combatentropy
Veteran
sorry, wasn't sure
Apple dumps Intel and shows Final Cut Pro on new Mac with Apple CPU !
- Thread starter joe1946
- Start date
combatentropy
Veteran
I love 13" screen for how I use my laptop, which is sitting up in bed with it on my lap. It's very close that way. At night, it's overwhelming, and I wish I had 11 or 12" screen. If the lowest brightness setting was dimmer, then of course that would also solve the problem.
If it's parked on your desk, though, I agree it can feel small. Then again you're going to get carpel tunnel that way
Keep wrists below elbows.
ahalpert
Major Contributor
Eh tricky not impossible.
I have a 512gb ssd in my mbp and can handle numerous projects worth of proxy files simultaneously. (And editing proxies off the internal ssd is stupid fast.)
Current USB speed is so high that you can get similar speeds working from an external ssd.
I wouldn't go below 512gb internal for my next MBP but I could do it with 256 if push came to shove.
You'll never get a large enough internal ssd on an mbp to handle all the raw footage you have on hand anyway, so external storage will always be part of the equation. One man's opinion
Samuel H
Major Contributor
I just read the very detailed and in-depth Anandtech article about the A14
https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/4
(I link to page 4 because it has a time-vs-performance chart that's just incredible)
The core microarchitecture is huge, much wider than intel and AMD designs in most aspects. This usually means much-lower clock speeds, but they're already making it run at reasonably high GHz within the thermal constraints of a phone, so frequencly is unlikely to be a problem. Everything points to this CPU really being a huge performer.
Their conclusion is striking
The first version of anything is prone to be buggy, but this looks extremely interesting.
https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/4
(I link to page 4 because it has a time-vs-performance chart that's just incredible)
The core microarchitecture is huge, much wider than intel and AMD designs in most aspects. This usually means much-lower clock speeds, but they're already making it run at reasonably high GHz within the thermal constraints of a phone, so frequencly is unlikely to be a problem. Everything points to this CPU really being a huge performer.
Their conclusion is striking
The first version of anything is prone to be buggy, but this looks extremely interesting.
AnandTech is one of the most BS-free tech publications I know of, as well as among the most tech savvy.
It's an interesting competition. Apple and ARM are heading down the road of very wide instruction decode, while both Intel and AMD are using by comparison narrow decode, but storing the decoded instructions in a micro-op cache, so they can dispatch more instructions than they can decode. In the low power arena, I think it's a given. Unfettered by power constraints, the energy efficiency from the microarchitecture is largely irrelevant because the interconnects, caches, and cache coherency logic require power, and that will likely overwhelm the power advantages from the architecture.
Intel has some impressive new architectures in the works that looks to be solidly competitive with Zen3... due to launch in 2022. AMD will be launching Zen5 in 2022 if it keeps is roadmap alive, and Apple would be on M3, assuming it follows a simple numerical naming scheme.
It's not the first version, it's the 12th. I don't expect bugs to be a problem, the question is whether or not Apple can deliver perfromance enough for professional use. A year ago the answer would have been "Lol!" but today the answer is a definite maybe.
IMO the deciding factor is going to be whether or not Apple's bean counters think it's worth the R&D effort and cost to develop the interconnects, cache coherency protocols, and so on for large scaleup like what AMD and Fujitsu have done.
ahalpert
Major Contributor
Yeah, this is what pushed me to get an MBP instead of putting the money into another desktop workstation for my last upgrade. I can edit at the playground or while my son is napping on my chest. Any speed advantages of the desktop machine are far more than offset by how much more work I can sneak into my life.
That being said, I have a 15" and would prefer a 16" even on my lap. As long as it's not unwieldy, extra real estate is nice! A larger viewer, more tracks on screen, more of the timeline in view...however you want to use it. I like editing with 2 or 3 displays, so going down to 1 is a bit of a crunch even if the mobility is awesome.
As for screen brightness - if I'm not color grading, I can use Night Shift at night and that dims it a bit more. You can also probably get a 3rd party app like f.lux or twilight (I forget which is for mac) and bring the brightness down even more. I haven't done that because I'm reluctant to clog the OS with anything. Plus, if I'm pulling an all-nighter, I can turn off Night Shift and crank the brightness for a very effective melatonin suppressant!
NorBro
Major Contributor
They've had so much time perfecting these phones and pads.
