The M2 is a great chip. That's why I think a chip that can beat it would also be great. Since the M3 is not available, benchmarking against the M2 seems like a pretty reasonable thing to do. The M2 was a nice bump over the M1, and the M3 doesn't promise huge gains over the M2, so AMD's result here is still pretty exciting.
Actually M1 performance per watt is unbeatable, M2 has lower perfomance value per watt consumed. There almost zero chance AMD or Intel can release anything remotely close to M1
This just means it has more cores (edit: it’s using SMT and homogeneous cores for 16 logical cores). A more fair comparison would be against an M2 Pro. It has the same TDP as the pro.
Great to see some competition coming but yeah sounds very clickbaity, it’s comparing the base M2 to what will apparently be the top of the line new AMD chip, instead of comparing it to the M2 max for example.
“
the Ryzen 7 7840U, the initial flagship APU of the new series, apparently offers 9% better 3D rendering performance, 14% better responsiveness, and a whopping 72% better multiprocessing performance than the standard M2
There will be four new chips: the aforementioned Ryzen 7 7840U, the midrange Ryzen 5 7640U and 7540U, and the affordable Ryzen 3 7440U
“
Also they’re not going to win on the marketing front with that naming strategy. Just awful. M1/2 and Pro/Max are so much simpler. It’s like Intel’s naming, such a mess.
The range of SKUs is a problem AMD and Intel (particularly) brought upon themselves. It’s not easy for anyone trying to buy a laptop to truly estimate the power of the processor in there without first going through a review video. Id take the narrow range over that mess.
> The range of SKUs is a problem AMD and Intel (particularly) brought upon themselves.
Its a problem for simplicity of naming, its not a problem beyond that (and the alternative is a problem if you aren’t a vertically integrated producer supporting a fairly narrow end-user product line.)
The 7945HX is already faster than the M2 Max and already available. The U line are competitors to the base M2 models.
Apple's names are pretty awful, I get Pro/Max/Ultra confused and they aren't even unique identifiers either. Which M2 Pro? 10 core or 12 core? Why do CPUs with the same name have different core counts?
> Also they’re not going to win on the marketing front with that naming strategy. Just awful. M1/2 and Pro/Max are so much simpler. It’s like Intel’s naming, such a mess.
Apple has an easier time with processor names because they have a single customer that wants a simple product lineup. Amd and Intel sell their processors to OEMs that want to have products for every budget, so you end up with lots of processor skus and lots of laptop skus and have to do a bit of research to understand what you would be buying and have to decide what you want. On the plus side, you have lots of choices, on the minus side, you have lots of choices.
Simple processor naming doesn't seem to stop Apple from naming every iPad 'the new iPad' though. But then again, other OEMs like to use the same product name and number for 3-7 different devices these days too, so everything is horrible, as usual.
Apple produces fewer chips but their naming isn't much better.
Is the M1 Ultra faster than the M2 Pro? Is the M1 Max faster than the Ultra, or is the Max not the maximum? How does the M1 Pro compare to the M3?
AMD generally uses the "bigger number within a generation is faster" and Apple uses some confusing adjectives. Either way the names only give a vague indication of performance.
The naming of products has gotten really silly and useless, and I put this blame on Intel and Apple. Apple started down the path of Pro, Pro Max, Ultra, Max, and Plus. Remember the iPhone XS? Idiotic naming there. Intel, otoh, just started doing crazy numeric strings with a random character added to the end. And this was after bad naming schemes like "Core 2 Quad". Part of the issue with GPUs and CPUs is the need to do binning to make IC manufacturing commercially viable so companies end up with a ton of SKUs, but the other part is just bad marketing.
I know that advertising, marketing, and such generally have a bad reputation, but good branding, marketing, and advertising is essential to inform consumers about what they are buying (and thereby ensuring satisfied customers in the end), to differentiate products in the market, and to maintain a high quality brand image.
Honestly, at this point, it would be better for Apple to just say: we have iPhone, bigger iPhone, more powerful iPhone, and both bigger and more powerful iPhone. On the CPU/GPU front, they should just list the core count and the iteration number. For Intel and AMD, the better route would be listing the specifications and creating a short hand for these specifications for those who are well versed in their product offerings, but then re-introducing things like "Pentium" or "Athlon" for the highest volume SKUs that people are most likely to buy.
As regards the original article. Everyone already mentioned the TDP issue with the comparison between AMD and Apple, but Apple is also going to be introducing the M3. The rumor mill has it that the M3 will be a far more powerful product than M1 and M2, but this remains to be seen.
>Honestly, at this point, it would be better for Apple to just say: we have iPhone, bigger iPhone, more powerful iPhone, and both bigger and more powerful iPhone.
They should call them: iPhone, iPhone Plus, iPhone Plus/2, iPhone Plus/3, iPhone Plus/4
Eh before the M1s I really disliked macs, and cursed that I had to use some at work. Even now I prefer AMD for desktop. But the unparallel energy efficiency of apple silicon for laptops is just a fact, and I will believe that AMD or intel can challenge apple in laptops when I see it, because till now they are hopelessly far away.
I notice the article only has base power draw in it. I’d want to know max TDP before anything gets crowned. This AMD is 15 watt base, the M2 is about 20 watts.
When the M1 came out, Intel announced chips that would crush and crush it they did, while drawing significantly more power. Intel announced with a 45 watt base TDP which didn’t look too far off the M1 Max, but the 12th Intels ended up peaking at 115 watts for the 12900HK.
It basically came down to if you were looking for something portable but would be plugged in when in use, the Intel offering was superior. If you wanted battery life, the Apple product remained superior.
Competition is good and hopefully AMD remains power competitive, but given a 7800 runs around 88 watts in normal use conditions and has 120 watt max budget, I would expect that performance AMD is touting to come with higher power draw.
Sadly at about about half the Geekbench 5 score. M2 is single core 2061, multicore 15281. 6850U has single 1469 and multi 7365 scores. M2 is also ahead in Cinebench R23 and splits the single and multi in R20 with the 6850U.
To go from that to 70% ahead on multicore would be amazing because it would force Intel and Apple to respond.
The cynic in me, having watched Intel and AMD fight for the desktop crown post Ryzen launch though is skeptical that the scores are some combo pushing power limits to rather undesirable levels and extremely selective benchmarking (It's 70% better at this one thing).
My understanding is that the only reason dGPUs generally don’t use system RAM isn’t that they don’t have access, but that the supporting software doesn’t use it out of choice because the of the speed hit (iGPUs, of course, do generally use system RAM as VRAM.)
> My understanding is that the only reason dGPUs generally don’t use system RAM isn’t that they don’t have access, but that the supporting software doesn’t use it out of choice because the of the speed hit
It looks like one of the advantages of Apples put-everything-on-the-package strategy is that they can have a very wide bus to ram, which makes using system ram for the GPU much more palatable.
Isn't the main difference that Apple is building something like 4-16 channel memory controllers in a mobile device while you normally don't get more than 2-4 channels even on a desktop? That's a lot of transistors and (potentially) power usage but if you can get on the latest node and have a market willing to pay for giant chips it lets you get impressive amounts of bandwidth.
I don't think you need the RAM to be on the same package for that, it just makes the timings easier.
Yes, I am not exaggerating. The Framework came to around $1.8k and the Macbook Pro came to ~ $3.9k. The only compromise with the Framework is that it is a 13" screen and the Macbook pro is a 14". But that's because Apple only offers up to 24GB ram on the 13".
