Should Qualcomm Feel Threatened By Apple's M1 Macs? (pcmag.com) 257
PCMag.com's lead mobile analyst calls Qualcomm "a little too unbothered by Apple's M1 Macs"
Qualcomm executives brushed off a question about Apple's new M1-based Macs during a question-and-answer session at the company's Snapdragon Summit today, where Qualcomm announced a new flagship smartphone chipset but no upgrades to its year-old chips for PCs... In general, reviews of Qualcomm-powered laptops such as the Microsoft Surface Pro X have celebrated the devices' long battery life, but lamented problems with third-party apps that were originally coded for Intel processors. That stands in stark contrast to Apple's new M1-based Macs, which don't seem to be slowed down as badly by older software...
"It's a great validation of what we've been doing for the past few years and [Qualcomm's product line] is just going to get stronger and stronger as we broaden our scope," said Alex Katouzian, Qualcomm SVP for mobile. Katouzian made sure to subtly call out ways in which Qualcomm's always-connected PCs are superior to Apple's newest Macs. The Macs lack 4G connectivity and still have poor-quality, 720p front-facing cameras... Katouzian also pointed out that (presumably unlike Apple) Qualcomm addresses "many tiers...and many price points" with its 7c, 8c, and 8cx laptop chipsets, letting Windows laptop makers drive prices well below the MacBook Air's $999 list price.
The core problem with Qualcomm's always-connected PC strategy is one that Qualcomm itself can't fix. While Qualcomm could, and probably will, soon announce a laptop chip that's based on the new Snapdragon 888 and has a level of raw power closer to Apple's M1, it's really down to Microsoft, as well as peripheral and app makers to solve the platform incompatibilities that have frustrated PC reviewers.
Hot Hardware cites Microsoft's promises of changes come in future updates to Windows 10, arguing that "with the arrival of x64 emulation and a growing library of native Arm64 apps, Windows 10 on Arm is going to be an even more powerful platform." From a performance perspective, while running Windows 10 on Arm, these [Snapdragon 8cx] chips may currently be at a disadvantage to the Apple M1, but some day in the not so distant future that might not be the case. We have no doubt that Qualcomm is likely working on a new Windows PC-centric SoC that is based on Snapdragon 888 or similar architecture. Qualcomm has promised a 25 percent uplift in CPU and a 35% lift in GPU performance over the Snapdragon 865, with the Snapdragon 888, which already offers a big boost over the previous gen Snapdragon 855/8cx. So, Qualcomm has the potential to put up a strong showing against the Apple M1, whenever its next-generation Snapdragon PC chip launches.
That may be, but John Gruber at Daring Fireball argues that currently "M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux." Those machines are just standing around in their underwear now because the M1 stole all their pants. Well, that just doesn't happen, your instincts tell you. One company, even a company like Apple, doesn't just embarrass the entire rest of a highly-competitive longstanding industry. But just because something hasn't happened — or hasn't happened in a very long while — doesn't mean it can't happen. And in this case, it just happened... M1 Macs completely upend what we can and should expect from PCs. It's a breakthrough along the lines of the iPhone itself in 2007.
"It's a great validation of what we've been doing for the past few years and [Qualcomm's product line] is just going to get stronger and stronger as we broaden our scope," said Alex Katouzian, Qualcomm SVP for mobile. Katouzian made sure to subtly call out ways in which Qualcomm's always-connected PCs are superior to Apple's newest Macs. The Macs lack 4G connectivity and still have poor-quality, 720p front-facing cameras... Katouzian also pointed out that (presumably unlike Apple) Qualcomm addresses "many tiers...and many price points" with its 7c, 8c, and 8cx laptop chipsets, letting Windows laptop makers drive prices well below the MacBook Air's $999 list price.
The core problem with Qualcomm's always-connected PC strategy is one that Qualcomm itself can't fix. While Qualcomm could, and probably will, soon announce a laptop chip that's based on the new Snapdragon 888 and has a level of raw power closer to Apple's M1, it's really down to Microsoft, as well as peripheral and app makers to solve the platform incompatibilities that have frustrated PC reviewers.
Hot Hardware cites Microsoft's promises of changes come in future updates to Windows 10, arguing that "with the arrival of x64 emulation and a growing library of native Arm64 apps, Windows 10 on Arm is going to be an even more powerful platform." From a performance perspective, while running Windows 10 on Arm, these [Snapdragon 8cx] chips may currently be at a disadvantage to the Apple M1, but some day in the not so distant future that might not be the case. We have no doubt that Qualcomm is likely working on a new Windows PC-centric SoC that is based on Snapdragon 888 or similar architecture. Qualcomm has promised a 25 percent uplift in CPU and a 35% lift in GPU performance over the Snapdragon 865, with the Snapdragon 888, which already offers a big boost over the previous gen Snapdragon 855/8cx. So, Qualcomm has the potential to put up a strong showing against the Apple M1, whenever its next-generation Snapdragon PC chip launches.