Do you know if anyone in the computer universe ever predicted - around 2005, 2010 - Apple would eventually do mostly everything themselves with computers?
Was that a crazy thought back then?
I wish SJ could see this...
Not that I know of.
And given how badly Intel faceplanted, if it weren't for AMD, there wouldn't be any real competition for the M1. Intel's been stuck on the same process and largely the same architecture for around five years now...
Yes it was, because no one could have predicted that Intel would go from being 18 months ahead of the rest of the semiconductor industry to two years behind over the course of three years... and even more, no one five years ago expected that AMD would still be around now.
Meh. He's have found a way to screw it up, and make up for it with even more ridiculous marketing.
NorBro
Major Contributor
I'm thinking about these powerful people at these companies all thinking about each other and what they know and don't know.
I don't follow the specifics about the business(es), but how Intel let this happen is unimaginable.
Then again - and before I wrote this post - I looked up Intel's revenue and Apple only accounted for at most 3 billion out of the 75 billion they made, so maybe it's not the end of the world.
I don't follow the specifics about the business(es), but how Intel let this happen is unimaginable.
Then again - and before I wrote this post - I looked up Intel's revenue and Apple only accounted for at most 3 billion out of the 75 billion they made, so maybe it's not the end of the world.
combatentropy
Veteran
Yes. It has always been Apple's style to control their hardware and software as much as they can. Steve Jobs was famous for his perfectionistic micromanagement.
In 2007 the first iPhone held a chip designed by Apple. Then in 2008, when they bought PA Semi, I thought Apple would soon move their whole line to ARM.
My take was, "What took them so long?" The answer partly is, I am not a chip designer. The other part is, I would not have waited till their chips matched Intel's. I've been pretty satisfied with processor power for many years. Then again, I don't run Windows, and when I browse the web, I have JavaScript turned off except for the websites that I have whitelisted. You should try it!
G
Guest
Guest
And maybe Intel has been sitting on their own next thing for years, since they've had no competition till very very recently.
NorBro
Major Contributor
Dude, haven't you heard?!
I'm on M1 now; my computer can handle JavaScript for all websites.
NorBro
Major Contributor
Everyone likes a good comeback story...
No, it was a series of screwups that combined had the effect of a major screwup.
The process problems stemmed from an obsessions with getting better scaling from a new process than anyone else... and that backfired in the form of abymsal yields.
The supply problems came from poor production planning; Intel took on manufacturing so many cellular modems that it wasn't able to keep up with demand for its processors, opening yet another door for AMD -- but that door opened while AMD was still behind in single-threaded performance by a non-trivial margin.
combatentropy
Veteran
Intel is falling apart. I may be intermixing its history with Boeing, but it's the same trajectory and it has happened with many others.
1. A nerd forms a company, hires other nerds. Naturally, their focus is on technical quality.
2. good quality + good timing = $millions.
3. Head nerd retires to Bermuda. Replaced by Jerry Maguire or Ebenezer Scrooge.
4. Costs are cut. The best engineers leave, to go where they're more appreciated.
5. Company is a hollow shell of what it once was.
---
It could turn around at any time, but its prospects will only worsen. I think what you'll see is that a lot of other companies are going to move away from Intel, to ARM or RISC-V.
Large companies tend to sit on what's working for them, even if it's technically possible to move to something else and actually grab the lead. This is because very large companies are filled from head to toe with coasters. Change is hard, painful, no fun for the 9-to-5er who just wants his health insurance and to go home and see his kid's soccer game.
But their tendency to sit around is rivaled only by their pack mentality. As soon as a company, for whatever unlikely reason, breaks out and moves ahead, then the other sleeping giants wake up. It becomes a trend! (These companies love talking about trends, ironically.) It's suddenly the hot topic at the next board meeting. Some get tired and go back to sleep, but some companies will start moving in the same direction, trying to catch up to the leader.
A good example of this is when Canon came out with 24p. As we all know, it was some kind of accident. This is because if they knew what they had on their hands in the 5D Mark II, then they would have saved the feature for the $15,000 C300. This is even though for years it had been blindingly obvious to us that a large-format, 24fps camera would be awesome. Remember, this was in the age of 1/3" sensors. People were attaching monstrosities to their camcorders, those ground-glass 35mm adapters. It also helped that Red camera had started. Anyway, after the 5D with 24p came out, what do you know! Just a few years later, it seemed like every camera company was making 24p DSLRs.