The "Apple tax" isn't really a thing for midrange specs any more, but they will absolutely charge you out the nose to max things out. The top-end macs are targeted at people who won't even notice if the laptop's price is doubled.
We notice… but grumble and pay it because we’ve justified it to ourselves. There are of course outliers with heaps of money that don’t care about the cost, but few businesses are so cavalier with cashflow as to not notice and check why someone is buying a $7k laptop or $25k desktop
They charge you out the nose even to do small upgrades. $200 for bumping the SSD from 256GB to 512GB and $200 to go to 16GB RAM. You can easily find brand name good PCIe Gen 4 2TB NVMe drives for less than $200 now.
Don’t forget they just announced Z1 for Steam Deck like machines (and Aerith for steam deck). They’re certainly not not in the game for low TDP. AMDs little GPUs are also much more capable at gaming because they get the decades of AMD experience and software in that space. M1 can maybe run a few games
There's no way AMD is going to make up that much ground in such a short amount of time. You will end up being correct in that AMD will be faster in one or two cherry-picked benchmarks. I will add these benchmarks will have zero relevance to anything regarding actual CPU power, and AMD will be greatly lacking in real performance.
> There's no way AMD is going to make up that much ground
What facts do you base this assertion on? Do you have insider access to the benchmark results? Maybe wait for the benchmark results before coming to Apple's rescue. Remember that it wasn't so long ago that AMD schooled Intel so hard that their CEO fell off.
The AMD comes ahead in all benchmarks available on that site: https://www.cpu-monkey.com/en/compare_cpu-amd_ryzen_7_7840u-.... I'm not sure where it would fit in terms of perf-per-joule, but there seems to be a very limited amount of benchmarks available (e.g. there seems to be no benchmark for Blender for the AMD, which would be one of the more telling results). The jury is still out for this.
There's already Geekbench 5 scores of the 8 core 7840U up from devices like the Aokzoe A1 Pro with scores around 1900 in single core and 10K in multicore. The M2 scores mentioned above are for the 12 core M2 Max. The Aokzoe is a Steam Deck like handheld too so a larger chassis might allow for a little more out of the chip.
I confess to not really pay attention to the Apple ecosystem. Does anyone know enough about these benchmarks and chips to really know what is going on here? Is this straight up a difference in conventional superscalar/multicore throughput and efficiency?
Or are is there an aspect of Apple to non-Apple comparison here with the compiler, OS, and library ecosystem? I.e. do any of the benchmark calculations get executed off CPU in the GPU or other "matrix" coprocessing units in the Apple SoC? I am only interested in comparing CPU to CPU or GPU to GPU, but would want to exclude software-based differences like different compilers or support libraries.
The M1 has a lot of ALUs, an 8-wide instruction decoder, and a huge reorder buffer. It can decode and execute a lot of instructions on every clock and plenty of space to execute them in an advantageous order. Ignoring power draw, Intel and AMD beat the Mx then they hit higher clock speeds, the Mx top out at 3GHz, and/or pack in more P-cores.
I clearly need to invest in a new NUC. I have some anemic Celeron build that supposedly draws 10-20 W, but the performance is laughably bad. If these "real" chips have such low max draws, they should be able to fit within my informal <20W idle envelope, but still provide performance when needed.
Even max TDP doesn't correlate well to energy efficiency in normal usage. Like, this chip could never use more than 15W and still be less energy efficient, if lightly threaded workloads clock up to drain the full 15W.
Which a boost clock of 5.1GHz strongly suggests might happen.
No way! That would be too expensive for us! To be able to really "crush" Apple, vendors MUST charge the same amount of money for RAM upgrade, and 8GB -> 128GB would be $3000!
> My M1 Air lasts 12+ hours and M1 Pro lasts 10+ hours
Neither of you have said anything about how you spend your time on the laptop itself, which makes both numbers equally useless.
Of course someone who spends their time compiling Rust projects and writing code will have less battery time than someone who writes emails and reads HN most of the day.
But we have no way of knowing if this is what's happening. But my guess is that your usage patterns are different.
My M1 was used for Google video meetings and development, with some firefox thrown in occasionally.
Don't get me wrong - 5 hours is still better than any other laptop I've used, I'm just constantly surprised by people who claim that any competing laptop has to last 8 hours or more.
Yesterday, in response to one of my comments, someone claimed to get regular and consistent 16 hours from their M1.
I really cannot see how that's possible unless they only use it for google searches or reading social media.
I dunno about those tests[1] showing 10h of video meetings, but it's certainly not impossible. At least one other person in this thread reports the same.
Besides, that means that the best case scenario for video only is 10h; I specifically said that I have an IDE running too.
[1] A test using the hardware codecs would certainly do well. I don't have control over what google meets supplies in the way of codecs.
If I forget to plug my M1 Pro laptop in in the morning, it will notify about 10% battery in the afternoon. I'd estimate that's somewhere in range of 4-6 hours. Mostly Android Studio, Teams, Firefox.
So does it mean a passively cooled laptop like MacBook Air?, I mean the M2 performance are good but what makes it attractive are battery life and fanless yet capable laptop.
Maybe fanless is important for people who don't use the laptop for work, but as someone who mainly works with computers and not much leisure, I don't care if it's fanless or not, as long as it doesn't throttle just because something compiles for 3 minutes instead of 20 seconds and uses all the cores to the max. Usually, passively cooled laptops are really poor at that, compared to ones that have active cooling.
A fanless-first design would mean I am not subject to the continual whine of whatever teeny high RPM fans that come standard. As someone stuck on corporate Windows machines, I have 24/7 fan noise from the combination of work-spyware + Teams + Onedrive continually indexing and stomping over each other.
Edit: I rarely do much CPU limited workloads. More RAM heavy than anything else.
The Mac Studio actually has a low RPM always-on fan. It's quiet enough that it can get away with that, and not turning it on and off avoids the noticeable transitions in volume. Under very heavy load the fan speeds up slightly.
Pretty much the only time I hear the M1 fans are when I'm doing something with tensorflow-metal, ffmpeg'ing or rendering video, or otherwise deliberately thrashing the CPU. e.g. Minecraft (albeit modded with Sodium et al) makes it noticeably warm but no fans.
I’ve had an M1 mini for a year now. I think it has a fan, but I’ve literally never heard it, even though it sits right in front of me. Even when using all cores to do video encoding.
Not your fault obviously, but I meant it matters in the sense that part of the problem also lies with the companies and industry as a whole shipping highly inefficient SW and having the solution be "just buy a new MacBook with Mx chip to run a videochat app bro!!11"
Honestly shitty SW should get slapped with an environmental TAX just like motor vehicles. I remember running Skype video calls and Pidgin chats in ~2005 on a 1.2 GHz single core CPU with 512MB of RAM.
Don’t let hardware makers off the hook for bad cooling system design. Or Intel off the hook for pumping more and more power into their processors to compete with AMD.
If you look at Teams when it’s running, it’s not like it’s maxing out a CPU core or anything. It’s just that many laptops are pitifully bad.
Huh. Most corporate windows machines are loaded with crap. The very same hardware behaves much much better with any modern Linux.
That, and apple has a 3 year lead on chip manufacturing side. I moved from M1 MacBook Pro a same size Lenovo with Ryzen 6800h and I cannot tell the difference speed wise. Noise wise, too. It's not even a Zen 4 CPU
The MacBook Air and Pro have the same performance for just over 7 minutes of full power usage / compiling, FWIW. At that point the Air throttles and the Pro never never throttles.
That’s the entire and sole performance differentiator between the Air and the Pro.
