How Apple Silicon Compares... - Part 1
Vložit
- čas přidán 28. 01. 2024
- Visit www.squarespace.com/macaddress and use code macaddress to get 10% off
Now that there are games running natively on Metal and Apple Silicon, we can finally see how Apple Silicon Macs game at their full potential. And so Jonathan asked LTT Labs to help him test that, only to get more than he bargained for.
Buy an Apple 16" M3 Max MacBook Pro: geni.us/Kxn1r
Buy an Apple M2 Mac Mini: geni.us/Obk5xh
Buy an AMD Ryzen 7 5800X3D CPU: lmg.gg/Edsgm (Canada: lmg.gg/ulCr8 )
Buy an AMD Ryzen 7 5800X CPU: geni.us/1yi0A
Buy an AMD Ryzen 7 5700X CPU: geni.us/7Bpjy
Buy an AMD Ryzen 5 5600G CPU: geni.us/mF0juNT
Purchases made through some store links may provide some compensation to Linus Media Group.
Purchases made through some store links may provide some compensation to Linus Media Group.
► GET MERCH: lttstore.com
► GET EXCLUSIVE CONTENT ON FLOATPLANE: lmg.gg/lttfloatplane
► SPONSORS, AFFILIATES, AND PARTNERS: lmg.gg/partners
FOLLOW US ELSEWHERE
---------------------------------------------------
Twitter: / macaddressyt
Instagram: / macaddressyt
Facebook: / macaddresschannel - Věda a technologie
I love how differently you write these compared to the rest of LMG. Constructed like a news piece.
Looks more professional to me
100% agree. Less chaotic
Likely the different structure is because it doesn't adhere to the LTT formula. Especially as there is little to no attempt to explain why what the video is about should matter to those not interested in apple products. (pretty much every more fact oriented video on LTT explains in the beginning why what they're making a video about should matter to anyone.)
Seriously, Jonathan knows how to tell a story. Wish he made more videos more often!
I was thinking how much I enjoy this channel. And I'm super happy it isn't a clone of LMG. It has its voice and style.
I would get a new mac, only if another 8 gigs of ram didn't cost 200 dollars more...
your loss
@@chewbster His gain actually since modern amd laptops can match performance and battery life for macbook pro at a far better value
I am running an m1 air with 8 gigs and it is painful. I'm going to have to spend $800 more for used hardware to have less RAM than a new Windows machine.
Docker, fusio 360n, cura, and a browser are a pick 2 situation on 8 gigs and there's still huge swapping that will kill the NAND over time.
Mac's used to be the come up a few years later once they were steadily discounted. But that was when they were upgradeable.
@@piyh3962why did you buy the 8 GB model if that’s what you plan to do? That’s on you tbh
While I understand the reasoning behind sticking with AM4 I feel like comparing Apples latest M3 (3nm silicone) that just came out a few months ago (and their M2 Pro/Max/Ultra 5nm lineup that also just came out last year) to AMD's previous gen (7nm silicone) that came out 4 years ago was always going to be an unfair fight in Apples favor with CPU bound workloads. As for "no low end AM5 chips", you can compare the 7700 or 7800X3D with their 8 cores to the base m2 or m3 chips that also have 8 cores... the 6 core 7600 also exists, but no apple M chips have fewer than 8 CPU cores and are at max 5nm all the way back to M1. IMO a more fair test would be doing these comparisons using similar process node CPUs & GPUs like AMDs 5nm 7000 series vs Apples 5nm M2 (or M1) series tested at similar wattages or accounting for PPW. Despite this, I'm still looking forward to how the results turn out in part 2 and appreciate that you guys are at least trying to make an "apples" to oranges comparison (see what I did there... sorry lol) with PC Hardware.
Sounds like a bad plan if they are looking for what is comparable…
@@Tawmii91 Perfect plan, the end result will be that an M3 Mac has performance similar to an much cheaper PC from years ago.
@@Tawmii91 My point isn't about what X86 CPUs runs the closest in performance to certain Apple CPUs in order to test GPUs, I understand the methodology they are using for these videos. I'm just saying I'd be more interested in the results of a similar test where the CPUs (and GPUs) have the same silicone node size and core count (M2 5nm vs Ryzen 7000 5nm and RTX 4000 5nm), and are either tested at similar wattages or include PPW in the results. This way you are closer to a true apples to apples comparison and can more clearly compare ARM vs X86 rather than arriving to the obvious conclusion that the smaller node cpu/gpu options (M2 and M3) are more efficient than their larger node PC counterparts.
