What will be the PC response to Apple's M1 ?

[dmccloud]

Once Intel is running on 5nm then the performance advantage will evaporate. We live in a Windows centric world so it’s great for Apple to have a lead but the incompatibilities of these ARM machines mean they’re not useful to most normal users. Mac OS X will continue to be a niche product, even more so due to its inability to run Windows x64 too.

Here's the thing: Intel can't even get 10nm locked in, let alone get any 7nm parts even manufactured. By the time Intel gets down to 5nm, Apple will have moved on to TSMCs 3nm or 2nm processes. And let's be honest here - Windows in any form actually consitutes a smaller segment of the overall computing market than it did even a decade ago. There are a LOT of people worldwide that are using either Android, iOS/iPad OS, or Chrome OS as their primary computing platform. You also have people using some sort of Linux distro as well. Microsoft has been more of a software and services company than a pure OS manufacturer for well over a decade at this point, so deemphasizing the OS and refocusing on subscriptions and SaaS (Microsoft 365, Game Pass Ultimate, Azure) would not have a significant impact on their bottom line financially.

 

[thunng8]

LOL, this is the TL;DW (too long; didn't watch) summary:

"The MacBook Pro is objectively faster than the Dell model, but my advice is to buy the slower, crappier Dell."

Did you even watch the video? The video did not such thing. He just says if you are after a Windows laptop, the Dell is still a great laptop and the best option for Windows - obviously the mac with m1 is significantly faster (sometimes more than double) for all tasks and is a game changer if you are open to buying a mac.

 

[SlCKB0Y]

Apple is spending lots of transistor budget on large System cache ( and large l2 caches ) and other stuff. And zero on a 4G/LTE/5G Modem. There is a decent chunk in Qualcomm's chip that Apple just doesn't do right now.

The only real advantage to having these types of modems on the SoC is due to the high premium of board realestate of the devices (phones) that these SoCs go into and the simplicity of having an off the shelf all-in-one solution. The performance cost of off-chip network controllers are orders of magnitude less than the cost associated with having other components (Eg. memory) off-chip. It's basically negligible.

If you are designing a SoC/package to go into a device with a significantly larger board (Eg. laptop) I don't see the need to take up chip/package space with a modem when you could dedicate that to extra cores, memory or other functionality that is highly dependant on latency and/or the speed of interconnectivity.

 

[SlCKB0Y]

And let's be honest here - Windows in any form actually consitutes a smaller segment of the overall computing market than it did even a decade ago.

The vast majority of the compute devices in the world are either Linux-based (data centre servers - from bare metal to cloud like AWS, Google, Oracle, IBM etc and Android phones, ChromeOS, embedded devices such as TVs), or BSD-based (all of Apples devices, enterprise network and storage devices by Juniper and Dell and gaming consoles like PS3/4 and Nintendo Switch).

Microsoft has most of the desktop market and Xbox. In terms of servers it's realistically limited to SME on-premise and Azure.

Also because of the permissive licensing of FreeBSD, even Microsoft used BSD code in the Windows family of OSs, at least until Windows 2000, particularly in the TCP/IP stack. Hotmail's thousands of servers ran FreeBSD when Microsoft acquired them and continued to run BSD for a number of years post acquisition. The rumour at the time was when they finally did migrate to Windows server, it was so much less efficient they needed double the amount of servers for the same workload.

The Linux and BSD families of OSs essentially run the computing world.

 

[SlCKB0Y]

Stuff like gaming (Nvidia GeForce, Google Stadia, online Xbox Game Pass). VR like Oculus Quest 2.

At some point in the not too distant future, a quality user experience will have less to do with the specs of the device in your hand versus the cloud computing server support that those devices connect to.

Yes and no. For all of the streaming media services you mentioned sure, but Gaming as a service is still heavily dependant on low latency for a good experience and even in highly developed countries like the USA, there are vast swathes of the populace outside of major metro areas who still do not have internet connectivity which is suitable for these types of services.

Densely populated countries with very good internet infrastructure such as Singapore, Japan, South Korea - Sure. But otherwise, unless they change the underlying technology in a fundamental way, desktop gaming isn't going anywhere for a long time.

 

[SlCKB0Y]

The first Commodore Amiga was in many ways better than pretty much any PC or Mac of its time. Where is the Amiga now?