I've got an old USB laptop "cooler" which is basically a laptop stand with some integrated fans. If I'm doing something intensive on my MBA it will keep it from throttling for a little while longer than the passive cooling.
I don't necessarily want fanless, but the system being inaudible under moderate load and not a banshee at high load is important. With proper engineering this is possible without too much of a performance hit but a lot of laptop manufacturers either don't put in the time to tune their offerings for this or do what Apple used to and chase thin-at-all-costs even in workstation machines (e.g. Dell Precision laptops being downgraded to close relatives of the XPS line).
Between my personal iMac Pro which is easily the second best-cooled Intel Mac desktop made in the past decade (after the cheesegrater tower) and the employer-provided 16" M1 Pro MPB I use for work I've been spoiled. My tolerance for fan noise has become very low, to the point that I'm debating replacing the case on my Ryzen 5000 gaming tower with something with big slow 200mm fans (like the Fractal Torrent) so its noise level can match that of my iMac and MBP.
I got a beefy Ryzen 3000 personal desktop in 2020 when my 2014 laptop started giving in (noisy even with new thermal paste, lid broken).
I can't stand the noise of the fans when it's next to me, it has a water cooler fan. I recently bought quieter fans to replace the existing ones with, however haven't fitted them due to calling issues.
Just wanted to say that your suggesting of getting big slow fans has intrigued me. Right now the desktop sits in the living room, and I only turn it on at night when I'm in the office and the family's in the bedroom, or out of necessity.
I remote into it, so things like games are out of the picture as it gets really loud when the GPU fans start going.
The size of the fans makes a huge difference, way more than you'd think.
One of my friends has a tower built with a i7 12700k in a Torrent Compact case (2x 200MM fans) which is very quiet on its own, and the giant hunk of metal NH-D15 cooler with a 140MM fan it uses adds almost nothing to that. Its PSU can also run with its fan off most of the time and even when it does spin up isn't too bad.
The 180mm Fractal fans are ... not very good. They're the best 180mm fans (there is not much choice, just Silverstone and some cheap stuff), but pretty shitty compared to the available 120/140mm choices or the 200mm Noctuas (which will not fit cases designed for 180mm fans).
Nice catch, I hadn’t noticed that Torrents came with 180mms instead of 200mms. Even if they’re not amazing they keep my friend’s temps reasonable without making much noise.
The number of cases that will fit 200mms is tiny and none of them are particularly great which is a bummer.
Fanless is important enough for me that when I was about to go home to a hot village over the Nov-Jan break (I was going to work till mid-December), I went and bought the M1 Air so I could ditch my loud work laptop. The work laptop was a Dell with a Quadro, known to overheat and get loud when temp > 45°C. The last time I was home I had observed that it idled at least below 40, while around 35 in the city, so with 10° warmer weather in December, it was going to be loud when idle.
Even when my new job gave me a new device, I hesitated to get the M1 Pro as it has a fan. Only got it after watching enough reviewers say it almost never turns on.
I'm 34 and my ears started constantly ringing in 2019 after my then employer gave me a faulty 'new' laptop whose fan was always at 100%. I complained about it numerous times to no avail.
We didn't fully grasp how loud it was until one day we were in the office past 5pm when the central cooling turner off. It contributed a lot to the office noise.
There were about 10 of us in the office, when I turned off my laptop almost everyone startled, asking what loud object had just turned off (we had been complaining that it's hot, so were aware the HVAC was off).
So yes most people compiling code may prefer beefier machines with fans, I've sadly started a lifelong journey of avoiding fans.
We live in a small apartment, I'm even considering getting a new fridge because our current one is loud.
I'm conflicted on this one. I have really good Sony ones, and whenever I wear them for even 30 minutes, when I take them off I feel worse.
I don't know if it's all in my head and they objectively don't make it worse, however I also now avoid wearing them.
I've probably used them for < 20 hours in the last year, half of that while driving long distance. My car also creates a lot of white noise.
Different tyres will change the noise of a car too.
When I bought my last car, it came with a mishmash of different brands of cheap crappy tyres. They were quite noisy all the time, but when you hit exactly 80km/h there was some kind of resonating drone as well, that was very weird...
When I replaced them with good, high-end softer tyres, the noise was reduced by a massive amount all the time, and no more drone.
Yeah I have read such also, and indeed also for me it is not ideal to wear for long time, but it is still definitely significantly better than having to be listening to a laptop fan or similar. Listening to music with it I feel is better than just using it for noise cancelling, but I cannot be certain it makes a difference. Otherwise, if I do not have such mechanical noises around, I disable noise cancelling.
I've really gotten a lot of milage out of wearing earplugs on a regular basis. It is a bit of an unusual practice, but I really like the quiet. Might be an alternative solution to your problem.
Alternatively, I've found that using a laptop pad with a large (200mm) fan can help to keep the machine cool without generating much noise.
> Sometimes I wear my over ear headphones with nothing playing just to dull it.
If you like the over-the-ear format, you can also get rated[1] earmuffs. I've tried out a set, but I also wear glasses and their clamping force causes my glasses frames to put too much force on my head for comfort.
I've had two MacBook Pros die in the Amazon rainforest due to humidity. I think the fans pull in humid air and that the water gets stuck. My brother has had an M1 Air for about 2.5 years in the jungle with no issues whatsoever. Also since there is no power grid, it's good to have more battery life (sometimes it rains for days).
If anyone reads this, any suggestions on how to make electronics last longer in such a humid environment? I've been thinking about building some sort of dry box with silica gel but I'm not really sure how... And to use an external monitor instead of the built in one...
As someone that works in an open floor plan I _hate_ when my coworkers laptops sound like jet planes because the fans are going full blast. It’s so distracting.
I'm at least happy your career seems to have been limited to indoors office work, much of the professional industry is very difference when it comes to noise. Doesn't mean you're not right, but I wish didn't have experience spending 7 hours / day next to very loud construction and cleaning machines, as it kind of shrews my perspective compared to my current peers it seems.
I use macbook air for work, and I am totally satisfied with it. For heavy stuff I use cloud or desktop computer anyway, having fans working max for hours it was a torture anyway in my older laptop, and portability is quite important for me for work. Yeah a 12c macbook pro max would be more ideal for heavy loads, but if we compare with an i5/ryzen 5 (or even i7/ryzen 7 probably) I would go with the macbook air.
You probably don't realise how fans can be a massive drag on productivity.
I was so mad about myself that I hesitated buying M1 for so long after buying one.
I've become much much much more productive on this machine, as I now longer feel this drag when I have to do something and then brace myself for the fan noise. I didn't realise that but I was putting a lot of tasks away just so that I don't get annoyed by fans.
My M1 Max has fans, but they are so well tuned even if they do turn on sometimes I don't hear them and most of the time they are off.
Given how much the fans are affecting productivity, I really couldn't care less if my compile time took 3 minutes rather than 20 seconds as long if it was completely silent, as I may have been hesitating for a couple of minutes to run the task in the first place.
Also it is such a comfort to have a meeting without having to say "sorry guys, these are my fans going off".
I also never noticed any performance issues. Anything I do is snappy, no hiccups and works regardless if I am plugged in or not.
I had situations with my Dell where I couldn't plug in and the anxiety that it could shut down at any time (battery meter wasn't trustworthy) was painful plus massively worse performance on battery.
Apple is a different league. As much as I hate Apple I see no alternative right now.
I'm not sure what the point here is, if you want an actively cooled laptop, just buy that. Just because you personally need something different it doesn't mean that feature isn't important to other people.