You're forgetting that the point isn't to compare CPUs - then your point would be fair. The point is to _eliminate the CPUs from the equation, so they can compare the GPUs_.
there's also disabling cores features on Ryzen, am I right? Like the whole strategy about AM5 not having budget options is absolute nuts, I am watching the whole video shaking my head on their decisions as to why they can't use the latest processors they compare. Also, AM4 is in DDR4 like that should also taken into account specially when they're testing gaming
Honest question: Isnt the data going to be skewed by excluding the newest CPUs from Intel and AMD? As an example the 5950x was released more than 3 years ago while the M3 was released at the end of last year? I think it would rbe really interesting for the public to see how the AM5 CPU from AMD rcompare too
Also it is a bit surprising to see both the intel 13700 and 13900 being excluded while also testing the M3 Pro?
They are just trying to match performance not exceed it. They are trying to test the Apple GPU, not CPU. Older CPUs are fine considering that
It's the only way they could get the Studio Ultra to be on top of every table.
@@peanutnutter1 Also, shouldn't the ultra chips be compared to AMD's threadripper chips from the time? the m1 ultra has like 20 cores even if it is a mix of p core e core.
@@avgHanimeEnjoyer You can get a new TR 7980 system for about the price of Studio Ultra
It´s a GPU comparison test. The idea is to match the equivelant CPUs so it does not skew the real test.
Love them or hate them, Apple silicon chips are impressive
Especially when compared to old Amd AM4 chips and entry level
Ikr? Amazing how Apple M3 manages to almost beat a 3 year old chip. Truly an amazing feat
What comes from this video (and even then it's barely touched), the true impressiveness comes mostly from the efficiency department, really.
As @willie69420 said, the test was only against AM4 chips, because they couldn't find any AM5 low-end enough to compare to slower Apple silicon ( 8:22 )
What really sets Apple apart from the PC space right now is how it performs while still being in a micro or even laptop form factor (and in the laptop FF, the fact it still have such battery life is truly amazing).
But yeah, being compared to a multi-years old architecture isn't the most flattering thing ever when only considering performance (and price to performance).
Honestly, I don't follow closely enough Apple hardware stuff to know, but I guess there were important redesigns between the different iterations of Apple Silicon, akin to a change of overall socket/generation for a CPU. So it feels a bit strange/"icky" having what is essentially 3 generations of Apple Silicon against only one in the CPU department.
I suspect parts of why they chose only one CPU platform was due to time budget allocated for Mac Address with the Lab team tho, and to keep charts readable in some ways.
All in all, the tradeoff is understandable and I'm quite interested in seeing the next parts, and how the team change their routine to accommodate for Apple benchmarking.
while being on 3nm and the am4 chips are on 7nm...@@willie69420
but apple killed it by not giving those platform (all of M seriess) upgradeability. Soldered DRAM is understandable, it saves power when the chip is closer to the CPU. but the SSD? its atrociuos and so anti consumer and anti green. give it a empty M.2 NVME slot and will erase 60 to 70% bad publicity. if they want to milk it, just give the slot to Pro line.
Jonathan running games with screen mirroring on is insane
Probably to avoid closing the MacBook that'd increase heat. In MacOS you can't display only on an external display like on Windows, you mirror and then turn the brightness all the way down on the MacBook. There is no "only show on Display #2" option afaik. Silly but yeah that's how the native display settings work.
@@pieterrossouw8596you can run external only but charger needs to be plugged in
@@x011V3Rx which would have you keeping the lid closed which may retain heat.
@@pieterrossouw8596 why not extend Main Display and change the display in-game?
@@pieterrossouw8596 I'm a Mac user and find this sad
When doing the gaming performance video please make sure on the high end you use the best "Max" chips and not "Ultra" chips. Many games don't run right on Ultra chips, because of similar issues to SLI. Ultra is two separate chips bridged together. A game needs to be specially optimized to run well on them.
Oh boy, got to the end of the video and the 40 core GPU M3 Max, wasn't chosen for the test :-( . It should have been the first pick. It is currently Apple's most apt gaming chip.
I liked Nikolas' presentation of the results. Seemed very confident on camera and sounds like he knows his stuff. Keep up the great work MA team!
I'm very confused by some of the methodology here - why does the longevity of the AM4 platform matter at all? Why weren't any newer chips used?
Not saying there weren't good reasons for all that, but they weren't clear to me.
Also, I'm obviously super out of the loop on Apple silicon, but it's a bit jarring to see M2 beating M3 so regularly in these results and everyone just chillin'.
It's been far too long since the last MA video! Hoping in 2024 you guys get the bandwidth to make them more often! It's my favorite of the LMG channels because Jonathan is such a great host.