That may be, but John Gruber at Daring Fireball argues that currently "M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux." Those machines are just standing around in their underwear now because the M1 stole all their pants. Well, that just doesn't happen, your instincts tell you. One company, even a company like Apple, doesn't just embarrass the entire rest of a highly-competitive longstanding industry. But just because something hasn't happened — or hasn't happened in a very long while — doesn't mean it can't happen. And in this case, it just happened... M1 Macs completely upend what we can and should expect from PCs. It's a breakthrough along the lines of the iPhone itself in 2007.
No. (Score:5, Insightful)
Yes and no. (Score:3, Insightful)
No, because Apple isn’t selling their SOCs to other companies. If you want ARM and aren’t Apple, your only option is Qualcomm.
Yes and no. There is also the problem of a consumer deciding to buy an ARM Mac rather than a Qualcomm based laptop. Those considering the Qualcomm based options would seem to be those not overly concerned about legacy x86 software so the cost of switching to Mac might be low.
Re:Yes and no. (Score:5, Insightful)
Who exactly is buying Qualcomm laptops anyway?
The battery life might be awesome but, bang for buck, most would stick to cheaper AMD/Intel.
Re: (Score:3)
Who exactly is buying Qualcomm laptops anyway?
No one *YET*. The choices are very scarce, and emulation of x86 software which many people rely on is a very new thing. The problem with ARM laptops is their niche and attempt to pitch as something unique / high end make them expensive (e.g. Surface Pro X). Quite the opposite to say, tablets.
As for bang for buck, I ask you which buck. You said yourself battery life may be awesome. Given the state of CPUs in 2020 I doubt 99.9% of users would notice if their CPU was performing like a 2015 machine. The CPU has
Re: (Score:2)
It's a chicken and egg scenario. A midrange core i5 will be cheaper than a Surface Pro X so long as no other OEMs drive the unit costs down (e.g. MS pairing a Snapdragon with 16GB of LPDDR4x), or MS opens the market up to other players such as mediatek (e.g. their arm Chromebooks).
But yes, if 2015 performance is adequate then battery life and ultra-quiet is something ARM laptops will deliver. Perhaps 2021 will see Cortex-X1 narrowing the benchmark gap; IIRC the Surface Pro X uses a fork of a 3 year old ph
Re:Yes and no. (Score:4, Insightful)
I expect ARM laptops to evolve the way Tables have. Remember Windows Tablet Edition? It was basically, try to do what windows does, with a slight bend. Apple will win here, because they are solely focused on these computing devices, to the extent that they make their own silicon now. Microsoft is focused on upending AWS, the X-Box, defending Office, and the like. Apple still is a device company first. Microsoft never has been, and relied on the OEMs to make the vast majority of windows hardware. Those guys will not compete with Apple.
Re:Yes and no. (Score:5, Informative)
No it isn't. In fact, Microsoft has the most expertise at this, having an x86 emulator since the mid 90s or so. The NT4 source code has a built in x86 emulator so non-x86 architectures can run x86 code. Microsoft also bought Connectix, which acquired Insignia Software, which is well known for their SoftPC (and VirtualPC) line of products for running DOS and Windows on Macs through the 90s and naughties. You could run Windows XP on OS X through an emulated PC prior Macs going x86.
Apple also has much experience with emulators in the OS since running 68k under PowerPC since the early 90s, PowerPC code on x86 and now x86 on ARM.
It's all very mature technology now, and OS level it's even easier since it only has to run the emulator the application portions - the moment the application calls into an OS library, it switches the call stack and runs the library natively. (In the original 68k, this emulation happened for a lot of the OS as well - it wasn't all native PowerPC code. It allowed Apple to get PowerPC MacOS running quickly with very little native PowerPC code since the emulator will emulate the OS libraries as necessary, which means as Apple ported over more things from 68k to PowerPC, the system got faster).
Re:Yes and no. (Score:5, Informative)
In 1987, Acorn released an excellent 8086 MS-DOS PC emulator. In terms of speed, it emulated at about 8086 speed in the 80286 PC-ÁT era. I used it at the time to edit WordPerfect documents, DB2 files and to compile Modula 2 code on the TopSpeed compiler.
Manual [computinghistory.org.uk]
Review in PCW, January 1988 [computinghistory.org.uk]
Re: Yes and no. (Score:2)
Re: (Score:3)
I can't help but wonder what bugs/security vulnerabilities might be introduced by these emulators - the security features normally built into the silicon probably won't work the way they're intended to. Sure, the emulation is probably only meant for legacy desktop applications, but something tells me that some people, probably some IT people who don't think beyond the application layer, will be using this on servers for applications that the developer is either A) too lazy to recompile, or B) nobody has the
No security issues, Rosetta is not emulation (Score:3, Informative)
I can't help but wonder what bugs/security vulnerabilities might be introduced by these emulators
None, because it's not emulation.