> I'm not sure what the point here is, if you want an actively cooled laptop, just buy that. Just because you personally need something different it doesn't mean that feature isn't important to other people.
The person that you're responding to is essentially expressing this exact same sentiment to the person above them? For many of us, fanlessness is irrelevant; in other words, "if you want a passively cooled laptop, just buy that". Hell, I may be an outlier here, but as far as I'm concerned battery life is also irrelevant, and power consumption only matters to the extent that it increases my power bill. I use laptops because they're easy to carry around when closed, not because I'm coding at the top of Mt. Everest or anywhere else that is far from an outlet.
The person that you're responding to is essentially expressing this exact same sentiment to the person above them?
I'm glad you asked. The answer is that no they are not. They are saying 'it can do this'. The person I responded to is saying 'I don't personally want that'. These are two separate things.
If you said 'tell me about this truck' and someone said 'this truck is good for towing boats' and you said 'I don't have a boat', they would say 'you asked and I explained it'. Why are you acting like a news article is trying to sell you personally on something?
>but as far as I'm concerned battery life is also irrelevant, and power consumption only matters to the extent that it increases my power bill. I use laptops because they're easy to carry around when closed
I'm somewhat in the same boat, but I sometimes use mine when traveling (without being plugged in), but fairly rarely.
However, another thing I really like about having a laptop is that my current one, at least, is very very quiet. Years and years of having a noisy desktop PC with loud fans really made me hate fan noise. With a modern laptop, I can do normal computing tasks and almost never hear the fan (and when I do, it's hard to hear), and I don't have to build some custom, non-portable liquid-cooled machine to get that.
I use it professionally and hate the fan, but in experience the MacBook Air M1 holds up well to sustained (Rust) compilations. None of the infuriating bursting behavior of old Intel chips.
Fabless is nice, and the air can do a good job, but the real kicker is the non-airs also just generally don’t make the fans kick in - even building massive amounts of code doesn’t seem to do it. As per usual the real “I need a fan” killer is the GPU, which is to me the least compelling bit (because I’m not super interested in gaming on laptops in general)
It depends on the work. A lot of people just do web development where either they are fiddling around with APIs or interacting with some VM over SSH or are just creating presentations and reports. Even if you are working with a massive codebase, there are tools to just selectively run the tests that are relevant to the changes you made before pushing it all off to a CI/CD pipeline.
Even if fanless is a luxury, a lot of professional developers can just make do with something with the power of a 2016 Macbook Air. Then the nice-to-use screen, keyboard, trackpad etc are extras worth having, much more than a buzzing fan.
So in the context of an article about this AMD chip, you instantly downgrade M2 to merely "good" while extolling the virtues of the cooling and battery life?
I just love imagining the Asahi Linux folks making a blood oath to reverse engineer these machines so that they may one day enjoy the peace and quiet of a laptop with no fan.
Well from the perspective of an engineer: I get and respect all the hard work that went to making the the M2 and generally ARM revolution that Apple had unleashed.
From the perspective of average joe: yeah I hear about performance every new device announcement can we do something about the shitty battery life and loud fans on non apple Machine my phone can do more with less power and noise. (For real though, I just want an M1 iPad Pro like device that runs Windows)
So yeah I guess yeah excluding battery life and cooling ,I don't really give a shit about Apple silicon TBH, and I think I'm not the only one.
> If you think 8gb of ram is enough, which it isn't, especially on macOS.
Depends what you do, and for many people you'd be wrong.
I've been using an 8GB M1 Air for a couple of years and never had an issue caused by memory constraints.
And that's with multiple instances of Rider and GoLand, VS Mac, VS Code, dozens of Brave tabs, multiple terminal sessions, locally-running Postgres, MS Word, Calibre, and more running simultaneously on a cold and silent machine. Also AI art generation using Stable Diffusion models, 768x768 rendering, and 4x upscaling, without issue. Plus running Parallels with Windows 11 and Visual Studio (ARM) was totally acceptable.
The only thing I haven't tried, and that I expect it would choke on, is multiple VMs or loads of containers.
It's amazing what you can do with 8GB RAM. I'd echo your original especially on macOS but to show the exact opposite, as my 16GB 10th Gen i5 Windows 11 machine lags with a much smaller workload whilst heating the house.
The thing to crush about the M chips is performance per watt, not raw performance. They are respectable chips performance wise but Intel and AMD have faster (but hotter) chips. But as far as I know there is still nothing that beats them for efficiency, at least for the high performance desktop and laptop oriented market.
As long as you don't start counting watt/performance/price, as then usually AMD wins out as Apple does have "premium" pricing for most (if not all) of their products
As I said elsewhere (https://news.ycombinator.com/item?id=35854112), if I could chose between a fanless M2 laptop that might throttle, and a M2 laptop with fans that would never throttle, I know for sure what I and many other developers would go for.
The only time I’ve heard the fans on my M1 Max is running ML stuff (Demucs) across every core for several minutes, so the CPU was completely maxed out. Even then it was quiet. Day to day for me it hardly ever gets warm and I never hear the fans.
"Cinebench R23 Multi" and "Cinebench R23 Single" are the software workloads used (multi-thread and single-thread). "Package Power" specifies (approximately) where the power is measured and what components are included—but reviewers like NotebookCheck aren't equipped to actually measure power at the chip or package level, so they're actually relying on software estimates that add another layer of vendor-specific sources of error on top of the inconsistencies about what's in the package (eg. Apple puts DRAM in the package but Intel and AMD don't).
If the P-cores are where the real workload runs, then the efficiency of the P-cores obviously is paramount. P-cores aren't just for running off wall power.
Are you assuming that P-cores only get powered on after E-cores are fully saturated?
I only meant "real workload" in the sense that when real people are using their computer, the P-cores are doing real work; that's true even for bursty workloads that don't keep the processor 100% busy, so power is being spent by the P-cores and their efficiency matters.
> Are you assuming that P-cores only get powered on after E-cores are fully saturated?
No, that's not how M1/M2 work.
Apps are generally on the P-cores, but daemons are on E-cores, and the daemons are doing essential work like keeping you connected to wifi. And the apps aren't running all the time if you're just browsing or editing text. It's all part of the end goal of having a lot of battery life.
The tdp for the m2 is 20 watts. The article mentions a 15 watt tdp for the AMD chip. We will have to wait to see when the chips are released but it appears performance per watt will be at a minimum competitive with the m2.
I’d maintain a certain amount of scepticism until independent benchmarks are out. Comparison by tdp, particularly between manufacturers, tends to be… less than useful; if nothing else there is no common definition of tdp.
As mentioned elsewhere in the comments, there was an Intel chip that appeared to beat the M1 on performance per watt if you looked at ‘tdp’, the issue being that Intel was using a particularly eccentric definition of tdp.
By then you can run your own benchmarks, which was the point of “until independent benchmarks are out”. You need some skepticism when discussing characteristics they give for an upcoming chip.
The point is, fundamentally, that you shouldn't take a manufacturer's word on real-world behaviour, particularly for something that isn't even released yet. Like, you can if you want, but going on history you will nearly always be disappointed.
Sure, and when the chip is available and that test can be done independently, that’ll largely answer the question. For now, it’s pretty much just marketing claims.
I don't have a citation on hand (though I love GamersNexus, their focus on video makes it hard to find citations for such things), iirc TDP is actually not comparable between processor generations, let alone manufacturers. Intel and AMD calculate it differently, and AMD's calculation includes the thermal solution attached to the chip, so hypothetically they can make it almost whatever number they want.
