But can it finally run Minecraft better than a 10 year old cheap Intel?
EDIT: For the downvoters - this was a joke (But can it run crysis?), but since this got attention - my pet peeves with Apple is that they advertised M chips as universally powerful and good, a serious competitor to Intel. It’s a mobile chip with instructions mostly for media encoding and decoding. It’s like comparing Prius to a bulldozer. If your workstation involves watching YouTube and using final cut pro, then sure, but anything remotely more advanced mostly falls apart on it. Don’t even get me started on AI and how slow the unified memory is
EDIT: Please stop ruining my impressions of average Lemmy person. Majority right now are acting really fucking stupid and immature just because I touched your favourite corporation. Read my explanation below. Shit, there was even a guy who posted a link thinking he’ll get me, while failing to read it and the article pretty much agrees with me (industry professionals)
But the big data centers doing all the video processing for the big video services (including both permanent videos from a library and things like live streaming) are encoding the videos with settings that require less computational power to decode. The idea is to be able to let even old budget smartphones still be able to display the video with very low power requirements on the client device. There’s no universe where consumers decoding digital video will be a high-power computational task.
Restaurants have sharp knives in the kitchen, but generally serve food that requires only minimal cutting effort from the table knives set out with the rest of the table settings. Dining will always be easier than cooking, by a margin that makes the difficulty of dining not worth mentioning, so it would be bizarre to criticize a knife as being only good for cooking and eating food, when plenty of dining tableware knives out there would be insufficient for kitchen work.
You’ve made the mistake of lumping decoding and encoding together based on the algorithmic/mathematical similarity of those tasks, when everyone else is more inclined to discuss the very different end user use cases of those computing needs.
Many plugins and apps I use don’t really work with GPU/Hardware acceleration when it comes to rendering, same applies to encoding in different codecs. I’d know, because I’ve been unfortunately doing this shit for nearly 20 years and building my workstations (definitely not ARM, screw your downvotes and love for it) around it.
Pretty much every serious studio out there uses either EPYC or Xeon and to me it seems ridiculous that apparently majority here doesn’t see the problem with my initial argument of apple marketing these chips as God-tier and beat-them-all, when clearly, as it has been proven before, apple heavily misleads with their marketing and it’s not as simple as it seems.
EDIT: And people who feel like arguing by bullshitting accusations (like the guy above about me not knowing anything) are basically how redditors argued.
You put editing and watching YouTube under the same umbrella
Video encoding and decoding is generally under the same category - video processing.
What editing software do you use?
Like, right now, or have used (pretty long list)? My favourite is still After Effects just because of how used to it I am, but I seriously do not feel like listing all the plugins and extra apps (probably any professional knows about mocha/syntheyes or nuke). That’s my main, I’ve even learned to mostly skip premiere (still gotta use media encoder for obvious reasons). For 3d stuff and effects - Cinema 4D (FumeFX, xparticles, realflow, etc). Good enough, detective?
Video encoding and decoding is generally under the same category - video processing.
That makes sense. I guess I’ve just not been at the level were people call it that so that is interesting.
Like, right now, or have used (pretty long list)? My favourite is still After Effects just because of how used to it I am, but I seriously do not feel like listing all the plugins and extra apps (probably any professional knows about mocha/syntheyes or nuke). That’s my main, I’ve even learned to mostly skip premiere (still gotta use media encoder for obvious reasons). For 3d stuff and effects - Cinema 4D (FumeFX, xparticles, realflow, etc). Good enough, detective?
Doesn’t take a detective to know that you know more than me. I’m at a much much smaller scale, mostly just Houdini at this point. My boss would laugh at me if I mentioned XEON for the office haha. Apologies for being so hostile.
Actually I tried some of their silicon-optimized modern ports like the Resident Evil 2 Remake on a MacBook Air (The one that doesn’t even have active cooling) and I was taken aback by just how well it ran.
Of course it’s not a gaming PC, but it for sure punches well above its class with the games it runs.
For a bit of fun I recently installed Horizon Zero Dawn on my M2 Air using Crossover, just to see how it fared.
