INTEL is it the end?

[oldSock]

Hi everyone,

 

Post summery:

Today i am looking into ARM and the possibilities that exist around it.

Also looking into INTEL and X86's future in general.

Hypothesis of what the future of computing may look like. 

 

Let's get into it,

 

When you do a google you will find that "ARM is a family of reduced instruction set computing architectures" The short of it is, it can do a lot of things really well and can even be dedicated with custom instructions to be supper efferent both process wise and power wise. Now ARM isn't new and fact is, it was a good solution for industry long before a consumable. Anyone from Texas Instruments to NVIDIA with their Tegra APX 2500 back in 2008 is nothing new to many of you. But what does all of this mean?

 

ARM is very good at prototyping and ARM is very good at keeping cost down "relatively" not to mention it is widely used in almost anything today, we might even find one in a smart toaster soon. Point is ARM if there processing being done chances are you will find an ARM CPU doing the work. So yes stating that arm is versatile is an understatement and judging from its beginnings to where it is now it is safe to say ARM is really hitting the market hard. Apple being the most recent but not the first to deploy it in desktop and laptop. I would argue that the Raspberry Pi was ahead of its time as it was could be used as a desktop, laptop, for prototyping and things like firewalls and the like. Let's be honest for a moment Linux  and Raspberry Pi really demonstrated what is possible. So i stand to be corrected but i would say Apple based their choices of the success of the Raspberry Pi.   

 

So what does this mean? Well for developers it means a lot more freedom to create multi-platform implementations as the operating systems are becoming similar in architecture. So with its obvious limitations ARM will become document and is why Microsoft is also looking into ARM find the SOURCE here. This will open many doors "depending" on how Microsoft is going to play it.

 

But what about the X86 CPU's? Well depending on what you use your computer for, we may see the end of this technology and it will happen sooner rather then later. See it is not about the gamer, or day to day electronic consumer it is and will always be about industry. If both commercial industrial industries pushes for arm solutions "desktop,server then the private sector will follow as manufactures make their choices on money/investment and demand.

 

As demand grows for ARM the X86 will finally be out done and with it a lot of functionality might actually be lost. The reality is for a long time a "do it all" CPU was really handy but it was also stood still hitting the 64bit mark at 2003 it never really moved past that. Yes core counts and other innovations did happen. However it is been a very very long road for the X86 i think emulation will be its future until it gets dropped. This will not be a fast process but from the look of things it may actually not be a slow one as well. I think we are about to see a industry change on scale we haven't seen since IBM lost its dominance so many years ago.

 

As for my Hypothesis, i think ARM is going to take over, but we have seen this type of thing before. But what we haven't seen is the price difference. Example an INTEL NUC or mine computers may be replaced by Raspberry Pi solutions that will greatly reduce cost and make computing over all much more affordable. This is already happening as many developers is actively developing incredible solutions to work with devices like the Raspberry Pi. Also consider the NVIDIA Jetson Nano that is also a great tool. But with Microsoft looking into ARM means the adoption of ARM solutions is a reality. 

 

The bigger question here is, what does this mean for NVIDIA, AMD and INTEL? Well if they don't adapt they might find themselves out of the loop. I mean think about it, a Raspberry Pi running windows with a nice GPU while remaining cost effective sounds like something i would invest in. Just saying... But yea time will tell. 

 

 

 

 

As always thank you for reading please take care

cheers

 

[captain_to_fire]

Intel seems to be not going down without a fight even though both AMD and Apple are doing better nowadays. 

Also, even though Intel is having a lot of troubles with their transistor size not shrinking beyond 14 and 10 nm, they’re still better financially than when AMD is worse troubles before Lisa Su came. 

[mahyar]

intel might be struggling now but they have great engineers 

and if you fallow news  you know the still have some punch reserved

[AndreiArgeanu]

It's far from the end, I don't know why people just expect a whole architecture, or a multi billion dollar company to disappear when there's a little competition, like yeah, sure it'll lose some popularity, but the technology won't just fade away all of a sudden, there's potentially billions of computers out there running x86, and making them obsolete all of a sudden isn't something that's going to happen overnight.

