple stuff if you're really into 'the platform', etc. and still set on it regardless for the long haul despite the minor speed bumps along the way, and aren't really strongly considering switching to another platform any time soon, is not to compare benchmarks from Apple-to-Android or Apple-to-Windows since that literally makes no difference to you who's staying put and maybe just not upgrading as frequently and sort of trivial to know, but if anything it is more than fair, reasonable, and in fact logical its to compare the new product to its predecessor it is in essence replacing as the following generation, even if still sold along side the previous, especially when more money is being asked for it in exchange. How can that possibly be cast as some obscure train of thought?

The bottom line is the SSD is slower M2 Air than previous base model M1. You can say 'it doesn't matter', 'its marginal', 'its imperceptible' and there's a debate to be had and all sorts of things, and even if it does not matter, the fact remains that it is slower. That is a statement of fact, what you do with it is your call its just a consumer electronic device it's not really going to matter. It's as if an iPhone gets worse battery life than its predecessor, that is a regression you can say 'its not that bad' or 'I charge my phone about just as frequently as before' or 'its worth it' but the fact remains there too in those situations and what else can you say but thats disappointing? Is it the end of the world? Nah, but we hope that they pick the ball back up after dropping it, with the next one.

If the mini 7 screen is kinda meh, I'll probably continue to hold onto my mini 5. I want a new mini, I want to give them money for it, I Just dont want a regression in screen quality, jelly gate aside, for a wad of cash-- it is a perceptible downgrade to my eyes. If I were to put up with that, I'd only consider it for really cheap I'm not gonna spend top dollar on something that is worse than what I have in a key way.


I've never really cared about how one platforms stacks up against another, as a concession of dabbling with Mac during the IBM processor days, I accepted a slower processor than Intel is what it is. I think A(x) series chips and M are now leagues beyond Qualcomm and Intel for their respective classes but battery life is important too and other things. I dont care to compare megapixels of the iPhone vs an android, doesnt matter to me, nor do I think that 8MP of one generation iPhone is identical to its successor 8MP camera because the device on paper has same MP, the optical sensors matter and software... does the device produce good pictures? Are they better and clearer than previously? is my benchmark there.

...but I certainly think its fair to compare 2020 Air to 2022 one and expect at least parity on everything and hopefully surpassing it in some/many/most ways.

--

It does kinda stink that they miss the mark on things that should be obvious to not have to deal with.

Like make the display controller the other orientation, figure it out, and use less crappy panels for Mini 6.
You're already charging $100 more and the display has never had these worries with the previous one.

And for M2 Air, have two NANDs. You're already charging $200 more. Make it right, offer for those wanting it - to replace single nand 256 with 2x128 or 2x256gb if out of those and if really feeling generous and repurpose those 1x256gb for other 2x256 Pro's, etc. as a supply chain guru running the show figure out how to navigate it without losing your heiney on profit.
And ship new units moving forward with two NANDs and apologize and move on.

I would love to see them do something that's a stand up act at this point. They have the resources to. Instead the statement is 'it looks like its slower' but overall its faster because its M2. No it is slower, the drive at least , can't speak to the overall experience but independent of that.

I was excited about the prospects of a 16" air, or another throwback to the 12" or something ultra compact like that, Both extremities, but now I'll keep expectations pretty tempered on some tomfoolery with prices, features, specs, or some combo.

I find the only problem is the SSD issue not the thermal throttling cause the MacBook Air designsed to slow down after 5-10 minutes.

Now the SSD Apple should have mentioned the Speeds on their website.