They need more viewers. There's a good chance if you're in a developed country the majority of the people you know own an iPhone. Tell them about this channel.
@@yupimrandy lol
the scripts and editing is nice too
and they often film outdoors which i really like
I hope the colors for the bar graph are differentiated more, as Apple/Intel are near impossible to tell apart if you’re red/green colorblind
then turn colorblind mode on your computer or monitor
@@OmniKoneko correct answer. it's not their job to do it
M3 came out 3 months ago.
Intel 13500 came out a year ago.
All the AMD's they are testing, came out 3-6 years ago.
How is it fair to use old AMD CPU's?
I think they are testing for comparable raw cpu performance
But the newer ones have better technologies - at least it's still the same micro architecture (zen3 and zen4)
@@aravindpallippara1577 The M3 have newer better technologies too.
Theyre testing GPUs. This was part 1, setting up comparable CPUs.
@@dtemp132 These CPU's were not comparable. They are using 3 year old ones.
The 8600G and 8700G would be more comparable.
@@saulgoodman2018the test selected them as comparable in calculation.
i love the way that you were open and honest about the testing process, and what the challenges were.
I didn't get the point of comparing M# with AM4 chips and a low end 13th gen intel chips.
Mac adress is honestly my favourite channel from LMG
I really appreciate how difficult this must've been for everyone, especially when it comes to finding equivalent hardware between two VERY different designs. I can't wait for part two!
With the testing having so little Intel data, the coloration of the Intel chips on the graphs ended up looking like highlights. This was a bit confusing. I think a quick audible mention of the relevant (in this case Apple) chip's color would have been helpful, some better color selection, and/or zooming in on graphs to where only 4-5 items are showing would help so much. Other than that great video, please keep up the great work.
A laptop AMD Ryzen 9 7945HX runs 34k points in Cinebench R23 multi and 1940 single, an Intel i9 14900HX is 32.4k and 2.19k respectivelly, maybe 7800X3D capped at different frequencies to have a similar single core performance in R23 to each Mac cpu would have been easier.
Yeah i was surprised they didn’t put current gen processors on here, then i realized the best was a LAST generation M2 chip and they also didn’t include the m3 max chip. So it feels a bit more fair to compare last gen to last gen i guess.
It’s great that your videos are always documentary style. It adds a nice variety to the LMG output.
Finally a new episode!!!! This has become my favorite channel of the LTT world. We need more man!
I really like the style of Mac Address pieces. This right here was like a proper deep dive. Well done, Sir!
The Mac Address creative team have finally did it, brainstorming the absolute best idea on this channel.
I'm looking forward for the Part 2
I ended up getting an M2 Mac Mini last year as my office computer. I honestly havent looked back at windows. I keep a windows laptop for the programming software my company uses, but thats pretty much it. I used to think OSX was messy and lumpy. But i worked out that it actually caters to my ADHD with all the floating windows ontop of eachother. This was all off the back of your review, so thanks for that.
I also switched to an M2 Mac mini, MacOs is so much cleaner, I was getting sick of built-in Windows ads and clunky new features.
I have to use a Macbook Pro at work, and I have the exact opposite opinion funnily enough. I miss the days where I used Windows or Linux at work every day. I find the lack of niceties to be quite appalling, having to install random 3rd party apps to fix fundamental issues (MacMouseFix because the scroll wheel sucks by default, Rectangles to have proper(ish) window management). And to me, the fact that any desktop app that remains "open in the background" has a dot despite having no window opened is quite ADHD-enabling.
The shortcut layout doesn't make any sense: option, command and control literally all mean the same thing, and are super inconsistent. On Windows, with rare historic exceptions, if you want to do something within the application, it's Ctrl. If you want to do something a little different, it's Ctrl+Shift or Ctrl+Alt. If you want to do something navigation related, it's Alt. If you want to do something with Windows, or window management, it's the Windows key. Linux usually follows this too, with some variations but nothing major. On macOS, it's super unintuitive, with window management shortcuts on Cmd (like Cmd+H), and regular app stuff that isn't navigation being on Opt.
Also, macOS updates take so damn long to install. Sure, Windows updates often, but it doesn't take 10 minutes out of my day, usually 2 or 3 if it's a big update. I don't know why but "update during the night" does not work and never updated.
Apple as a company is also terrible, see how they maliciously comply with the DMA when they had every opportunity to comply the normal way, and get a massive reputation boost from techies.