The x86 instructions are all transcoded into ARM instructions ahead of time (or just ahead of time if you have a JIT).
As such there's no security issues introduced since you are not pretending to be an x86 processor executing instructions, you are the real hardware processor executing real hardware instructions and doing so within the full security the CPU offers...
This gets to t
Re: (Score:2)
Score: +1, Ironic.
Re: (Score:3, Funny)
I'm not sure what you mean by "transcode"... I'm not sure what distinction you're trying to draw.
Indeed, if you don't know what transcoding means, I could see where you would not understand the distinction.
Re: (Score:2)
No it isn't.
Yes it is, don't be silly we're talking about general acceptance of a product here, not the underlying ability for some guru to emulate something with their OS. They haven't had Windows on ARM running x86 code emulated natively as part of the OS seamless to the user as a product until October last year. Even MS's previous ARM offerings didn't simply run x86 out of the box.
Re: (Score:3, Informative)
Who exactly is buying Qualcomm laptops anyway?
My first Chromebook was ARM and it was pretty damn inexpensive. Not sure if the CPU was Qualcomm or not.
Re: (Score:2)
Those considering the Qualcomm based options would seem to be those not overly concerned about legacy x86 software so the cost of switching to Mac might be low.
That makes no sense. The Qualcomm windows laptops all run x86 software.
Re: (Score:2)
Those considering the Qualcomm based options would seem to be those not overly concerned about legacy x86 software so the cost of switching to Mac might be low.
That makes no sense. The Qualcomm windows laptops all run x86 software.
A user wants email, web browsing, web based Google apps and cloud storage, etc. They don't have legacy x86 apps. An x86 system is still an option, they are merely no locked into x86.
Also I once had an ARM Chromebook, it might have had a Qualcomm CPU.
Snapdragon-powered Chromebooks (Score:5, Informative)
The Qualcomm windows laptops all run x86 software.
There is more to the world than Windows laptops.
Snapdragon-powered Chromebooks
https://www.qualcomm.com/produ... [qualcomm.com]
Re: Yes and no. (Score:5, Funny)
Is everyone on this site just a bunch of clueless Mac users now?
No.
Everyone on this site is a bunch of well-informed Mac users now.
Re: (Score:2)
Is everyone on this site just a bunch of clueless Mac users now?
No.
Everyone on this site is a bunch of well-informed Mac users now.
Isn't it amusing how windows people think that people using Unix are clueless? I live and die in terminal, and love it when troubleshooting network issues, such as when I ask a Windows user to type ipconfig/all in a command line...
"What's a come and line?"
You understand the divide you are talking about is not between Unix/Windows but between IT/Non IT?
Re: (Score:2)
You don't understand the reference that 99% of Windows users seem to think that Mac users are clueless when it comes to computers.
Re: Yes and no. (Score:5, Insightful)
It can’t run the latest 64-bit apps, only older 32-bit apps and is restricted by the same stupid address space limitations that brings. The proper x86 support is coming but is not there yet, so right now, it can’t do anything competent which requires real horsepower, unlike these new M1 Macs.
In addition to this, the emulation fails to play some 32-bit video games outright. This indicates it likely will not work for all 32-bit applications with strong graphical requirements either.
Apple provides some of the worst backwards compatibility in the entire computing industry but right now they have Microsoft licked. Of course, rumour has it that AMD are working on ARM chips with high performance x86 compatibility - if that comes to fruition, Microsoft will be saved on the desktop at least.
Re: (Score:2)
That would be a game changer. Also, I think Nvidia must be concerned about market position. With AMD making graphics cards and processors, and Intel making processors and trying to make graphics cards, where does that leave Nvidia?
Re: (Score:2)
Something tells me they're going to be just fine, as long as CUDA remains standard and continues to outperform the competition.
Re: (Score:2)
Windows on ARM will probably flop again. There just isn't much of a compelling reason for it. Your x64 apps run slower or not at all, and the battery life gains aren't really of much interest to most people (12 hours is more than enough, 50% more than a full work day).
They are probably hoping that it takes off in tablets again, but people just don't seem interested in Windows tablets. They prefer a mobile OS with touch-first interface and mobile apps, perhaps why Microsoft wants to get Android apps running
Re: (Score:3)
and the battery life gains aren't really of much interest to most people
Also, the battery life gains work IF you have aggressive power saving. If you want to spend 12 hours watching videos via whatever route has been optimised, then sure, you get 12 hours. If you use something that uses a lot of CPU like an IDE, the battery life goes through the floor. The Macbook Pro battery life with xcode running is best measured in microseconds. Probably better with an M1, but not anything like into the "I don't need a
Re:Yes and no. (Score:5, Insightful)
and the battery life gains aren't really of much interest to most people
Also, the battery life gains work IF you have aggressive power saving. If you want to spend 12 hours watching videos via whatever route has been optimised, then sure, you get 12 hours. If you use something that uses a lot of CPU like an IDE, the battery life goes through the floor. The Macbook Pro battery life with xcode running is best measured in microseconds. Probably better with an M1, but not anything like into the "I don't need a power adapter today" levels.