Right, that was the first thing I looked at and I don’t understand the article author’s logic: compare performance to M2, then don’t compare the 25% lower power (admittedly stated) consumption.
Just a super weird writing choice.
Also puts to bed the “x86 is fundamentally unable to be as low power as arm” rationale for the poor intel and amd perf/watt. I think the sole reason for that is that they never _really_ competed on that metric.
Does anyone know the state of linux scheduling and efficiency cores? It is my understanding that part of what makes the Apple experience smooth is tight control over what processes get pinned to performance cores and enforcing background jobs remain as such.
For example, the Gnome indexer, tracker-miner, should only ever be allocated to background status.
Not really surprising. The m2 is just the same concept as the amd apu, but with sleeker branding and marketing. Amd has been doing apu since 2011, ages longer than apple
The whole m2 hootenany on hn reminded me of when retina came out. Its just a fancy branding for a hidpi display. Nothing inherently special.
You can't get HiDPI displays from anyone else, and nobody else can do the software right because they insist on trying to support arbitrary resolution instead of just 2x.
As far as desktop monitors go, there's ye olde LG Ultrafine 5K, and that's about your lot. There used to be a Dell, too, but, as is standard with exotic Dell monitors, they rapidly discontinued it.
HW exists of course, but the software is another story. With macOS you have zero issues. I've been using a 4k setup for probably 6 years now, in addition to my retina MBP's.
Seriously, I do not doubt the performance (I hope they do not melt their cpus and motherboards this time), but energy efficiency, which is important in laptops? I will believe it when I see it (by in depth reviews, not by what benchmarks companies release), considering the desktop ryzen 7000s got to a completely different direction. Intel had made some ridiculous energy efficiency claims in their last generation that were totally out of reality in practice.
The desktop Ryzen 7000 parts are still saddled with power-hungry die-to-die interconnects and an IO die that's built on an older fab process (though not quite as old as previous Ryzen IO dies), so the idle power is horrible. And the parts without the 3D stacked L3 cache will happily crank up the voltage to crazy levels in order to get the last few percent of clock speed. But if you cap their power/voltage/frequency to reasonable levels and look at the power consumed by just the CPU cores, it's clear that AMD has a very promising building block to use for their monolithic mobile processors.
But I agree that AMD's marketing claims as filtered through low-quality journalism that doesn't even distinguish between peak and base power consumption is a poor way to judge whether AMD has succeeded in building a great low-power processor.
Well I do not take any claims of any company seriously before I see actual reviews anyway, they usually just cherry pick what they want, so I do not specifically blame AMD for that.
And yeah I agree, we run our new AMD desktops in eco mode at work, it was ridiculous to routinely get like 95C temperatures for whatever random chores. Once we used to overclock AMD cpus, nowadays we underclock them lol. I have not really checked how many watts it draws after we put them in eco mode though, I will do it next time.
Maybe not so implausible? AMD had a good run with the Athlon/K7 and the K8. The latter was so successful that Intel was forced to adopt its 64-bit ISA. The Pentium 4 (NetBurst) was a dead end design, and Intel had to backtrack and build off their laptop CPUs that were based on the older Pentium Pro design.
Apple bought P.A. Semi in 2008, a company founded by Daniel Dobberpuhl of DEC Alpha fame. Clearly they had something brewing on the CPU front.
Apple "clearly had something brewing on the CPU front" sure, they were just getting into the mobile phone industry; hearing that Apple would eventually make their own smartphone chips wouldn't have been surprising.
What might've been surprising is them getting into the big, high performance desktop-class CPU segment and leading the industry in efficiency and performance HEDT chips for desktops.
And Intel's relative irrelevance is certainly surprising, as is them losing the fab game to TSMC.
> The Pentium 4 (NetBurst) was a dead end design, and Intel had to backtrack and build off their laptop CPUs that were based on the older Pentium Pro design.
Don't forget Itanium, Intel's other dead end architecture. It's actually kind of amazing that Intel recovered from both at the same time (for the most part).
In what way were they collecting monopoly rents? It seems like what happened was someone came out with a better product and took a bunch of their business. Isn't that exactly what monopolies are expected to prevent from happening, and, as I understand it, the reason they are considered so bad?
The Commission filed an administrative complaint against Intel Corp., the world’s leading computer chip maker, charging that the company had illegally used its dominant market position for a decade to stifle competition and strengthen its monopoly. The complaint alleged that Intel engaged in a course of conduct to shut out rivals’ competing microchips by cutting off their access to the marketplace. In particular, the complaint alleged that Intel unlawfully maintained its monopoly in relevant central processing unit, or CPU, markets, and sought to acquire a second monopoly in the relevant graphics markets, using a variety of unfair methods of competition. In August of 2010, Intel agreed to a settlement containing provisions that would undo the effects of Intel's past conduct, and prohibiting Intel from suppressing competition in the future.
You should probably include Intel in that list, given that they acquired DEC's chip fabs in 1997[1], the lion's share of Alpha IP in 2001[2], and hundreds of Alpha engineers over the next few years[3].
Really? IIRC the PPC970 (you could definitely debate to what extent this was Apple vs IBM, of course) and Athlon64 both handily beat Intel's stuff when they launched ~2003; it's just that they failed to maintain that lead.
I want this as a server. It has 20 PCIe lanes (not sure whether the integrated GPU uses any of them), so getting 4x NVMe plus 10 or even 25 GigE should be possible. Bonus points for a server platform like this that still has a battery.
The previous generation of laptops from Apple were so bad that I'll be vice-like gripping my M1 laptop hard before even looking at the benchmarks. Performance alone is such a bad metric to even mention cause I just think about being 15% throttled 99% of the time, having a huge WHIRRRRRR and only lasting about 1 hour.
I can't believe how much we've been focused on core speed when memory bandwidth and latency has been such a huge issue... at least for my workloads. My stuff screams on a M1 Chip.
What exactly are you asking for here? Because you've already been informed that AMD and Intel systems using LPDDR5 at the same speed and bus width are now available. On paper, every new laptop using soldered LPDDR is now over 100GB/s for DRAM bandwidth. Are you asking for synthetic benchmark results comparing achievable throughput against theoretical bandwidth, or are you just being obtuse?
I am asking for a name of one that has been demonstrated to actually do that. Not promise. Not “on paper”. Actually do. Extra credit if it actually exists in a store to actually buy. Today.
There's nothing nonsensical about the claim that there are AMD and Intel systems using 128-bit LPDDR5-6400. AMD's mobile processors started supporting that config last year, so in spite of this year's mobile processors being a bit late the market is already full of systems so equipped. Intel's mobile processors as of earlier this year also support this memory configuration and speed, and systems have been shipping from multiple vendors for well over a month; they're not exotic or hard to find. There's even at least one announced but not yet shipping Intel-based system that will use LPDDR5-7500!
It's called "overfitting specs". They ask for a very specific set of specs that only Apple laptops have and ignore anything else. If you can find anything that competes they'll move the goalposts and complain about touchpad or something else.
Sorry, still not clear exactly where your goalposts are. Is there a public benchmark where M2 scores above 100GB/s that you'd like to see comparable runs for a recent x86 mobile platform?
For whatever reason, it isn't so easy to find real-world memory bandwidth results for consumer stuff; it's a fairly standard benchmark for server hardware, but less so for consumer. Here's one for an old Ryzen 9: https://www.anandtech.com/show/14605/the-and-ryzen-3700x-390..., but couldn't find anything current offhand.