Also a very stupid thing: unless it's DisplayLink, you can't use a USB C hub with 2 display outputs on it. It will only mirror the same screen and not extend on both screens. This works fine on Windows and Linux, I have no idea why it doesn't on any Apple Silicon-powered laptop. It's not the end of the world but it's really stupid.
Also, as a developer, I can't do anything to support Mac systems if I don't have a Mac myself. You can't officially do anything, not even a VM, and have to resort to hacky methods. This just sucks.
The main thing that is better is the hardware: phenomenal trackpad, phenomenal screen, great keyboard (although a VERY hard bottom out on keys which makes it a bit uncomfortable for heavy typers). But the software hasn't been great.
@@matthieulucas9059I have had to use a MacBook Pro at work for three years. I thought I'd get used to it. I haven't.
@@matthieulucas9059 I can't agree more. I am doing finance, and my company is "Macs-only". I ended up making them install Parallels and windows on the Mac and had to sign with blood that I am not opening a browser in Windows. Macs are simply not built for non-creative work: you can't do almost anything fast, and it is purely software design issue. Windows management is simply horrific on Macs for instance. Hotkeys are the mess etc.
But to be fair to Mac, I absolutely love my MacBook Pro at home that I use for photography (that's my huge hobby) and even some once-in-a-while gaming. Browsing internet is also a pleasure. Just don't put any deadline next to MacBooks :)
Is short: I believe one should choose their tools for the job. Windows is built for boring finance guys like me, macs for cool creatives like our designers and marketeers
@@matthieulucas9059 If you don't want to use a third party app for windows snapping, you can do this natively on your Mac with keyboard shortcuts.
As someone who has been using MacBook air’s since 2012, it’s amazing to me that being able to grab your laptop 3 days later and it’s still charged is a surprise to people. I only use windows on my desktop, I had no idea that was such a big issue with windows stuff
Happens on Linux too unfortunately. One of the biggest reasons why I'm leaning towards a MacBook instead of a Framework.
correct me if I'm wrong, but aren't those tested 3 years old amd CPUs? The intel one at least is farely recent
Yes, but thats not really the point. As he points out the single threaded benchmarks point out that the 5000 series of AMD chips is slower core for core than most of the Mx lineup, but the problem comes with you try and step up to 7th gen AMD, and current gen intel. The 7th gen AMD doesn't have any low end chips to pit against the low end m1, m2, and m3. So to get as wide an array of CPU's that best reflect the Mx competitor they are going with AM4 so that they can stretch backwards and forwards in time as best as possible while still using the same MB and ram to make the tests as apples to apples as possible. Yes, the ideal would be to use the 7th gen AMD, but for one that would make m2 and m1 the old boys as they predate 7th gen, and it would mean that they wouldn't be able to compare against similar level processors cause there is no low end for AMD just yet. So either have am4 and am5 and get even farther from apples to apples, or use am4 exclusively and get as close as possible while maintaining as much similarity between the runs.
@@mizinoinovermyhead.7523 They should've just not compare the Max and Ultra variants. They are in principle HEDT chips, no one is going. to buy them to game. The same as no one is buying threadrippers to game.
@@mizinoinovermyhead.7523 Ever heard of... Ryzen 5 7600X or Ryzen 7 7700X? Not to mention that Apple silicon is far more expensive than AMD CPUs, so you might as well get a Ryzen 9 7950X(3D) because that's what you'd be getting for the same price.
@@mizinoinovermyhead.7523 What low end chips? Apple's crap costs a fortune. They plan to use a generation old AMD's CPU's against modern Apple M3's. I'm not even sure if they plan to use Nvidia GPU's.
@@DukenukemX They did note that next vid will put amd AND Nvidia up to the test.
Buuut also, they should really do two/three different tests:
Price equivalent comparison (Likely the most important to gamers)
Performance equivalent comparison (tests how optimized games are on mac, which seems to be their current goal)
Architecture equivalent comparison (TSMC 5nm vs 5nm, who utilizes the wafers better?)
This was brilliant Jonathan. I really like the semi-documentary format
I can't wait to see how this turns out! The Apple vs Windows divide has been going on for longer than most LMG staff have been alive to the point where cross-platform comparisons don't really come up - you're either on one side or you're on the other, and X is better for task A whereas Y is better for task B and that's the end of it. I will be so glad to see what the results are, regardless of how they come out, so I have something to compare Apple Silicon to (I've never had a functional Mac of any sort). Can't wait to see the results!