1. A nerd forms a company, hires other nerds. Naturally, their focus is on technical quality.
2. good quality + good timing = $millions.
3. Head nerd retires to Bermuda. Replaced by Jerry Maguire or Ebenezer Scrooge.
4. Costs are cut. The best engineers leave, to go where they're more appreciated.
5. Company is a hollow shell of what it once was.
---
It could turn around at any time, but its prospects will only worsen. I think what you'll see is that a lot of other companies are going to move away from Intel, to ARM or RISC-V.
Large companies tend to sit on what's working for them, even if it's technically possible to move to something else and actually grab the lead. This is because very large companies are filled from head to toe with coasters. Change is hard, painful, no fun for the 9-to-5er who just wants his health insurance and to go home and see his kid's soccer game.
But their tendency to sit around is rivaled only by their pack mentality. As soon as a company, for whatever unlikely reason, breaks out and moves ahead, then the other sleeping giants wake up. It becomes a trend! (These companies love talking about trends, ironically.) It's suddenly the hot topic at the next board meeting. Some get tired and go back to sleep, but some companies will start moving in the same direction, trying to catch up to the leader.
A good example of this is when Canon came out with 24p. As we all know, it was some kind of accident. This is because if they knew what they had on their hands in the 5D Mark II, then they would have saved the feature for the $15,000 C300. This is even though for years it had been blindingly obvious to us that a large-format, 24fps camera would be awesome. Remember, this was in the age of 1/3" sensors. People were attaching monstrosities to their camcorders, those ground-glass 35mm adapters. It also helped that Red camera had started. Anyway, after the 5D with 24p came out, what do you know! Just a few years later, it seemed like every camera company was making 24p DSLRs.
Last edited:
NorBro
Major Contributor
The Japanese actually licensed 24p from these people below.
Adding 24p was only an "accident" in the stills camera (more likely an oversight since no one predicted that people would actually care to shoot video on photography cameras).
https://www.przyborski.com/?page_id=1537
After the 5D's success - and people like Vincent Laforet and Philip Bloom helping push these kind of cameras shooting video - of course the rest of the companies jumped on it.
Adding 24p was only an "accident" in the stills camera (more likely an oversight since no one predicted that people would actually care to shoot video on photography cameras).
https://www.przyborski.com/?page_id=1537
After the 5D's success - and people like Vincent Laforet and Philip Bloom helping push these kind of cameras shooting video - of course the rest of the companies jumped on it.
combatentropy
Veteran
Clermond
Veteran
I canceled my iPad Air4 preorder. Main purpose was playback R5 files on location from a card reader. I’m very tempted by the 13” MBP but as there is no USB4 hub (yet) I have no clue how I could copy files from a reader to the external SSD and power the book at the same time. Same in the office with a bigger second screen
Thomas Smet
Veteran
Kind of important to also keep in mind benchmark scores for the M1 are only really using 4 cores. The 8 core CPU uses 4 high performances cores and 4 high efficiency cores for simple tasks. So any high performance benchmarks are only using 4 cores really and its beating the top 16" MBP Intel 8 core CPU.
Intel should be really worried.
Technically a 16" MBP could use the same chip since its already beating the Intel 16" MBP. I hope that's not the case and Apple goes with a 16 core CPU with 8 high performance cores which would put the 16" MBP into MP territory. Doubling the GPU cores as well would give 5.2 TFLOPS or just about what the high end 5600m can do on the 16" MBP. Except it doesn't cost an extra $800 for that GPU.
Whats really exciting about all of this is we no longer have to nitpick over paying $200 more for a 0.2ghz boost in CPU speed which means practically nothing in the real world. We no longer have to figure out the confusing meaning between a i3, i5, i7 and i9 CPU. We no longer have t worry about different GPU upgrades either. We get the same top performing CPU and GPU no matter what. Any configuration now is based entirely on ram and storage.
We also don't have odd CPU and GPU differences between different tiers of computers. The MBA, 13" MBP and Mac Mini all use the same CPU and GPU. Makes it much easier to pick a computer that best suits ones needs. That decision is now based entirely on portability or sitting on a desk and if one doesn't need to pay for a display, keyboard, mouse and webcam because they already have that stuff. Its no longer about well this CPU is 200 MHz faster so maybe I should get the Mini instead. None of that garbage matters anymore and we don't have to pay the base price and get the crappy i3 CPU.