This. To give some hard numbers, Apple claims 11 hours of battery life on the MacBook Pro 16". Running Xcode builds on my work laptop, I can burn almost all the way through the battery in an hour. On my 15" MacBook Pro (8:40 rated), OBS can put me at the low battery mark in... 45 minutes, IIRC.
Getting more battery life through aggressive power management only helps if you don't actually need the CPUs to be running at full power. For those of us who do, computers won't be able to go all day for people doing real work until the rated battery life is 40–80 hours. Basically, we need computers that last a week or so without charging under light use, because if we get there, then they'll *maybe* make it through one entire work day under serious use.
If these new M1 devices really do get a factor of 4 increase in battery life, then given a reasonable mix of compiling and editing code with an hour of meetings and lunch, they *might* *sometimes* make it through a single workday without charging. We really still need another factor of two on top of that to do so consistently.
Re: (Score:2)
Indeed, having needs that are 2-3 standard deviations outside the usual is always a tough spot to be in.
If Apple (or anyone else in the industry) could get that kind battery life, the
Re: (Score:2)
Indeed, having needs that are 2-3 standard deviations outside the usual is always a tough spot to be in.
I'm pretty sure it's not 2–3 standard deviations. The standard deviations are really wide. You have a solid majority of people who just do web browsing with Safari and very little else. But you have 36% who run Chrome (which uses way, way more battery). And probably at least 10% do serious stuff, whether it's Xcode, video editing, photo editing, audio recording, music composing, etc., all of which are horrible power pigs by comparison. So probably close to half are significantly above the normal
Re: (Score:2)
Getting more battery life through aggressive power management only helps if you don't actually need the CPUs to be running at full power.
I'll agree, also since the MBPs are notorious for throttling, I imagine they don't hit the full TDB since they can't dissipate the heat. While the M1 based laptops will have a speed boost, they'll be closer in power draw, and hence battery life, since the intel based ones aren't maxed out poperly.
If these new M1 devices really do get a factor of 4 increase in battery life,
Re: (Score:2)
(12 hours is more than enough, 50% more than a full work day).
A flight from London to BKK is about 12h.
The flight back is about 14h.
So: I beg to differ.
Re: (Score:2)
(12 hours is more than enough, 50% more than a full work day). A flight from London to BKK is about 12h. The flight back is about 14h.
So: I beg to differ.
That's not a very common work flow. Do you actually do that on a regular basis?
Re: (Score:2)
Are you seriously going to spend 14 hours working on a flight? And in my experience long haul flights like that have power sockets anyway now. And you can always bring a USB power pack.
Re: (Score:3)
Even when the OEM is Microsoft themselves, their support coverage is spotty. The original Surface only got 5 years of software support (2012-2017), while the Surfa
Re: (Score:3)
This certainly illustrates that Microsoft doesn’t have the attitude of “if you can’t make it better, then don’t bother”. Microsoft has often had good ideas, but frequently they feel like they could been baked a bit longer.
Apple on the hand often seems to the party, but will have refined it enough that people care. They didn’t necessarily invent the concept or manufacture the core product, but they brought it to market fully formed in a way that makes people excited. Recen
Re: (Score:3)
What was actually good about the iPod, apart form the marketing? The first few versions didn't even have USB.
They had firewire, which was superior.
Re: (Score:2)
https://www.qualcomm.com/produ... [qualcomm.com]
Re: (Score:2)
Depending on your employment situation, of course. There is no way in hell I would let Windows in my business. I value my security. Hence I have a compelling reason NOT to use it.
YMMV.
Re: (Score:2)
Color me curious. I spend the VAST majority of my time inside a web browser (using thing's like Google's G-Suite and OpenOffice) and occasionally in an Xterm window writing code. If I can do that, and if it has decent video/audio performance for when the kids hit me up for a Disney video while we're on the road, and it is as frugal with battery life as Apple is boasting....then I'm good.
My sticking point with Apple in the past has been three-fold. (1) Their devices tend to be more expensive, (2) my choi
Re: (Score:2)
Color me curious. I spend the VAST majority of my time inside a web browser (using thing's like Google's G-Suite and OpenOffice) and occasionally in an Xterm window writing code. If I can do that, and if it has decent video/audio performance for when the kids hit me up for a Disney video while we're on the road, and it is as frugal with battery life as Apple is boasting....then I'm good.