Couple things to keep in mind:
1) All these results need to carry the caveat: These are last, and last-last gen PC processors (soon to be last-last and last-last-last). Instead of trying to find mac's that are comparable to PC performance, you're trying to find PCs that are comparable to mac performance, which is a pretty important caveat since you're literally having to go back a couple generations to find stuff that's as slow as the normal consumer grade stuff that Mac is producing. Their Ultra chip(s) are non-starters for most people because the price tag on those things is absolutely, gob-smackingly insane, and that's before you even start getting reasonably sized ram and storage added in.
2) As someone else mentioned, please do not, under any circumstance, test these with that abhorrent screen mirroring setup you demo'd it on. That's....just nutty (both for playing or for testing).
3) I was actually expecting Mac's to do a lot better than they did. Seeing an M1 processor losing to a 3000-series AMD chip in some of the tests was pretty eye-opening. Again, I don't think the Ultra is really worth talking about. You can build an absolutely state-of-the-art gaming PC that will blow the doors off the Ultra Studio/Pro machines in 90-95% of tasks (let alone games), for the cost of an Ultra machine. Ultra may be powerful, but it's just a horrible value proposition. Yes, a small subset of tasks are very fast on it, but those particular tasks are extremely fast on ALL apple silicon, so you don't ever buy Ultra just for those tasks.
4) I still can't get over that you're comparing 3-4 year old tech from AMD to state-of-the-art Apple stuff. If that's not a telling scenario for the Apple ecosystem, I don't know what is.
You make a fair point about how old the comparison PC parts are, but did you stop to consider the 1) power draw difference (like 30-40 watts vs 100-150 watts for the 5800x) and 2) that it is a laptop CPU vs a full size desktop CPU that, at launch, was nearly 500 dollars alone (also 5800x)?
I do agree to some regard though
@@ashtonbeca Let's say you live in Germany, one of the places with the highest energy costs, play 2h/day, what does energy cost? Let's say the power consumption difference is 200W (my PC on full power, Ryzen 5 2600 + RTX 3070).
365days * 2h/day * 0.4€/kWh * 1kW/1000W * 200W = 58.4€
So for 3 years of usage, you won't even get a RAM upgrade from 8GB to 16GB for Macs 😂.
@@ashtonbeca current gen amd while idles quite high they don't want to go above 70watts even with best gpus. apple packages the same internals in their desktop equivalents though
Good to see more LTT Labs content. So many other CZcams channels just list the manufacturer specs and claim it to be king then cry that LTT isn't doing it right. But this is a cool project and the fact the CPU is stacked next to similar low-end CPUs. I do give credit to Apple for having the low-power thing down, but ARM is good for mobile. The M chips are just Apple's twist on ARM, along with all the no-compromises OS changes to support the chip.
This is so dope, love how much is being accounted for. Something I'm learning that is massively beneficial is at the things Apple choses to focus on which better benefit the MAIN PEOPLE their professional Mac are targeted at i.e. creative work, amongst other thigns.
Can't wait for part 2!
This was a great video, I have no interest in owning a MAC but the presentation style and infomation Jonathen puts forward is very entertaining to watch.
Really appreciate just the attempt to see if this was even realistically possible. Although, my M3 max chip feels a little left out.
Quality content! Keep it up ❤
I bought an M2 Pro MBP just recently. I'm very new to Mac and Apple's ecosystem (only started using an iPhone with 11 when it dropped) but I gotta say, the performance of this thing is impressive. Also I used the Apple refurb store thanks to this channel's coverage of that whole process a couple of years ago.
Really cool info. The single core performance was also super enlightening considering a surprising amount of programs run solely or default to single-core. Just hadn't thought of it before. Thanks!
Pretty amazing guys! Keep it up!
As always, brilliant!
LTT lowered their uploading frequency in order to increase quality, but quite frankly I haven't noticed that much of an improvement. Mac Address on the other hand is an example of what that improvement might look like in an LTT video. It's so much better produced, edited and presented than LTT videos that it's surprising some of that hasn't carried over to LTT.
They're pretty much the only LMG videos I consistently watch, and I don't even have a Mac (yet).
That’s what happens when you use a mac 😂
But I agree. These vids are very informative and well written.
While I do somewhat agree, I will also say that I am not sure if it was ever intended for LTT to be as "professional" as MA. Part of the charm for LTT are also those more improvised videos and I think they do know thats their brand (I mean, sometimes they even construct a video specifically to bolster that improv-spirit, for example when they show setting up a massive TV without even thinking about how to do it best before).
What the extra time for LTT was mostly about is eliminating Errors and wrong information being given during a video as that happened way too frequently
@@latech7671 Today's video with ElectroBOOM is a good example of what you're talking about, it's just improvised antics and I thought it was great. But many of their videos aren't actualy improvised, they're quite scripted and could benefit from a production overhaul.