What kind of work are you doing to need that on a laptop of all things? The usage cases I am aware of are all better suited by non-portable computers dedicated to the task.
That sounds incredibly specialized and I am needlessly curious.
AMD should really stop with this crappy marketing and look at how many of their promises evaporated quickly once people tried their new products. Radeon fiascos all the time, alienating Threadripper users by false promises, blown up Ryzen 7k3Ds running on voltages from marketing materials, crappy x3D hybrid CCD architecture disadvantaging many workloads etc. Somebody should tell their marketing team something unpleasant but needed.
Marketing benchmarks used values that led to blown up 3Ds. Mobo fixes might lead to performance regressions (i.e. no longer the best CPU for a given game).
255 comments
[ 4.9 ms ] story [ 300 ms ] threadCompetition is great.
Clickbait is not.
Does Apple upset you, yeah?
> and a whopping 72% better multiprocessing performance than the standard M2
More than it used to.
“ the Ryzen 7 7840U, the initial flagship APU of the new series, apparently offers 9% better 3D rendering performance, 14% better responsiveness, and a whopping 72% better multiprocessing performance than the standard M2
There will be four new chips: the aforementioned Ryzen 7 7840U, the midrange Ryzen 5 7640U and 7540U, and the affordable Ryzen 3 7440U “
Also they’re not going to win on the marketing front with that naming strategy. Just awful. M1/2 and Pro/Max are so much simpler. It’s like Intel’s naming, such a mess.
Intel’s latest move is just absurd.
An extremely narrow range of SKU’s in a product line with an extremely short history makes that easier.
Its a problem for simplicity of naming, its not a problem beyond that (and the alternative is a problem if you aren’t a vertically integrated producer supporting a fairly narrow end-user product line.)
Apple's names are pretty awful, I get Pro/Max/Ultra confused and they aren't even unique identifiers either. Which M2 Pro? 10 core or 12 core? Why do CPUs with the same name have different core counts?
Apple has an easier time with processor names because they have a single customer that wants a simple product lineup. Amd and Intel sell their processors to OEMs that want to have products for every budget, so you end up with lots of processor skus and lots of laptop skus and have to do a bit of research to understand what you would be buying and have to decide what you want. On the plus side, you have lots of choices, on the minus side, you have lots of choices.
Simple processor naming doesn't seem to stop Apple from naming every iPad 'the new iPad' though. But then again, other OEMs like to use the same product name and number for 3-7 different devices these days too, so everything is horrible, as usual.
Is the M1 Ultra faster than the M2 Pro? Is the M1 Max faster than the Ultra, or is the Max not the maximum? How does the M1 Pro compare to the M3?
AMD generally uses the "bigger number within a generation is faster" and Apple uses some confusing adjectives. Either way the names only give a vague indication of performance.
The naming of products has gotten really silly and useless, and I put this blame on Intel and Apple. Apple started down the path of Pro, Pro Max, Ultra, Max, and Plus. Remember the iPhone XS? Idiotic naming there. Intel, otoh, just started doing crazy numeric strings with a random character added to the end. And this was after bad naming schemes like "Core 2 Quad". Part of the issue with GPUs and CPUs is the need to do binning to make IC manufacturing commercially viable so companies end up with a ton of SKUs, but the other part is just bad marketing.
I know that advertising, marketing, and such generally have a bad reputation, but good branding, marketing, and advertising is essential to inform consumers about what they are buying (and thereby ensuring satisfied customers in the end), to differentiate products in the market, and to maintain a high quality brand image.
Honestly, at this point, it would be better for Apple to just say: we have iPhone, bigger iPhone, more powerful iPhone, and both bigger and more powerful iPhone. On the CPU/GPU front, they should just list the core count and the iteration number. For Intel and AMD, the better route would be listing the specifications and creating a short hand for these specifications for those who are well versed in their product offerings, but then re-introducing things like "Pentium" or "Athlon" for the highest volume SKUs that people are most likely to buy.
As regards the original article. Everyone already mentioned the TDP issue with the comparison between AMD and Apple, but Apple is also going to be introducing the M3. The rumor mill has it that the M3 will be a far more powerful product than M1 and M2, but this remains to be seen.
They should call them: iPhone, iPhone Plus, iPhone Plus/2, iPhone Plus/3, iPhone Plus/4
My M1 is great.
RISC-V is inevitable.
All I find are comparisons with the desktop parts.
When the M1 came out, Intel announced chips that would crush and crush it they did, while drawing significantly more power. Intel announced with a 45 watt base TDP which didn’t look too far off the M1 Max, but the 12th Intels ended up peaking at 115 watts for the 12900HK.
It basically came down to if you were looking for something portable but would be plugged in when in use, the Intel offering was superior. If you wanted battery life, the Apple product remained superior.
Competition is good and hopefully AMD remains power competitive, but given a 7800 runs around 88 watts in normal use conditions and has 120 watt max budget, I would expect that performance AMD is touting to come with higher power draw.
To go from that to 70% ahead on multicore would be amazing because it would force Intel and Apple to respond.
The cynic in me, having watched Intel and AMD fight for the desktop crown post Ryzen launch though is skeptical that the scores are some combo pushing power limits to rather undesirable levels and extremely selective benchmarking (It's 70% better at this one thing).
It looks like one of the advantages of Apples put-everything-on-the-package strategy is that they can have a very wide bus to ram, which makes using system ram for the GPU much more palatable.
I don't think you need the RAM to be on the same package for that, it just makes the timings easier.
What facts do you base this assertion on? Do you have insider access to the benchmark results? Maybe wait for the benchmark results before coming to Apple's rescue. Remember that it wasn't so long ago that AMD schooled Intel so hard that their CEO fell off.
Or are is there an aspect of Apple to non-Apple comparison here with the compiler, OS, and library ecosystem? I.e. do any of the benchmark calculations get executed off CPU in the GPU or other "matrix" coprocessing units in the Apple SoC? I am only interested in comparing CPU to CPU or GPU to GPU, but would want to exclude software-based differences like different compilers or support libraries.
Idle at 6-7W, but a full-powered multi-core CPU when you need it. Plus USB4 and 32 or 64GB RAM.
Which a boost clock of 5.1GHz strongly suggests might happen.
the same 8 hours of battery life, or more
above 128gb RAM at the same speeds or better
and then equivalent computational performance or better
No way! That would be too expensive for us! To be able to really "crush" Apple, vendors MUST charge the same amount of money for RAM upgrade, and 8GB -> 128GB would be $3000!
It already has better computational power
Why? My new M1 did maybe 5 hours on the battery. Got worse as it aged over a year.
> My M1 Air lasts 12+ hours and M1 Pro lasts 10+ hours
Neither of you have said anything about how you spend your time on the laptop itself, which makes both numbers equally useless.
Of course someone who spends their time compiling Rust projects and writing code will have less battery time than someone who writes emails and reads HN most of the day.
But we have no way of knowing if this is what's happening. But my guess is that your usage patterns are different.
Don't get me wrong - 5 hours is still better than any other laptop I've used, I'm just constantly surprised by people who claim that any competing laptop has to last 8 hours or more.
Yesterday, in response to one of my comments, someone claimed to get regular and consistent 16 hours from their M1.
I really cannot see how that's possible unless they only use it for google searches or reading social media.
I dunno about those tests[1] showing 10h of video meetings, but it's certainly not impossible. At least one other person in this thread reports the same.
Besides, that means that the best case scenario for video only is 10h; I specifically said that I have an IDE running too.