In pretty much every way it was technologically superior, both in terms of hardware and OS and accomplished this whilst being much more resource efficient. By that point in time though, they had very little chance against the giant partnership of IBM and Microsoft and suffering from a lack of serious business applications it failed to break into the business market and was relegated mostly to the gaming market. From there, a combination of terrible mismanagement under the ownership of Commodore and the release of the much more affordable 16bit consoles by Nintendo, Sega and others, it quickly lost marketshare in the gaming sector.

It also never really gained a strong foothold outside of the PAL regions of the UK, parts of Europe and Australia.

But it's legacy is that it is one of the single most influential computers of all time and there still exists a very small but absolutely hardcore Amiga user base to this day. There are still new Amiga upgrades and accessories being released all the time.

 

[polyphenol]

Ryzen is everywhere in the DIY market. They still are having an uphill battle getting into the prebuilt market, and that's where the sales/profit is. Intel can provide more discounts (or kickbacks, they've been caught before), and more chips (they still have bigger capacity than AMD) to the PC brands than AMD can.

Though both Dell and HP offer several Ryzen models in their ranges classed as "business".

My experience suggests that, regardless the manufacturer of the processor, many office workers have been caring an awful lot more about noise than absolute performance. If the box on the desk in front of them is effectively silent, they are happier than if it regularly goes into take-off mode.

For that reason, M1-like technologies will be very much appreciated. Whoever they come from.

 

[SlCKB0Y]

here is actually Intel's response ... not surprising, what else were they going to say?

Intel EVP on Apple testing new chip design: ‘We feel very good with where we are competitively’​

Intel EVP on Apple testing new chip design: ‘We feel very good with where we are competitively’

Gregory Bryant, Intel EVP and General Manager of the Client Computing Group joins Yahoo Finance Live to discuss Intel’s transformation strategy for 2021 and break down how the company is setting itself apart from competitors.

finance.yahoo.com

 

[polyphenol]

Whilst being cross-examined at Ward's trial, Rice-Davies made a riposte which has since become famous. When James Burge, the defence counsel, pointed out that Lord Astor denied an affair or even having met her, she dismissed this, giggling "Well he would, wouldn't he?" (often misquoted "Well he would say that, wouldn't he?")

Not a lot of real substance there.

 

[SlCKB0Y]

There's been plenty of times different companies have come to work together to standardize things, USB, Displayport, codecs, etc. If a common architecture is agreed upon, the common software could be built for it.

I would say that for every time companies have managed to reach a consensus on a standardised format, there are a much greater number of instances where consortiums of companies have released competing technologies with one (not necessarily the technologically superior solution) withering and dying, usually at the expense of early adopter consumers. It's almost always messy.

These "format wars" have been going on forever (Railway gauges, AC vs DC power, NTSC vs PAL, VHS vs Betamax).

 

[polyphenol]

I would say that for every time companies have managed to reach a consensus on a standardised format, there are a much greater number of instances where consortiums of companies have released competing technologies with one (not necessarily the technologically superior solution) withering and dying, usually at the expense of early adopter consumers. It's almost always messy.

These "format wars" have been going on forever (Railway gauges, AC vs DC power, NTSC vs PAL, VHS vs Betamax).

Even if there is an agreed standard, all too often one company (whether or not a member of the consortium) goes off and adds to or changes a standard.

 

[SlCKB0Y]

Even if there is an agreed standard, all too often one company (whether or not a member of the consortium) goes off and adds to or changes a standard.

A good example of this is that once Microsoft achieved basically a web browser monopoly via its anti-competitive behaviour against Netscape and others, it attempted to co-opt the HTML standard by adding support for many non-standard features around the period of IE 3 and 4.

To achieve this monopoly they basically extorted OEMs to not include Netscape by default on the PCs they sold under the threat of having their Windows licensing contracts revoked. They also extended this to other competing technology such as Java and resulted in the major antitrust lawsuit United States v. Microsoft Corp. in 2001.

Anyway, getting a bit off-topic but I thought it might be interesting to some of the younger folks here. ?

 

[Serban55]

First response

Microsoft reportedly designing its own ARM-based chips for servers and Surface PCs

Custom chips will reportedly power Microsoft’s cloud services.

www.theverge.com

cant wait for competition

 

[thenewperson]

Anyway, getting a bit off-topic but I thought it might be interesting to some of the younger folks here.

It's always useful because I've noticed a lot of people think just including I.E. by default alone was what led to MS being hit by antitrust stuff.