My sticking point with Apple in the past has been three-fold. (1) Their devices tend to be more expensive,
Having used Mac, Windows machines, and Linux for decades now, My experience is that while initial outlay for the Mac is a bit higher, my Macs tend to last around 50 percent longer than Windows machines. That negates any price difference.
(2) my choices for software/hardware are more limited, Once upon a time, when people built their own Windows machines, this was an issue. In present times, I find that more of a liability, as Microsoft has this nasty habit of tossing older equipment away with no recourse. While Linux especially just continues support. I discovered that once when I was putting together several dual boot machines that used USB-Serial converters. Linux side? Just worked. Operating system went out and loaded the drivers. Windows side? Nope. Turned out the Converters I was setting up were from the days of palm pilot. But it still worked.
Latest is a perfectly good HP laser printer that a Windows update dropped support for.
and (3) things only integrate well if you buy into their entire ecosystem:
Mac iPhone iPad AppleTV etc
It's great that everything works well together and the interface is similar across devices, but the cost and lock-in has given me pause.
They all work - I have Macs and iPhones, and wife has an ipad. I also use Windows and Android pads. If you are happy with your setup, and price is the most important thing to you, then it's all good. My experience is that the Apple ecosystem is superior, and I'm a Unix guy anyhow, but for some, the most important thing above all other considerations is price.
I get a lot of this, as when I have have people spec out computer systems, they look at price over performance, and it's almost impossible to get them to think otherwise.
Re: (Score:2)
... (2) my choices for software/hardware are more limited ...
Bogus claim on the software side given your Linux PC. Macs run virtually everything a Linux PC will. Nearly all open source software supports Linux, *BSD and Macs.
... (3) things only integrate well if you buy into their entire ecosystem ...
Lock-in is not required. Android based phones and tablets work fine with Macs. Just use gmail on the Mac to share the addressbook, etc. Use whatever non-Apple music service you prefer on the Mac. Similar for movies. Similar for productivity software. Etc. Whatever you are using on your Linux PC will most likely be available on the Mac too.
Re: (Score:2)
Yes and no. There is also the problem of a consumer deciding to buy an ARM Mac rather than a Qualcomm based laptop.
Nobody with a Macbook on their shortlist was ever going to buy a Qualcomm based laptop instead.
Re: (Score:2)
Yes and no. There is also the problem of a consumer deciding to buy an ARM Mac rather than a Qualcomm based laptop.
Nobody with a Macbook on their shortlist was ever going to buy a Qualcomm based laptop instead.
LOL. So no Chromebook buyer ever seriously considered a Mac. Good to know. ;-)
Re: (Score:2)
You should consider suicide.
But using a green technique, of course. That means no lead shot or hazardous chemicals.
Re: (Score:2)
No, because Apple isn’t selling their SOCs to other companies.
Yet. Perhaps there's a future where in which Windows runs on ARM just fine and Apple starts selling their silicon to PC manufacturers.
Re: (Score:2)
Apple is going to make Windows and Linux run faster? Why would they do that?
Re: (Score:2)
Re: (Score:2)
Yes, because people will just stop buying whatever they were buying from these other companies and will start buying from Apple instead.
Re: (Score:2)
While you’re correct that horse trainers didn’t directly compete against car manufacturers, if you were a horse trainer at the turn of the last century, you’d be a fool to dismiss the potential impact cars posed to your business. Granted, we have the benefit of hindsight, but it should have been obvious even to them that their primary business was supplying horses to the carriage operators who did compete against auto manufacturers, so they were effectively competing via proxy, and thus in
What lead? (Score:2)
Microsoft are screwed if ARM-based PCs take off, no matter what they do. Windows in general is already behind
Re:What lead? (Score:5, Informative)
While true that Apple and Intel are neck-and-neck right now in single core performance and that Intel has some investments that will likely pay off soon, Intel’s processors have been turning around single-digit percent performance gains for the last several generations, which is far short of the double digit percent performance gains Apple continues to be seeing. You can’t look at a graph like this one from Anandtech and not feel like the writing’s on the wall, and it’s from before the M1 was announced.
https://images.anandtech.com/d... [anandtech.com]
Re: (Score:3)
Re: (Score:2)
It's not like this is even the most threatening part of Apple anyway.
The Apple marketing machine is a much more effective weapon, at a point apple could literally sell 20-30 million of 6502 powered machines right now.
Those other machines were wearing pants? (Score:2)
Well, there’s your problem right there - thermal throttling caused by poor airflow.
There are probably more than a few supercomputers (Score:2)
that would disagree with the statement:
"That may be, but John Gruber at Daring Fireball argues that currently "M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux." "
Re: (Score:3)
Indeed. Also it's quite easy to embarrass the M1 based macs. Trivial, really. Something like "I need more than the distinctly mid range 16G of RAM". They're still selling intel based MBPs for people who need something a little beefier.
Re: (Score:2)
It is unlikely to remain 16G of RAM for long.