There's over a hundred people working there now, it's a huge company and considering the volume of content they produce not every single video needs to have that "improvised" quality.
....and no Linus
Watched this on floatplane yesterday. Part 2 will be fun
I bought an M1 Mac Mini when it first came out for Apple development. I ended up loving it so much I bought the M3 Pro MBP and have been converting more and more to Apple. I hate it, but also, love it! Lol😂
The fact that I’m watching this video before the main LTT video is huge, I love this channel
I think there's one thing that you're not considering when it comes to comparing x86/64 to ARM: the process node. Apple is notorious for pushing TSMC's process tech and AMD picks it up after Apple has done their thing. Perhaps it might be worth considering an additional test that compares equivalent nodes as well? (I think it would be M1 vs Ryzen 5000 or thereabouts)
Kudos to Apple for pushing the industry. (not talking about soldered ram)
@@jivewig I'm not sure what you mean, but if you're implying that process node technology wouldn't be improving without Apple, that's not the case. TSMC is the foundry for dozens of major chip design firms, not just Apple. If it weren't Apple, it would've likely been Nvidia.
@@Omonarc I’m not saying it wouldn’t have improved. You said Apple is pushing TSMC and I said that’s good.
@@jivewig Oooooh, I'm sorry if I sounded accusatory there.
@@Omonarc it was my fault, I just realized there was a typo in my previous comment thanks to Apple auto-correct. I’ve edited it and you can see what I meant to say.
I am so hyped for part II. My m1 MacBookPro runs BG 3 natively in 4k on highest preset with FSR set to Balanced... It is incredible...
So happy to see partnership with LTT labs
The M2 Ultra and M3 Max are cutting edge processors which are found in $3000+ computers. They are close the the 7900x in terms of single and multi core raw performance, and it is a disservice to compare them to anything but their true competition. Even the M3 is surprisingly close to the 7600.
Would be neat to see power consumption as part of this series, too. I love this though! I have a 5800X3D PC at home as well as an M3 Pro 12c MBP. Looking forward to the GPU results!
Such a cool concept!
It's cool that Labs is starting to hit these things in a way and with precision that wasn't possible for LTT before. Excited to see the results!
I didn't buy a Mac for gaming -- I bought it to learn Mac-only software and refresh my OS knowledge [I hadn't used a Mac regularly since the System 7 days], and it also plays Minecraft, lol. Nice thing about Minecraft being my main game is it's on Mac, Windows, and Linux, and I can just host a server to let me play the same world from all my machines :3
Man, I love the CBC-esque look this is going for.
I understand why you guys went with desktop comparisons but a laptop CPU test for the future would be mega interesting. especially if you guys did comprehensive video editing tests
I bought a Mac for gaming! 2017-2021 era when they supported eGPUs. I've since acquired a Windows PC to game and an M2 MacBook Air for everything else, including video editing.
Well I'm excited for the rest of this series.
The macs are really cool, but I am thinking about a framework for my next laptop (whenever that is, gonna wait till my current one needs to upgrade more)
FrameWork is going to run out of investor Money by Q1 2025
I work at a creative agency that is all Mac. We have about 100 employees. We had an it company come in and help do some upgrades. They were all PC. Well because the needed Mac’s to help us, they bought a few. They all converted and use Mac’s now. The biggest reasons were the speed and battery life in doing just normal office tasks.
I like this style of video, especially with Nikolas providing the expert input.
For how much more efficient Apple Silicon is, I’m really impressed by the results. I wonder how much of the gains come from having the SOC package vs the core architecture
Horst was so glad that someone finally bought a Mac thanks to one of his videos. Really good host!
Love the totally different style of MA very Documentary in style
Love it. It's always really interesting to see comparisons of hardware post-Apple silicon. Apple makes some outrageous claims in their marketing material which are really hard to put into context. Are they really fast? Are they worth the money? Are they really as power efficient as claimed? Looking forward to seeing how this series pans out!
Bunch of experts meet in a fancy lab to evaluate testing methods - proceed to running cinebench.
You should test the newest amd cpus too. At least for comparison.
YEP. It seems like they are going in with already skewed results.
the aim here is to find a CPU platform that is about the same so that they can then compare GPUs to figure out what PC GPU is comparable. The goal is not to find out what cpu is faster but rather find out what CPUs are the same perfomance (in gaming situations) so that they can remove the cpu differnce from the equation when comparing GPUs.
@@hishnash It is still a 3 year old CPU they are comparing to a 3 month old CPU. How is not comparable?
@@saulgoodman2018how thick are you that you don’t understand?