[Seyrren]

This is not the first time Intel has been dethroned, before the core 2 duo series AMD was better as well, their CPU's far outperforming Intel's, i still remember my overclocked Sempron 2600+ it was a beast! which out performed many Intel CPU's at the time. So as i said last time this happened we ended up with the core 2 duo series which was a major innovation (both considering the IPC and spreading multicores to the desktop market achievement) and changed greatly the landscape of CPU's, i'm looking forward to see what they will come up now.

[Sarra]

As Linus said on the WAN show, if Intel is in hot water right now, it means they screwed up 6 years ago.

 

Worst case scenario, in 6 years, they will be back in the lead, and AMD will be... Well, I seriously hope competitive.

 

There's one market where ARM is less likely to become 'king', as well as Apple, and that's with gaming, and to a lesser extent DIY. ARM is trying to break into the enterprise sector, but I think that Intel will be able to at least stay in a strong second, only to AMD, with Apple... Do they even make anything that's actually considered enterprise? I know some of their crap is expensive enough to be, but I don't think it was actually designed to be used in a real server farm, other than as a workstation or whatever.

 

Anyway... Intel is a multi-billion dollar company. Being dethroned as a performance lead is a setback, but not 'the end'.

 

If AMD could recover from Bulldozer, then Intel can recover from... Well... When you have a product name that's just a can of alphabet soup that someone threw against a wall, that's tough to recover from.

 

Core i5-10210U wtf

[captain_to_fire]

8 minutes ago, Sarra said:

ARM is trying to break into the enterprise sector, but I think that Intel will be able to at least stay in a strong second, only to AMD, with Apple... Do they even make anything that's actually considered enterprise?

AMD has Ryzen Pro and Epyc. Qualcomm has a datacenter chip. If Apple wanted to get back to servers, now is probably the right time to do so since their M1 chip is performing well, they can easily scale it and bring back Xserve. 

[Sarra]

1 minute ago, like_ooh_ahh said:

AMD has Ryzen Pro and Epyc. Qualcomm has a datacenter chip. If Apple wanted to get back to servers, now is probably the right time to do so since their M1 chip is performing well, they can easily scale it and bring back Xserve. 

They could, but will they? I doubt it.

[akio123008]

Why on earth is it that when a company isn't doing well for a bit, people automatically assume it's "the end"? What do you think Intel is? A guy in an attic designing CPUs? 

 

And about the whole x86 thing, I don't think it matters as much. Worst case Intel's going to have to produce ARM cpus as well. And if companies like MS can do it, I see no reason why a company with decades in CPU design experience wouldn't be able to pull it off as well.

[Arika]

If every company disappeared as soon as someone else took the lead, there would be nothing left...

 

What is with everything thinking intel is dying?

[Laborant]

ARM has a huge advantage with having experienced personnel around the globe, developing all sorts of connected devices, like smart lamps, hi-fi-systems and so on, that often run on some sort of ARM-processor. ARM is more than just smart phones and tablets. Basically everything that needs to implement some sort of communication protocol (like Thread or MQTT) runs on an ARM-Processor (or maybe increasingly RiscV).
 

Having this in mind, there are lots of people around the globe, being sort of comfortable writing in ARM-assembly. What people do know x86 architecture well enough to optimise it? Developers at AMD, Intel and some HAL-developers at Microsoft, as well as some Linux-People. This narrows the pool of experienced employees down. 

 

What Apple did is something, no other company could achieve in the next years - they redesigned their ARM-Cores and its surrounding architecture to fit their use-case. They adapted their operating system to work together with these optimisations, allowing this Rosetta-translation alongside with other accelerators. This is unique to the whole industry. While Microsoft works together with Qualcomm, they cannot achieve this level of integration. Nvidia has some experience in designing their custom ARM-Cores (Denver) but they do not have their operating system (to see L4T as being their OS would be a far stretch, with it being an adapted Ubuntu). AMD and Intel do not have their own operating systems as well.
The Windows-Bubble (which is huge) will continue to stick with x86 for years to come. If Apple continues to pull ahead in performance, efficiency and developer comfort, Microsoft could lose in the regular customer market. Companies are way slower in jumping ships like this - imagine a corporation having to re-train all its employees for new software running on Mac. That's expensive!

A big "if" remains in Microsoft, Windows 10X and themselves developing on their own ARM-based silicon (but data centre focused as it seems). But for them being on Apple's level, it will also take years. I do not think that any country would allow Microsoft buying Qualcomm - because of Qualcomm's role in the Android-Bubble.