The 16 GB ram limitation and lack of eGPU support sucks but whatever. I bet something like 90% of 13" Mac users never upgrade past the 8 GB anyway. The people that walk into a Apple Store or Best Buy to buy a Mac don't get that option. Before 2018 no MacBook had more than 16 GB support so it isn't a huge deal breaker and the machines will still be able to do a lot. These are also entry level machines. Sure the 13" has pro in the name but really its still entry level and not meant for the level of power users that actually need 32 GB and 64 GB of ram.
The eGPU thing kind of sucks. eGPU was never just about replacing an inferior GPU but also about adding to an existing GPU. Applications like Blender and Resolve will use multiple GPUs for rendering. As sucky as the Intel Macs will be they still have that option to use up to four GPUs in Resolve and pretty much no limit with Blender. I think Apple and FCPX failed the concept of eGPU by not allowing FCPX to use all the GPUs in the system and instead they opted for a use this one or that one only approach. Lots of reviews mistakenly thought eGPU wasn't worth it on Macs like the iMac but they failed to realize that it was FCPX that failed there and if they used Resolve they could use both GPUs together.
I hope Apple makes their own M1 based eGPUs in the future. Something like a SSD size eGPU would be amazing.
As fast as a GPU will get it will never be as fast as having 2 or 3 of them. Apple is going to have to do something for the MP that has a more modular design. Maybe not eGPU like but they will have to have some kind of user swappable GPU or pros would call for their heads again after finally getting a modular Mac.
Intel should be really worried.
Technically a 16" MBP could use the same chip since its already beating the Intel 16" MBP. I hope that's not the case and Apple goes with a 16 core CPU with 8 high performance cores which would put the 16" MBP into MP territory. Doubling the GPU cores as well would give 5.2 TFLOPS or just about what the high end 5600m can do on the 16" MBP. Except it doesn't cost an extra $800 for that GPU.
Whats really exciting about all of this is we no longer have to nitpick over paying $200 more for a 0.2ghz boost in CPU speed which means practically nothing in the real world. We no longer have to figure out the confusing meaning between a i3, i5, i7 and i9 CPU. We no longer have t worry about different GPU upgrades either. We get the same top performing CPU and GPU no matter what. Any configuration now is based entirely on ram and storage.
We also don't have odd CPU and GPU differences between different tiers of computers. The MBA, 13" MBP and Mac Mini all use the same CPU and GPU. Makes it much easier to pick a computer that best suits ones needs. That decision is now based entirely on portability or sitting on a desk and if one doesn't need to pay for a display, keyboard, mouse and webcam because they already have that stuff. Its no longer about well this CPU is 200 MHz faster so maybe I should get the Mini instead. None of that garbage matters anymore and we don't have to pay the base price and get the crappy i3 CPU.
The 16 GB ram limitation and lack of eGPU support sucks but whatever. I bet something like 90% of 13" Mac users never upgrade past the 8 GB anyway. The people that walk into a Apple Store or Best Buy to buy a Mac don't get that option. Before 2018 no MacBook had more than 16 GB support so it isn't a huge deal breaker and the machines will still be able to do a lot. These are also entry level machines. Sure the 13" has pro in the name but really its still entry level and not meant for the level of power users that actually need 32 GB and 64 GB of ram.
The eGPU thing kind of sucks. eGPU was never just about replacing an inferior GPU but also about adding to an existing GPU. Applications like Blender and Resolve will use multiple GPUs for rendering. As sucky as the Intel Macs will be they still have that option to use up to four GPUs in Resolve and pretty much no limit with Blender. I think Apple and FCPX failed the concept of eGPU by not allowing FCPX to use all the GPUs in the system and instead they opted for a use this one or that one only approach. Lots of reviews mistakenly thought eGPU wasn't worth it on Macs like the iMac but they failed to realize that it was FCPX that failed there and if they used Resolve they could use both GPUs together.
I hope Apple makes their own M1 based eGPUs in the future. Something like a SSD size eGPU would be amazing.
As fast as a GPU will get it will never be as fast as having 2 or 3 of them. Apple is going to have to do something for the MP that has a more modular design. Maybe not eGPU like but they will have to have some kind of user swappable GPU or pros would call for their heads again after finally getting a modular Mac.