 

[deconstruct60]

First response

Microsoft reportedly designing its own ARM-based chips for servers and Surface PCs

Custom chips will reportedly power Microsoft’s cloud services.

www.theverge.com

cant wait for competition

The "and Surface" there is likely more hand waving than substance. ( or pretty similar to SQ2 , SQ1 being independent products. )

That's for server ARM chips. That project is being run out of the Azure Business group.
The competition is Amazon Graviton 2. ( the upcoming Ampere Altra CPU rolling out to large cloud competitors. and with semi-custom Thunder X3 coming. )


Everyone who is a large player is in the cloud services space is probably going to have a substantive chunk of ARM based servers in their datacenter in 3 years. This is primary Microsoft just "keeping up with the Joneses".

Maybe something will fall out later for the Surface, but the primary design goals are likely to be substantively different.

More likely Microsoft will get some other vendor(s) to lay some better foundation here. Whether Qualcomm puts more resources behind their effort or Samsung cleans up their act or someone else comes up with a PC laptop focused solution. Strategically for Windows it would be far better if they fostered some generally available solutions so that their Windows "partners" could weave them into solutions also. Microsoft building a laptop chip that you have to buy to run Windows 10 on ARM would be bad long term strategically. That isn't the business they have. Microsoft far more so needs to 'boot strap' a more general solution. That would best be done by working on a couple of solutions with partners. The major problem so far is that only working with Qualcomm. ( who is just as much as interested (if not more ) in selling cellular radios are a mid-range PC SoC. )

If Qualcomm and Samsung just did larger variations on their top end smartphones with four Cortex-X1 cores ( instead of the one X1 in the Snapdragon 888) there would be something competitive enough for Windows 10. wouldn't be faster than M1 , but if Windows ran natively that would be a large upside to those firmly committed to Windows 10.


The real large gap that Microsoft is missing with Windows on ARM is not Surface. It is desktops and workstations. (and/or large , heavier "desktop replacement" laptop. ) . They could try to 'stop gap' that hole by stuffing some server packages into a box with slots, but that kind of missing the point. They are going to need a developer ecosystem to drive the growth on ARM servers. Doing that on lightest of light , max battery life laptops isn't the more productive path forward.

 

[FlyingTexan]

PC won’t be far behind. Microsoft has been secretly developing Windows Core. Windows made for all platforms a leading to ARM development

 

[Internaut]

The "and Surface" there is likely more hand waving than substance. ( or pretty similar to SQ2 , SQ1 being independent products. )

That's for server ARM chips. That project is being run out of the Azure Business group.
The competition is Amazon Graviton 2. ( the upcoming Ampere Altra CPU rolling out to large cloud competitors. and with semi-custom Thunder X3 coming. )


Everyone who is a large player is in the cloud services space is probably going to have a substantive chunk of ARM based servers in their datacenter in 3 years. This is primary Microsoft just "keeping up with the Joneses".

Maybe something will fall out later for the Surface, but the primary design goals are likely to be substantively different.

More likely Microsoft will get some other vendor(s) to lay some better foundation here. Whether Qualcomm puts more resources behind their effort or Samsung cleans up their act or someone else comes up with a PC laptop focused solution. Strategically for Windows it would be far better if they fostered some generally available solutions so that their Windows "partners" could weave them into solutions also. Microsoft building a laptop chip that you have to buy to run Windows 10 on ARM would be bad long term strategically. That isn't the business they have. Microsoft far more so needs to 'boot strap' a more general solution. That would best be done by working on a couple of solutions with partners. The major problem so far is that only working with Qualcomm. ( who is just as much as interested (if not more ) in selling cellular radios are a mid-range PC SoC. )

If Qualcomm and Samsung just did larger variations on their top end smartphones with four Cortex-X1 cores ( instead of the one X1 in the Snapdragon 888) there would be something competitive enough for Windows 10. wouldn't be faster than M1 , but if Windows ran natively that would be a large upside to those firmly committed to Windows 10.


The real large gap that Microsoft is missing with Windows on ARM is not Surface. It is desktops and workstations. (and/or large , heavier "desktop replacement" laptop. ) . They could try to 'stop gap' that hole by stuffing some server packages into a box with slots, but that kind of missing the point. They are going to need a developer ecosystem to drive the growth on ARM servers. Doing that on lightest of light , max battery life laptops isn't the more productive path forward.

I agree with this. The energy cost savings moving towards ARM based data centres are going to be huge. As for laptops and tablets, good enough will be good enough. There’s no short term need to beat Apple’s MX performance. There is an urgent need for something a lot better than current efforts in the Windows/ARM space.