 

I wouldn't mind getting a Windows on ARM desktop, if the applications I need run and the performance is at least on par with the x86 counterparts. This will certainly not happen in the near future.

[8tg]

18 hours ago, AndreiArgeanu said:

It's far from the end, I don't know why people just expect a whole architecture, or a multi billion dollar company to disappear when there's a little competition, like yeah, sure it'll lose some popularity, but the technology won't just fade away all of a sudden, there's potentially billions of computers out there running x86, and making them obsolete all of a sudden isn't something that's going to happen overnight.

Sparc.

 

edit

and powerpc 

[WhitetailAni]

I can see ARM taking over in laptops.

Desktops? Not really - except for Macs.

[8tg]

1 hour ago, Caroline said:

It would be nice to have ARM based laptops, and I mean more than a bunch of prototypes, something that could work as a daily driver for school, college or regular office work where you need a CPU to be fast but also energy efficient and most importantly cheap.

Same goes for desktops, -standard- (not gaming or custom) computers would benefit from ARM: smaller designs, less power consumption, etc.

Chromebooks are what you’re looking for.

Lots of ARM options in the chromebook market, the higher end mediatek options being pretty decent performers under chromeOS or other Linux distributions.

As soon as windows has a proper ARM port and architecture compatibility tool, you’ll see a lot more of the low end market shift to ARM because it’s substantially cheaper.

[Commodus]

20 hours ago, Sarra said:

As Linus said on the WAN show, if Intel is in hot water right now, it means they screwed up 6 years ago.

 

Worst case scenario, in 6 years, they will be back in the lead, and AMD will be... Well, I seriously hope competitive.

 

There's one market where ARM is less likely to become 'king', as well as Apple, and that's with gaming, and to a lesser extent DIY. ARM is trying to break into the enterprise sector, but I think that Intel will be able to at least stay in a strong second, only to AMD, with Apple... Do they even make anything that's actually considered enterprise? I know some of their crap is expensive enough to be, but I don't think it was actually designed to be used in a real server farm, other than as a workstation or whatever.

 

Anyway... Intel is a multi-billion dollar company. Being dethroned as a performance lead is a setback, but not 'the end'.

 

If AMD could recover from Bulldozer, then Intel can recover from... Well... When you have a product name that's just a can of alphabet soup that someone threw against a wall, that's tough to recover from.

 

Core i5-10210U wtf

The concern is that this might be less of the usual shifting tide in CPUs and more of a "Microsoft dismissing the iPhone" moment, that instance where the incumbent misses the boat on an industry change and loses relevance, even if it's still a big company. Microsoft didn't die after failing in the smartphone world, but it went from the all-conquering gatekeeper of technology to just one of several tech giants.

 

While ARM is a long way from dominating computers, it's no longer a far-fetched conclusion. And if Intel can't develop a good counter to it or doesn't transition to ARM, it might be reduced to just another chip designer. In some way it already has...  arguably, Qualcomm, MediaTek and others are more influential.

[Guest]

19 hours ago, akio123008 said:

Why on earth is it that when a company isn't doing well for a bit, people automatically assume it's "the end"?

To much movies. 

19 hours ago, akio123008 said:

 

 

[AndreiArgeanu]

2 hours ago, 8tg said:

Sparc.

 

edit

and powerpc 

Sparc from what I've seen never really gained traction, not only that but it released about a whole decade after x86, and while PowerPC did gain traction, again, it released quite a long while after x86 and it never really gained support in the mainstream desktop market, the most support Microsoft gave PowerPC was with Windows NT 3.1 and Windows NT 4.0 in 1996, of course there was Apple, but they gave up on the architecture as well after a while. Either way, both of these architectures failed in a market that was flooded with x86 computers.

[Sarra]

1 hour ago, Commodus said:

The concern is that this might be less of the usual shifting tide in CPUs and more of a "Microsoft dismissing the iPhone" moment, that instance where the incumbent misses the boat on an industry change and loses relevance, even if it's still a big company. Microsoft didn't die after failing in the smartphone world, but it went from the all-conquering gatekeeper of technology to just one of several tech giants.

 

While ARM is a long way from dominating computers, it's no longer a far-fetched conclusion. And if Intel can't develop a good counter to it or doesn't transition to ARM, it might be reduced to just another chip designer. In some way it already has...  arguably, Qualcomm, MediaTek and others are more influential.