[1] A test using the hardware codecs would certainly do well. I don't have control over what google meets supplies in the way of codecs.
I'm mostly doing web dev in VS Code and Chrome, with a language server plugin doing background work.
One difference with the M1 Pro is that I have the Slack app installed on it, and it's always shown in the "using significant battery" list.
100% CPU usage on all cores during 4-6 hours is descent.
Simply developing in IntelliJ doesn't even max out 1 core, nevermind all cores.
Edit: I rarely do much CPU limited workloads. More RAM heavy than anything else.
Even then, it's still quiet enough for the person next to me to barely notice.
Heat is also a downside when you have to carry it between meeting rooms.
A silent/cool laptop is highly preferred over something that does maybe 5% faster.
In the case of Apple’s M series you get both. Other laptops (gram) may be lighter but the silent/cool/fast is a winning combo for me.
How much of that is the problem of the SW?
Honestly shitty SW should get slapped with an environmental TAX just like motor vehicles. I remember running Skype video calls and Pidgin chats in ~2005 on a 1.2 GHz single core CPU with 512MB of RAM.
If you look at Teams when it’s running, it’s not like it’s maxing out a CPU core or anything. It’s just that many laptops are pitifully bad.
That, and apple has a 3 year lead on chip manufacturing side. I moved from M1 MacBook Pro a same size Lenovo with Ryzen 6800h and I cannot tell the difference speed wise. Noise wise, too. It's not even a Zen 4 CPU
I can exert some influence in my choice of hardware.
That’s the entire and sole performance differentiator between the Air and the Pro.
Between my personal iMac Pro which is easily the second best-cooled Intel Mac desktop made in the past decade (after the cheesegrater tower) and the employer-provided 16" M1 Pro MPB I use for work I've been spoiled. My tolerance for fan noise has become very low, to the point that I'm debating replacing the case on my Ryzen 5000 gaming tower with something with big slow 200mm fans (like the Fractal Torrent) so its noise level can match that of my iMac and MBP.
I can't stand the noise of the fans when it's next to me, it has a water cooler fan. I recently bought quieter fans to replace the existing ones with, however haven't fitted them due to calling issues.
Just wanted to say that your suggesting of getting big slow fans has intrigued me. Right now the desktop sits in the living room, and I only turn it on at night when I'm in the office and the family's in the bedroom, or out of necessity.
I remote into it, so things like games are out of the picture as it gets really loud when the GPU fans start going.
One of my friends has a tower built with a i7 12700k in a Torrent Compact case (2x 200MM fans) which is very quiet on its own, and the giant hunk of metal NH-D15 cooler with a 140MM fan it uses adds almost nothing to that. Its PSU can also run with its fan off most of the time and even when it does spin up isn't too bad.
Semi-passive PSU is kinda table stakes nowadays.
The number of cases that will fit 200mms is tiny and none of them are particularly great which is a bummer.
Fanless is important enough for me that when I was about to go home to a hot village over the Nov-Jan break (I was going to work till mid-December), I went and bought the M1 Air so I could ditch my loud work laptop. The work laptop was a Dell with a Quadro, known to overheat and get loud when temp > 45°C. The last time I was home I had observed that it idled at least below 40, while around 35 in the city, so with 10° warmer weather in December, it was going to be loud when idle.
Even when my new job gave me a new device, I hesitated to get the M1 Pro as it has a fan. Only got it after watching enough reviewers say it almost never turns on.
I'm 34 and my ears started constantly ringing in 2019 after my then employer gave me a faulty 'new' laptop whose fan was always at 100%. I complained about it numerous times to no avail.
We didn't fully grasp how loud it was until one day we were in the office past 5pm when the central cooling turner off. It contributed a lot to the office noise.
There were about 10 of us in the office, when I turned off my laptop almost everyone startled, asking what loud object had just turned off (we had been complaining that it's hot, so were aware the HVAC was off).
So yes most people compiling code may prefer beefier machines with fans, I've sadly started a lifelong journey of avoiding fans.
We live in a small apartment, I'm even considering getting a new fridge because our current one is loud.
A fanless laptop has been a revelation for me. I am not going back ever.
I've probably used them for < 20 hours in the last year, half of that while driving long distance. My car also creates a lot of white noise.
Different tyres will change the noise of a car too.
When I bought my last car, it came with a mishmash of different brands of cheap crappy tyres. They were quite noisy all the time, but when you hit exactly 80km/h there was some kind of resonating drone as well, that was very weird...
When I replaced them with good, high-end softer tyres, the noise was reduced by a massive amount all the time, and no more drone.
Alternatively, I've found that using a laptop pad with a large (200mm) fan can help to keep the machine cool without generating much noise.
I, on the other hand, find the noise overwhelming sometimes. Sometimes I wear my over ear headphones with nothing playing just to dull it.
Same idea, I suppose? Maybe I'll try some ear plugs.
If you like the over-the-ear format, you can also get rated[1] earmuffs. I've tried out a set, but I also wear glasses and their clamping force causes my glasses frames to put too much force on my head for comfort.
---
1. By rated, I mean they have an "NRR" rating.
If anyone reads this, any suggestions on how to make electronics last longer in such a humid environment? I've been thinking about building some sort of dry box with silica gel but I'm not really sure how... And to use an external monitor instead of the built in one...
I was so mad about myself that I hesitated buying M1 for so long after buying one.
I've become much much much more productive on this machine, as I now longer feel this drag when I have to do something and then brace myself for the fan noise. I didn't realise that but I was putting a lot of tasks away just so that I don't get annoyed by fans.
My M1 Max has fans, but they are so well tuned even if they do turn on sometimes I don't hear them and most of the time they are off.
Given how much the fans are affecting productivity, I really couldn't care less if my compile time took 3 minutes rather than 20 seconds as long if it was completely silent, as I may have been hesitating for a couple of minutes to run the task in the first place.
Also it is such a comfort to have a meeting without having to say "sorry guys, these are my fans going off".
I also never noticed any performance issues. Anything I do is snappy, no hiccups and works regardless if I am plugged in or not.
I had situations with my Dell where I couldn't plug in and the anxiety that it could shut down at any time (battery meter wasn't trustworthy) was painful plus massively worse performance on battery.
Apple is a different league. As much as I hate Apple I see no alternative right now.
The person that you're responding to is essentially expressing this exact same sentiment to the person above them? For many of us, fanlessness is irrelevant; in other words, "if you want a passively cooled laptop, just buy that". Hell, I may be an outlier here, but as far as I'm concerned battery life is also irrelevant, and power consumption only matters to the extent that it increases my power bill. I use laptops because they're easy to carry around when closed, not because I'm coding at the top of Mt. Everest or anywhere else that is far from an outlet.
I'm glad you asked. The answer is that no they are not. They are saying 'it can do this'. The person I responded to is saying 'I don't personally want that'. These are two separate things.
If you said 'tell me about this truck' and someone said 'this truck is good for towing boats' and you said 'I don't have a boat', they would say 'you asked and I explained it'. Why are you acting like a news article is trying to sell you personally on something?
I'm somewhat in the same boat, but I sometimes use mine when traveling (without being plugged in), but fairly rarely.
However, another thing I really like about having a laptop is that my current one, at least, is very very quiet. Years and years of having a noisy desktop PC with loud fans really made me hate fan noise. With a modern laptop, I can do normal computing tasks and almost never hear the fan (and when I do, it's hard to hear), and I don't have to build some custom, non-portable liquid-cooled machine to get that.