 

[tdar]

I agree with this. The energy cost savings moving towards ARM based data centres are going to be huge. As for laptops and tablets, good enough will be good enough. There’s no short term need to beat Apple’s MX performance. There is an urgent need for something a lot better than current efforts in the Windows/ARM space.

The 2 major cloud service providers-Amazon and Microsoft Azure are already using and promoting arm based instances. Amazon for example says that they offer 4to1 efficiency over a x386 instance. Interesting to note, both companies are building their own chips ( having them fabed) and their own servers.

 

[SlCKB0Y]

Tighter integration is not an advantage. It's a compromise. The integrated system runs more efficiently and more smoothly, but it's rarely the exact system you wanted. You bought the closest match the manufacturer was willing to sell, but you probably had to pay extra for features you didn't need.

I disagree with this and think you have it backwards. The biggest issues facing Intel and the x86/64 ISA on the hardware side and Windows on the software side is their feature bloat due them both being "jack of all trades, master of none" and the fact that they both have to maintain compatibility for legacy features stretching back decades. It's with these products that people are paying for features they don't need.

Apple has basically been able to start completely unencumbered by this and if anything, the compromise is that they are lacking features that some people might need for niche applications because of their laser focus on efficiency.

 

[SlCKB0Y]

The energy cost savings moving towards ARM based data centres are going to be huge.

Yes, it's going to be a game-changer - and not just in terms of raw power consumption but also heat management/cooling systems required in the DC.

 

[SlCKB0Y]

I've noticed a lot of people think just including I.E. by default alone was what led to MS being hit by antitrust stuff.

Absolutely. I've been in IT for a long time and anyone younger than maybe 25 or 30 probably doesn't appreciate the degree to which Microsoft was engaged in unscrupulous, unethical and downright illegal behaviour around the mid 1990s and therefore why they were so hated.

Whilst they have rehabilitated their public image drastically with regards to their consumer based software and services, anyone who deals with them in an enterprise capacity for things like SPLA licensing can still see examples now and then of their mafioso style behaviour. ?

 

[Argon_]

Absolutely. I've been in IT for a long time and anyone younger than maybe 25 or 30 probably doesn't appreciate the degree to which Microsoft was engaged in unscrupulous, unethical and downright illegal behaviour around the mid 1990s and therefore why they were so hated.

Whilst they have rehabilitated their public image drastically with regards to their consumer based software and services, anyone who deals with them in an enterprise capacity for things like SPLA licensing can still see examples now and then of their mafioso style behaviour. ?

Vista may go down in history as the turning point when Apple started clawing out a bigger chunk of the market share.

 

[SlCKB0Y]

Vista may go down in history as the turning point when Apple started clawing out a bigger chunk of the market share.

Yep, well whilst Microsoft was releasing the much maligned Windows Vista, Apple was in the process of completing their migration to Intel hardware (setting themselves up for the next decade), releasing the critically lauded OS X Leopard and launching the revolutionary iPhone, which completely disrupted both the phone and the computing industry, the effects of which are still being felt with the introduction of the M1 Macs.

It was a pretty good year for Apple. ?

 

[JeepGuy]

Vista may go down in history as the turning point when Apple started clawing out a bigger chunk of the market share.

I think, apple moving to intel, and introducing bootcamp had a bigger impact.

 

[Joelist]

You have to differentiate between what WILL be the PC response and what SHOULD be the PC response.

M1 did not drop out of the sky. It has been coming for a long time; roughly since the A Series stopped using ARM designs and Apple went to only using the ISA but having their own microarchitecture. Let's remember that in the mobile world the A Series has been running rings around all the other SoCs for years and that M1 is a child of the A Series. So M1 is not just blowing away the x86 competition but it also smokes ARM too.

So what SHOULD the PC response be? To really bear down on efficiency. To work on and either solve the inherent limitations in CISC and x86 that stop them from using an 8 wide architecture like Apple or create a new ISA that is not hamstrung like this. To come up with common frameworks that allow PCs to take advantage of things like DSP blocks, Machine Learning and other functional divisions. And by common frameworks it means all current PC development languages and environments will need to change also.

What WILL they do? Well, one thing to note is there is no monolithic "PC". It is a bunch of different companies doing different things and competing against each other. So a "PC" response is not going to match M1 because the fragmented nature of the PC world means it cannot and still be able to work on the wide variety of hardware and software. What it MAY do is look for ways to improve PPW which is a big challenge.