I got a perfectly respectable 25fps, or thereabouts.
This is a three year old, fanless, entry-level laptop, playing a game that’s not even written for its architecture at a rate that’s entirely playable. It’s actually crazy how good that hardware is.
Just like “optimization for games don’t matter anymore because everyone has 32gb of ram”, right? Because ARM is just simple RISC, right? https://simplifycpp.org/?id=a0882
What you consider unimportant is actually super important in cyber security by the way.
On one hand you’re claiming RISC is bad because performance. On the other hand you’re saying the difference is important due to cybersecurity.
On the first point I’ve already proven you wrong. And the second point… RISC is actually safer for a number of reasons than CISC.
So are you for, or against RISC? Because both topics you picked support RISC pretty much unilaterally.
Also might I add that right now the only truly viable open future architecture is RISC-V, which is getting to a point where we can blend together MCUs and CPUs based purely on available extra hardware on silicon. Meaning same tooling will be able to compile the same code regardless if you want to run it on a sensor array that lasts a year on a single coin cell, or a powerful processing node (okay okay arguably RISC-V isn’t completely there for truly high performance computing, but there’s already cheap SoCs with Pi3 comparable performance, and the market is just gearing up for better ones. Oh and if your goal is AI, pretty solid 16TOPS in a low power package).
Yeah, I played large modpacks at good fps decade ago.
It’s not about whether you can use AI (obviously they can, commonly even larger models than similar laptops due to unified memory), it’s about how fast it is.
Gemma ran at 50/tops
Qwen 27B? was way slower , 5/tops
8B models run perfectly fine, but are mostly useless for chat and agents.
8B is only good for specialists. Like one 8B model that can only write and correct python3 code.
And then only in English.
But can it finally run Minecraft better than a 10 year old cheap Intel?
EDIT: For the downvoters - this was a joke (But can it run crysis?), but since this got attention - my pet peeves with Apple is that they advertised M chips as universally powerful and good, a serious competitor to Intel. It’s a mobile chip with instructions mostly for media encoding and decoding. It’s like comparing Prius to a bulldozer. If your workstation involves watching YouTube and using final cut pro, then sure, but anything remotely more advanced mostly falls apart on it. Don’t even get me started on AI and how slow the unified memory is
Everything you’ve said here let’s me know that you have no idea what you’re taking about. Lumping video editing with watching YouTube lmao.
It’s called decoding and encoding…
EDIT: Please stop ruining my impressions of average Lemmy person. Majority right now are acting really fucking stupid and immature just because I touched your favourite corporation. Read my explanation below. Shit, there was even a guy who posted a link thinking he’ll get me, while failing to read it and the article pretty much agrees with me (industry professionals)
But the big data centers doing all the video processing for the big video services (including both permanent videos from a library and things like live streaming) are encoding the videos with settings that require less computational power to decode. The idea is to be able to let even old budget smartphones still be able to display the video with very low power requirements on the client device. There’s no universe where consumers decoding digital video will be a high-power computational task.
Restaurants have sharp knives in the kitchen, but generally serve food that requires only minimal cutting effort from the table knives set out with the rest of the table settings. Dining will always be easier than cooking, by a margin that makes the difficulty of dining not worth mentioning, so it would be bizarre to criticize a knife as being only good for cooking and eating food, when plenty of dining tableware knives out there would be insufficient for kitchen work.
You’ve made the mistake of lumping decoding and encoding together based on the algorithmic/mathematical similarity of those tasks, when everyone else is more inclined to discuss the very different end user use cases of those computing needs.
Which is heavy on the CPU and GPU…
You don’t know much about encoding or decoding either.
Many plugins and apps I use don’t really work with GPU/Hardware acceleration when it comes to rendering, same applies to encoding in different codecs. I’d know, because I’ve been unfortunately doing this shit for nearly 20 years and building my workstations (definitely not ARM, screw your downvotes and love for it) around it.