I'm 90% sure that most laptops do not include a fab :D (though maybe some are fabulous [sparkle sparkle glitter])
Even if fanless is a luxury, a lot of professional developers can just make do with something with the power of a 2016 Macbook Air. Then the nice-to-use screen, keyboard, trackpad etc are extras worth having, much more than a buzzing fan.
I just love imagining the Asahi Linux folks making a blood oath to reverse engineer these machines so that they may one day enjoy the peace and quiet of a laptop with no fan.
From the perspective of average joe: yeah I hear about performance every new device announcement can we do something about the shitty battery life and loud fans on non apple Machine my phone can do more with less power and noise. (For real though, I just want an M1 iPad Pro like device that runs Windows)
So yeah I guess yeah excluding battery life and cooling ,I don't really give a shit about Apple silicon TBH, and I think I'm not the only one.
Excited to see the competition though!
I got more in mind, but if your computing needs are checking your email and editing word files, it’s fine.
Depends what you do, and for many people you'd be wrong.
I've been using an 8GB M1 Air for a couple of years and never had an issue caused by memory constraints.
And that's with multiple instances of Rider and GoLand, VS Mac, VS Code, dozens of Brave tabs, multiple terminal sessions, locally-running Postgres, MS Word, Calibre, and more running simultaneously on a cold and silent machine. Also AI art generation using Stable Diffusion models, 768x768 rendering, and 4x upscaling, without issue. Plus running Parallels with Windows 11 and Visual Studio (ARM) was totally acceptable.
The only thing I haven't tried, and that I expect it would choke on, is multiple VMs or loads of containers.
It's amazing what you can do with 8GB RAM. I'd echo your original especially on macOS but to show the exact opposite, as my 16GB 10th Gen i5 Windows 11 machine lags with a much smaller workload whilst heating the house.
But with almost all non-Apple laptops running spyware-laden Windows 11, just how possible is this?
Of course, running Linux should show very different performance.
The Apple M-series is 3x more performant per watt than anything else.
Perf Per Watt (bigger is better):
https://www.notebookcheck.net/Apple-M2-SoC-Analysis-Worse-CP...Different test:
Cinebench R23 Multi Package Power Efficiency
M1 - 561 pts/watt
M2 - 433
AMD Ryzen 7 6800U - 374
Intel Core i5-1240P - 258
It should not be used to compare ARM vs x86 CPUS. Geekbench is better here.
I mean, 3x the performance per watt might be --lies-- --damn lies-- uh, benchmarks targeting the little CPU?
Isn’t almost everything designed for consumer use, even the Intel i9, these days?
I only meant "real workload" in the sense that when real people are using their computer, the P-cores are doing real work; that's true even for bursty workloads that don't keep the processor 100% busy, so power is being spent by the P-cores and their efficiency matters.
No, that's not how M1/M2 work.
Apps are generally on the P-cores, but daemons are on E-cores, and the daemons are doing essential work like keeping you connected to wifi. And the apps aren't running all the time if you're just browsing or editing text. It's all part of the end goal of having a lot of battery life.
SCNR!
As mentioned elsewhere in the comments, there was an Intel chip that appeared to beat the M1 on performance per watt if you looked at ‘tdp’, the issue being that Intel was using a particularly eccentric definition of tdp.
If you don't trust that, you can always measure the whole system usage at the plug.
Except, of course, for chips that are not available yet, such as those discussed here.
Just a super weird writing choice.
Also puts to bed the “x86 is fundamentally unable to be as low power as arm” rationale for the poor intel and amd perf/watt. I think the sole reason for that is that they never _really_ competed on that metric.
TLDR: The performance per watt gap is gone, but only when you compare at similar TDPs.
For example, the Gnome indexer, tracker-miner, should only ever be allocated to background status.
[1]: https://www.hardwaretimes.com/amds-first-hybrid-core-big-lit...
The whole m2 hootenany on hn reminded me of when retina came out. Its just a fancy branding for a hidpi display. Nothing inherently special.
Macbook Pro - 2560x1600, 227ppi
Asus Zenbook Pro 16X OLED 60Hz - 3840x2400, 283ppi
Asus Vivobook Pro 14 OLED 16:10 90Hz - 2880x1800, 243ppi
Asus ZenBook Pro 14 Duo OLED 120Hz 16:10 - 2880x1800, 234ppi
etc etc etc. I'm sure there's other brands.
Seriously, I do not doubt the performance (I hope they do not melt their cpus and motherboards this time), but energy efficiency, which is important in laptops? I will believe it when I see it (by in depth reviews, not by what benchmarks companies release), considering the desktop ryzen 7000s got to a completely different direction. Intel had made some ridiculous energy efficiency claims in their last generation that were totally out of reality in practice.
But I agree that AMD's marketing claims as filtered through low-quality journalism that doesn't even distinguish between peak and base power consumption is a poor way to judge whether AMD has succeeded in building a great low-power processor.
And yeah I agree, we run our new AMD desktops in eco mode at work, it was ridiculous to routinely get like 95C temperatures for whatever random chores. Once we used to overclock AMD cpus, nowadays we underclock them lol. I have not really checked how many watts it draws after we put them in eco mode though, I will do it next time.
Apple bought P.A. Semi in 2008, a company founded by Daniel Dobberpuhl of DEC Alpha fame. Clearly they had something brewing on the CPU front.
What might've been surprising is them getting into the big, high performance desktop-class CPU segment and leading the industry in efficiency and performance HEDT chips for desktops.
And Intel's relative irrelevance is certainly surprising, as is them losing the fab game to TSMC.
Don't forget Itanium, Intel's other dead end architecture. It's actually kind of amazing that Intel recovered from both at the same time (for the most part).
The Commission filed an administrative complaint against Intel Corp., the world’s leading computer chip maker, charging that the company had illegally used its dominant market position for a decade to stifle competition and strengthen its monopoly. The complaint alleged that Intel engaged in a course of conduct to shut out rivals’ competing microchips by cutting off their access to the marketplace. In particular, the complaint alleged that Intel unlawfully maintained its monopoly in relevant central processing unit, or CPU, markets, and sought to acquire a second monopoly in the relevant graphics markets, using a variety of unfair methods of competition. In August of 2010, Intel agreed to a settlement containing provisions that would undo the effects of Intel's past conduct, and prohibiting Intel from suppressing competition in the future.
[1] yes, I know ARM64 looks nothing like the ARM ISA of 1990.
DEC -> AMD (Dirk Meyer, Jim Keller)
DEC -> PA Semi -> Apple (Daniel Dobberpuhl)
[1] https://www.wired.com/1997/10/intel-dec-settle-alpha-chip-di....
[2] https://www.electronicsweekly.com/news/archived/resources-ar...
[3] https://www.cnet.com/tech/tech-industry/intel-gets-more-key-...
"They focused on servers, uh, you know, like for http. Those are really big in the future"
"Oh, weird"
Uh, the Switch is an NVIDIA Tegra SoC...
God it still hurts so much
For whatever reason, it isn't so easy to find real-world memory bandwidth results for consumer stuff; it's a fairly standard benchmark for server hardware, but less so for consumer. Here's one for an old Ryzen 9: https://www.anandtech.com/show/14605/the-and-ryzen-3700x-390..., but couldn't find anything current offhand.
That sounds incredibly specialized and I am needlessly curious.
Windows laptop companies, please DO NOT follow suit! Though some of them already did so, sigh.
Please Apple, please just replace your shiny aluminum with plastic (or some better alloy) and make your fanless offering 1kg.