Re: (Score:2)
No, but the power use will go up if they can't get the PoP stuff working with more. Driving high bandwidth signals over a circuit board will draw substantially more power than what they have right now. I don't think package on package solutions will ever reach the high end with memory amounts.
Re: There are probably more than a few supercomput (Score:2)
We should be threated by TSMC (Score:5, Interesting)
Re:We should be threated by TSMC (Score:5, Interesting)
Unfortunately it takes a huge amount of investment and a very long time to set up new fabs, and the 5nm stuff is proprietary so would need to be re-developed by competitors while TSMC continues to forge ahead.
China is trying to do is, investing vast amounts of money and offering incredible salaries to try to tempt TSMC staff away to help them catch up, but even they are saying they will need 4-5 years to get to """7nm""" in the best case. Quote marks because it won't be as good as established 7nm processes, it will still be fairly new, have some limitations and lower yields.
Intel has been trying for years and is still stuck on 14nm, and their parts absolutely suck because of it.
Re: (Score:3)
Intel parts are only competitive when you don't care about battery life and are able to provide extreme cooling.
Ryzen 5000 was an absolutely flawless victory. Faster, much lower power and cooler, cheaper motherboards... Now AMD are back it's really exposing the limitations of Intel's 14nm process.
Re: (Score:2)
Re:We should be threated by TSMC (Score:4, Informative)
Their 5nm monopoly means you are bound to them
Protip, "nm" these days is a version number on one manufacturers fab process. It is NOT a unit of measure.
TSMC 5nm process chips are 48 nanometer sized transistors. Intel and AMD are both also using 48 nanometer sized transistors. There is no monopoly here.
It's not exact, but if you're implying that Intel and TSMC are on equal footing, you're overstating Intel's position a bit. Intel's 7nm process won't be shipping until late 2022 or early 2023, so we can only compare with their 10nm process (which is actually shipping parts). TSMC's 5nm process chips have 173 transistors per square mm. Intel's 10nm process gets 100.8 transistors per square mm, which is slightly better than TSMC's 7nm process, but still way behind. Samsung's 5 nm process gets only 127 transistors per square mm, which again is way behind, but not as far behind as Intel.
Does it matter? (Score:2)
That may be, but John Gruber at Daring Fireball argues that currently "M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux."
In an age of computers, where the average PC is fast enough for most mundane tasks and only gamers and such need 'fast' computers, what does it matter when someone claims whatever as a reason to enter the walled garden?
The iPhone was just a marketing breakthrough (Score:4, Insightful)
Windows mobile (not WM10, the one before that, and before Windows Phone, think like Windows Mobile 2003 and 5) had screens with double the resolution, plenty of apps (real apps, not just bookmarks to web pages) including nice complete navigation programs and much better hardware like screens with double iPhone's resolution, expandable memory (instead of 4 or 8GB max, partly taken by the OS) and so on. 2007 iPhones didn't even have 3G (!!!) and GPS (!).
Without a doubt Apple did something right indeed, or at least desirable (like apps store with 1000+ apps that make fart noises?) but there was nothing at all revolutionary.
Windows Mobile was garbage (Score:2)
It was essentially a PC interface on a tiny phone screen - it just didnt work. And for consumers, once bitten etc... The original iPhone might not have had the best hardware but it was the interface that sold it, along with appropriate lifestyle marketing to hipsters.
Re: (Score:3)
along with appropriate lifestyle marketing to hipsters.
So, I'm a hipster because I prefer an on screen "real keyboard" for sending text messages over the 9 - 10 keys of a "normal mobile phone" and because I want a phone that can do internet and has a real web browser and an real mail client?
Sorry, no idea in what world you live, but if that makes me a hipster, I will now look for fancy stuff to trim my beard.
Re: (Score:2)
Back then - yes. Most people didn't give a stuff about doing any of that on a phone in 2007 plus the 2G connection was almost useless for anything beyond basic text emails.
Re: (Score:3)
"Hipster" is the old vacant trope that Apple haters pull out to explain Apple's success. Move on.
Re: (Score:2)
Sorry mate, there's more than a grain of truth in it. Apple devotees were to a large part in media back in 2007 with the social and cultural baggage that entails. Don't like that fact? Tough, suck it up buttercup.
Re: (Score:2)
The non-touchscreen versions worked fine and those were the ones that sold. The PDA variants did suck, but those translated poorly to cell phones and were dead ends.
All the best devices at that time were non-touchscreen. Palms had touch but were so unstable they were unusable. Once you jettisoned the touch interface, though, the Windows devices were better than Symbian mostly because Symbian apps were so terrible. Of course, hipsters didn't like them, perhaps that explains your bias?
Re: (Score:2)
but there was nothing at all revolutionary.
Yes, because they made that revolution with the Newton 10 years before that. (* facepalm *)
Re: (Score:2)
LOL I don't know which was worse, the Newton itself of the morons who advocated for them. Never was there a product so singularly focused on solving the wrong problem.