This is not a test to see who is best, let go of your apple rage.. they are looking for a similar performing cpu or “comparable” for the final gpu test.. geez
@@saulgoodman2018 the goal is not to find the fastest CPU.. The goal is to find the CPU that performs the same in games.
The aim of this video was to figure out the test harness so that they can then compare GPUs (in gaming).
They said the reason they did not use the latest AMD cpus is that there are no modern CPUs from AMD (on the desktop) that are as low perf as the M1/3 entry level SOC and thus would not be a usefull base to compare with. (hard to compare GPUs when the cpu your using runs differently)
I got an old 11" 2015 Macbook Air to learn how to use it (and because I don't have money to spend on an M1), even if I've daily drived an iphone and an ipad pretty much since the iphone 3gs days I never ever touched a mac, and well, lemme tell you, after a learning curve and adapting period it's being my work computer, some things just work, like airdropping files and pictures from my iphone to my ipad or macbook, the fact that everything in my icloud, from photos, contacts, passwords, email and my notes, my dear notes that fill my iphone were all immediately accessible in my mac was a godsend, I can't believe I refused for so long to get a macbook, I'll try my best to get on the Apple Silicon bandwagon after experiencing first hand what using a mac feels like
I know the team's bandwidth might be limited, but imho it would be prudent to also include the latest AM5/Zen4(or 4c) chips seeing how the Ryzen 5000 series are over 3 years old and yet you compare them with the absurdly expensive M2 Ultra!
Can you post in the descriptions below your video your lab results you have featured?
Testing Ryzen 5000 in single core test and wondering why they such, good job labs👏
I bought a Mac for gaming 😭 dolphin runs so well, and I can still play Roblox and Minecraft, and I can easily take it with me everywhere
Yes finally! This is awesome! Also I love BG3
It would be interesting to re-do the test when there are more ARM processors on the market, since the more direct comparison might be interesting, to see how much of the performance gains are due to the architecture, and how much is down to choices Apple made with the CPU/SoC.
Probably the best video about labs.
It would make more sense to compare a modern platform to apple silicon. or if you don't do that, compare the same generation of devices to each other. Using AM4 is just a legacy platform (That I use and love) but still not really a interresting test. So please compare same node size / same generation chips to each other. That would give interresting insight to the x86 vs ARM question. Still very nicely produce, thank you.
Good job showcasibg labs, I really appreciate the overreach MA videos have.
Why ubuntu 16. Does it did you compile with optimalisations and such? It could mean a big difference.
Great video! Yeah macs are always great with CPU performance. Curious how the latest gen will perform against the M3 max though.
They aren't testing latest Gen. These are all 5000 series CPUs instead of 7000 series. At least the 5800X3D was a 2022 release, but comparing it against the M2 Ultra, still released over a year later, seems odd. I get they couldn't have tested every cpu, but I would think they would aim for similar price/release date to compare against.
@@godminnette2 yeah they mentioned why in the video. Still curious though. Also surprised the M2 Ultra just destroys the rest of them. Though it’s probably better pitted against a thread ripper.
@@CreativeMindsAudio Its a poor excuse though, as many have mentioned they are creating a bigger memory bandwidth disparity by not using DDR5.
@@alexatkin for sure! at the same time though, my PC is a 5800X and my macbook pro is the last intel mac. I'd like to get an apple silicon computer, but so far this is serving me well. first i'd like to get a new GPU, unless the next tests show that the 4000 series nvidia stuff doesn't completely dominate the apple chips. if the GPU is pretty meh by comparison i may opt for a cheaper macbook upgrade.
Excellent video and super entertaining. I find it impressive we are getting desktop performance with apple silicon.
So did you guys get left out of early reviewer list for Vision Pro?
I love how you name everyone who works with you instead of just saying "Labs". The fact you treated people as people instead of being overshadowed by the company name means a lot.
I do like Mac book airs for being passively cooled no noice from a fan, and no issue with battery compared to a lot of windows laptop and weight is relatively low. I hope a windows/linux laptops can do that too soon ish
I even tried bg3 on the mac book air m2 24gb ram 2Tb ssd (big cpu version of the m2) i was surprised by it but for gaming i will stick with my pc but well I could use a Mac book(was bought used with rest warranty)
I LOVE the Sticker Bomb cover!!!
I did bought a mac (16" M2 Pro) for gaming, at least it was one of the few reasons. I bought my mac to play football manager, nothing too intensive but the game definitely runs better compare to playing it on my 7700k pc.