Pretty much every serious studio out there uses either EPYC or Xeon and to me it seems ridiculous that apparently majority here doesn’t see the problem with my initial argument of apple marketing these chips as God-tier and beat-them-all, when clearly, as it has been proven before, apple heavily misleads with their marketing and it’s not as simple as it seems.
EDIT: And people who feel like arguing by bullshitting accusations (like the guy above about me not knowing anything) are basically how redditors argued.
You put editing and watching YouTube under the same umbrella and then speak of using EYPIC and XEON CPUs?
What editing software do you use?
Video encoding and decoding is generally under the same category - video processing.
Like, right now, or have used (pretty long list)? My favourite is still After Effects just because of how used to it I am, but I seriously do not feel like listing all the plugins and extra apps (probably any professional knows about mocha/syntheyes or nuke). That’s my main, I’ve even learned to mostly skip premiere (still gotta use media encoder for obvious reasons). For 3d stuff and effects - Cinema 4D (FumeFX, xparticles, realflow, etc). Good enough, detective?
That makes sense. I guess I’ve just not been at the level were people call it that so that is interesting.
Doesn’t take a detective to know that you know more than me. I’m at a much much smaller scale, mostly just Houdini at this point. My boss would laugh at me if I mentioned XEON for the office haha. Apologies for being so hostile.
And yet the “mobile chip” in my Mac Studio absolutely crushes my 9800x3d while using almost half the power.
Actually I tried some of their silicon-optimized modern ports like the Resident Evil 2 Remake on a MacBook Air (The one that doesn’t even have active cooling) and I was taken aback by just how well it ran.
Of course it’s not a gaming PC, but it for sure punches well above its class with the games it runs.
For a bit of fun I recently installed Horizon Zero Dawn on my M2 Air using Crossover, just to see how it fared.
I got a perfectly respectable 25fps, or thereabouts.
This is a three year old, fanless, entry-level laptop, playing a game that’s not even written for its architecture at a rate that’s entirely playable. It’s actually crazy how good that hardware is.
idk, I use a gaming PC for that stuff.
The Mac’s are for work and creation.
Yeah, instruction sets don’t matter that much with modern processors. Try bullshitting harder.
Just like “optimization for games don’t matter anymore because everyone has 32gb of ram”, right? Because ARM is just simple RISC, right? https://simplifycpp.org/?id=a0882
What you consider unimportant is actually super important in cyber security by the way.
So super important that it’s not important at all.
https://chipsandcheese.com/p/why-x86-doesnt-need-to-die
Did you even read what you sent me? Do point out to me the part about cybersecurity, I’m waiting.
EDIT: Actually, it’s a great article.
Dude, pick a lane.
On one hand you’re claiming RISC is bad because performance. On the other hand you’re saying the difference is important due to cybersecurity.
On the first point I’ve already proven you wrong. And the second point… RISC is actually safer for a number of reasons than CISC.
So are you for, or against RISC? Because both topics you picked support RISC pretty much unilaterally.
Also might I add that right now the only truly viable open future architecture is RISC-V, which is getting to a point where we can blend together MCUs and CPUs based purely on available extra hardware on silicon. Meaning same tooling will be able to compile the same code regardless if you want to run it on a sensor array that lasts a year on a single coin cell, or a powerful processing node (okay okay arguably RISC-V isn’t completely there for truly high performance computing, but there’s already cheap SoCs with Pi3 comparable performance, and the market is just gearing up for better ones. Oh and if your goal is AI, pretty solid 16TOPS in a low power package).
I play Minecraft with the PrismLauncher on the M4 24GB. Shaders and DistantHorizon. 50-80fps
AI also works. Gemma 4 26B a4b MLX runs… somehow.
Yeah, I played large modpacks at good fps decade ago.
It’s not about whether you can use AI (obviously they can, commonly even larger models than similar laptops due to unified memory), it’s about how fast it is.
Gemma ran at 50/tops Qwen 27B? was way slower , 5/tops 8B models run perfectly fine, but are mostly useless for chat and agents. 8B is only good for specialists. Like one 8B model that can only write and correct python3 code. And then only in English.