Back in the day, my company went through a wave of ex-Apple coattail hires just as the Newton raged. They were everywhere and no one carrying a Newton ever accomplished anything. Within a year you never saw one. Palms, though, stuck around because they provided actual value.
If there is ever an opinion you can dismiss out of hand, it's from
Re: (Score:2)
and too big.
During that time, there were those that believed that the most important problem to solve was an interface learnable by someone who refused to make any effort to understand a computer. All those efforts were ignorant and misguided because they targeted a market that was closed minded to the solutions. Newton was one of those things.
Re: (Score:2)
Windows mobile and Win
Re: (Score:2)
You are partly correct. I had been using PDAs with Windows Mobile and 640x480 screens since around 2004 (the Toshiba e800 was my first), when the iPhone came out I had an Axim X51v, mostly for GPS when driving and watching shows while on the NY subway. So I saw those first iphones - remember they did not even support apps when they originally came out, and my first impression was that with that horrible half-VGA they were unusable for anything that I'd want to use them with, even if they could do the things
Re: (Score:2)
Windows Mobile was, indeed, best of breed at the time, but not the touchscreen version which was terrible. Problem was that the screens were small in order to allow for the full keyboards, just like other devices of the time.
The iPhone was not revolutionary in its design It wasn't even a smartphone, but the iPhone did one thing that had not been done before, it implemented a soft keyboard that worked without a stylus, allowing for the form factor that everyone wanted but no one had yet made work. Also, t
Re: (Score:2)
Without a doubt Apple did something right indeed, or at least desirable (like apps store with 1000+ apps that make fart noises?) but there was nothing at all revolutionary.
I loved my HTC Excalibur (sold to me as the T-Mobile Dash), and I even stuck it out with WinMo post-iPhone with my HTC Rodium (Touch Pro2) and HTC HD2...but there's a whole lot more to the story that got overlooked in your post.
First, let's discuss what the first generation iPhone was: It wasn't a smartphone as we know it today. It was a feature phone, done right. Forget about WinMo for a minute, and let's compare it to the Motorola Razr. Threaded text messaging wasn't a thing. Kinetic scrolling was new. Co
No. Why? (Score:3)
Apple is not threatening the PC market here. In fact, it is probably harder to use an M1 Mac for the typical Windows (and Linux) scenarios found on PC hardware. That should actually male Qualcom relax.
That said, strategically things are going in the direction of platform irrelevance, but they are nowhere near there yet.
Who should worry is Intel though. AMD has ARM experience and can mix AMD64 and ARM and produce pure ARM chips as well. Intel is sorely behind in this field, because Intel still does not really understand CPU design. At hart, Intel is still a memory maker that for a long time had superior fabrication processes exclusively available to them that compensated for their lack of a good CPU design. That fabrication process superiority is in the past now.
Re: (Score:2, Interesting)
At hart, Intel is still a memory maker that for a long time had superior fabrication processes exclusively available to them that compensated for their lack of a good CPU design. That fabrication process superiority is in the past now.
I would argue it's the other way around. Intel's design is actually pretty good. The issue is they can't fab chips at very high densities at a competitive price to AMD. Until the very recent AMD chips, Intel had them beat at per-core performance, even on chips using a five year old fabrication process. At the consumer level, performance is basically a wash for most workloads. In the datacenter, however, there is little reason to buy a $5000 32-core Intel CPU over a $4000 48-core AMD chip.
meh (Score:2)
Re: (Score:2)
Ya, if RISC-V comes up to speed, Apple would be forbidden to switch their production to RISC-V? I'll bet they already have RISC-V designs to play with, they'd be stupid not to.
Re: (Score:2)
All linux "Apps" run on Macs, too.
No idea why you do not know the most simplest things.
Intel arch has hit a wall (Score:3)
PCs will erode Apple's lead as 5nm chips become available and more optimisations/instructions get added to the ISA.
Sounds like you don't understand what makes the M1 so fast [medium.com].
The Intel x86 instruction set is at a wall where "adding more to the ISA" cannot help them.
And while Intel and AMD are working away, do you think Apple will have stopped advancing?
the real M1 miracle (Score:5, Insightful)
is just how stupidly well Rosetta2 works (Yes, yes, I know about Docker and a few others). Yes, Windows on ARM has x86 emulation but it's painfully slow and much less compatible, making Windows ARM dead on arrival. Pretty much everyone expected Apple's M1 to be fast and cool and have great battery life, but just how well Intel software runs on M1 Macs came completely out of the left field for most.
Re: (Score:2)
Docker's functionality is nothing like Rosetta 2. Were you thinking of QEMU?
Re: (Score:2)
Docker's functionality is nothing like Rosetta 2. Were you thinking of QEMU?
Docker doesn't work [medium.com] in the M1 based macs. They are working on it [docker.com].