I think this is the best Mac Address video yet.
love that imbedded sigh right out of the gate at 1:15 from the labs guy 😀
I love how at the end he said he bought a Mac as a daily driver but then right after said he loves that he can neglect it for 3 days and still have a charge. So it’s more of a “every 3 days driver” haha. I will say I do love my Apple silicon macs. They have been solid machines to work on and the perfect option for my wife to run her business from.
You have never gone a whole weekend without going to the computer? You should work on that. :)
Actually curious about this, would like for a price comparison but that would be even harder than the tests
Yeah, especially as form factors should be taken into account then.
A 1000$+DIFFERENCE
On apple for sure
Still at the beginning but FYI Baldur’s Gate 3 on Mac is actually using legacy versions of Metal from before the ARM era. It’s not quite native.
Cool video! I don't agree with your method though. You now picked CPUs based on them having similar performance to the Apple chips which gives the best comparison for the graphics cards in a later video. It doesn't show the best value for CPU's though.
I'd love to see a comparison of similarly priced PC builds in the future using the latest and greatest CPUs and GPUs like the €380,- 7800X3D.
Seeing the price of a Mac Studio M2 Ultra you could build an AMD Threadripper 7970x system.
Wonder if you are going to test that new Apple Vision Pro that came out this month.
How long till part 2, I’m on the edge of buying a new Mac book and knowing what it’s like in gaming and performance between m3, pro and max.
I feel like you should do the FLAC audio encode test and the Libraw test again with the discrete GPUs (especially with intel ARC GPUs) because some of the GPUs have dedicated media encoders in them (intel ARC GPUs have dual media encoders, while most other GPUs only have one media encoder)
The main problem with this comparision is that the memory used in Apple Sillicon is a lot faster than the DDR4 used in the AM4 platform. A more fair comparison would be with DDR5 memory in the AM5 platform.
What a great testing
i have the base m1 air w/ 8gb - i use it for dev work/coding, and it’s so incredibly fast, and i can bring it anywhere w/o a charger!
All I know is that I was shocked when I installed Baldur's Gate 3 on my M2 MacBook Air and found that it was possible to play it very enjoyably. Going ARM was the smartest thing Apple ever did and I hope the rest of the industry follows.
I am once again advocating for the use of TouchDesigner in the testing/bench-marking suite.
SUPER tight knit integration with NVIDIA SDKs on windows
Natively cross platform on macOS (apple silicon native bundle) and on Windows
EXTREMELY in-depth analytics on system resource use down to the 0.001ms Per Operation
Native support for GLSL Shaders (a PROPER way to benchmark a GPU across OS types just short of C-Ray, can build ur freaking own)
I think it'd be really interesting to redo this test with computers equipped with the new snapdragon arm based chips
Here are some key takeaways from the video:
1. The goal is to compare Apple silicon (M1, M2, M3 chips) to off-the-shelf PC CPUs and GPUs for gaming performance. This is now possible because some games can run natively on both Macs and PCs.
2. They tested 7 different benchmark tests on both Macs and PCs to compare CPU performance. The tests measured single-core and multi-core performance.
3. In single-core tests, Apple silicon dominated, especially in audio/video encoding tests. In multi-core tests, high-end Macs matched or exceeded gaming-focused AMD Ryzen CPUs.
4. They matched Apple chips to AMD Ryzen CPUs for gaming rigs: M2 Ultra -> 5800X3D, M2 Pro/Max/M3 Pro -> 5800X, M3/M2 -> 5700X, M1 -> 5600G. These are best guesses pending GPU tests.
5. Testing showed Apple silicon is very efficient but not necessarily faster for gaming. More testing is needed with GPUs to draw final performance conclusions.
6. The process demonstrated that benchmarking two completely different architectures has challenges and required compromises. There is still more to learn.
7. The exposure to Apple silicon performance and efficiency left a positive impression on the PC-focused testers. One tester ended up buying a MacBook Air to use daily.
9:26 as Dijkstra said, testing only shows the presence of bugs, never their absence. It’s okay at some point where you’re more or less certain you’ve done all the testing that’s practical
Honestly I’m pretty excited for Mac gaming despite the lack of market, because Apple’s hardware, as unrepairable as it is, is an excellent package
I'm a software developer and, coincidentally, I work on both an M2(latest OS/24GB RAM) and a 5600G(Pop_OS/32GB RAM). The Mac M2 is faster and more responsive when compiling and running code. Most of the time I rather work on my Mac than on my PC. I haven't benchmarked them but I've worked on both using the same IDE(IntelliJ) and the same codebases interchangeably and it's pretty noticeable.
More Jonathan featuring the behind-the-scenes staff at LTT !
Thank you.