Rosetta 2 is apparently not enough to get docker running.
M1 Macs may embarrass all other PCs BUT.... (Score:2)
Qualcomm should be threatened by Apple modems (Score:4, Insightful)
Qualcomm is greedy however. They own many other technologies that you are required to license from them for the right to have 5G on your phone; accelerometers, power management systems, etc. In these things Qualcomm has good tech but there are competing systems out there. To have 5G on your phone, you are required to license ALL of Qualcomm's patents, pay a royalty percentage of each phone's price PER PHONE, and pay a penalty to Qualcomm if you don't use tech you licensed from them, effectively creating an anti-competitive behavior for the tech where Qualcomm has actual competitors. If you want to use some other vendor's power management, well fine, but you're still paying Qualcomm $5 or $10 per phone for their system even if you're not using it. This seems egregious, but given the growing competition with China, the Feds seem unwilling to punish them for it as it plays heavily into overseas trade. Personally I don't have a problem with Qualcomm's license of a royalty per phone; Apple's argument is that the value of hte iPhone is in all the tech that goes into it, but I agree with Qualcomm in that the value of the iPhone is existentially about it's data connectivity whereas everything else is just cool features adn nice-to-have's; without the data those other techs are meaningless. The double-dip licensing though is pretty egregious. https://www.reuters.com/articl... [reuters.com]
This is the core Qualcomm business model, and the modems are the key to it. This is the crux of the Apple/Qualcomm patent fight. Apple attacked on 3 fronts; they sued in court directly, the funded Intel to make a competing tech, and it's believed they pushed the Federal government to run an anti-trust investigation into Qualcomm all at once, right when Qualcomm was vulnerable during the messy and ultimately failed takeover of NXP. The only one that mattered though was Intel; all this other stuff was just fluff if Intel couldn't get the modem to work to launch new 5G iPhones. It's no coincidence that Apple announced a settlement in the Qualcomm dispute the same day that Intel announced they were getting out of the modem business; Intel failed to produce the modem which would delay 5G iPhones and Apple couldn't have that. So they backed off, capitulated to Qualcomm's egregious licensing practices etc. https://9to5mac.com/2019/04/17... [9to5mac.com]
And then Apple bought Intel's modem division. And moved it to San Diego, 10 minutes from Qualcomm's campus. https://apnews.com/article/c77... [apnews.com]
All this other stuff between Apple and Qualcomm is just push-and-pull. Apple wants to hurt Qualcomm because they're tired of paying them a royalty per phone, and being forced into Qualcomm's supply chain. They folded on 5G, but they'll brute force it for 6G and 7G or whatever comes next. That goes right to the heart of Qualcomm's core business model.
To me, the M1 is a good processor, but if i were Qualcomm I'd see it as Apple's progression toward chip design and manufacture, and use that as an example of the modem that Apple might be able to build to compete with Qualcomm right at the core of what Qualcomm is. That's the real threat here.
Really? (Score:2)
"One company, even a company like Apple, doesn't just embarrass the entire rest of a highly-competitive longstanding industry. "
Tesla comes to mind and SpaceX.
Not at all... (Score:2)
Title posits wrong question (Score:2)
Should Microsoft worry not Qualcomm.
All these "M1 is the best" articles are crap (Score:2, Insightful)
This guy claims it's great that apple provides a chart with no numbers, no labels etc.
"M1 embarrasses all other PCs"
In what way? Care to quantify that in any way?
"M1 is fast and cool while every other CPU is slow and hot" anything to back that up other than the graph you drew in paint?
Re: (Score:3)
Most of the "M1 is awesome and everything else is crap" articles use Geekbench and a handful of Apple optimized applications like Final Cut Pro to prove their point.
Of course, when you measure the performance of applications that people actually use like Chrome or Office, the performance gains are much more narrow. And the M1 still gets trounced in most game benchmarks when pitted against PC's and laptops with discrete graphics processors that are available around the same price point.
That doesn't mean that
But will someone copy it? (Score:2)
It would be interesting to see if we get a re-run of the Intel- AMD battle if someone tries to copy the M1. I guess it would depend on how tightly locked up in patents it is, though these would only be valid in the USA , so it wouldn't be beyond the bounds of credibility for the chinese to come up with something.
Re: (Score:2)
I don't recall the exact details, but AMD had a license to make Intel chips. Intel granted it as part of an anti-trust settlement. So the IP cross-over there is quite different. Apple has no such agreements with anyone else.
Comment removed (Score:5, Interesting)
Re: (Score:3)
The current and past crop of top-notch level physicists have/had a good appreciation of philosophy and the arts. Where do you think new concepts come from?
Re: (Score:2)
The independent benchmarkers disagree with apples claims completely, and the write of the article we are commenting about is clearly compensated by apple in some way as he slathers apples cock with his sloppy word salad