While this is a valid point, I don't really see ARM replacing the bread and butter chips, like those used in enterprise servers, nor in HEDT applications.

 

Intel has sat at the top for so long, and made so much money from especially enterprise, as well as the prebuilt market, I don't think they want to lose that.

 

If ARM starts making products that actually make sense in the low or mid range prebuilt market, and Dell, HP, etc., start implementing them, plus the massive chunk of market share that AMD has been rather nicely carving out, Intel may stop being the industry leader, but I don't think it would lose relevance.

 

But, I don't think it's going to really have a 'Microsoft dismissing the iPhone' moment, at least, not yet.

 

I seriously doubt that ARM will EVER have a place in the HEDT, high, mid, or even low end gaming PC markets, and especially not in productivity/creativity markets, just by the basic concept of what ARM really is: a low power, decently efficient, easy to manufacture chip. I also see AMD making chips, both CPU's and GPU's, for the PS5 and the Series X, and again, I really doubt that ARM would ever had a place in that kind of market, either, though it's ironic to me that Microsoft went with AMD for the Series X instead of Intel.

 

My personal hope is that ARM sticks to mobile markets, AMD becomes a real thorn in Intel's side, and they have to actually innovate to stay relevant, instead of coasting on a Tick-Tock cycled, as they have been doing until 10th gen.

[Zodiark1593]

This kind of reminds me of the time on Anandtech Forums where someone attempted to argue with me that it was the end of separate graphics cards. This was about a decade ago. 
 

Intel’s been in worse positions before, namely right at about when the Athlon 64 launched. And lets not forget that AMD was in a very bad place during the Bulldozer days. 
 

Conroe and Ryzen show us that these two giants can pull some fancy rabbits when their backs are against a wall. Some recent stuff is showing that Intel hasn’t been sitting around the past couple years. Get some talented engineers, and executives that firmly separate financials from the engineering (because it’s not the engineers’ job to care about corporate finances), and miracles can be made. 

[oldSock]

7 hours ago, fpo said:

To much movies. 

 

Actually no... See a few things are factually happening. Intel is in a bit of a spot to the point that they are looking at outsourcing and factually is being pushed to do so. It is not just in the movies where this stuff happens. In real life if a company doesn't do well investors jump ship, they always have and always will. 

[Commodus]

6 hours ago, Sarra said:

While this is a valid point, I don't really see ARM replacing the bread and butter chips, like those used in enterprise servers, nor in HEDT applications.

 

Intel has sat at the top for so long, and made so much money from especially enterprise, as well as the prebuilt market, I don't think they want to lose that.

 

If ARM starts making products that actually make sense in the low or mid range prebuilt market, and Dell, HP, etc., start implementing them, plus the massive chunk of market share that AMD has been rather nicely carving out, Intel may stop being the industry leader, but I don't think it would lose relevance.

 

But, I don't think it's going to really have a 'Microsoft dismissing the iPhone' moment, at least, not yet.

 

I seriously doubt that ARM will EVER have a place in the HEDT, high, mid, or even low end gaming PC markets, and especially not in productivity/creativity markets, just by the basic concept of what ARM really is: a low power, decently efficient, easy to manufacture chip. I also see AMD making chips, both CPU's and GPU's, for the PS5 and the Series X, and again, I really doubt that ARM would ever had a place in that kind of market, either, though it's ironic to me that Microsoft went with AMD for the Series X instead of Intel.

 

My personal hope is that ARM sticks to mobile markets, AMD becomes a real thorn in Intel's side, and they have to actually innovate to stay relevant, instead of coasting on a Tick-Tock cycled, as they have been doing until 10th gen.

I wouldn't be so sure about HEDT and servers. You could be right, but we also haven't seen much effort to tackle that space, and likely won't until Apple offers an ARM-based Mac Pro (for HEDT, clearly). I imagine ARM being helpful for enterprise servers, though — massively parallel computing with modest power budgets is rather appealing.

 

On gaming and the low end you likely have a point, at least for now. However, we're already seeing Apple tackle productivity and creativity.... and you now have a $999 ultraportable that can handle complex image and 4K video edits better than computers (Mac or Windows) costing twice as much. I don't think Apple will necessarily up-end the entire industry, but I do expect ARM-based iMacs and Mac Pros to make a stronger case for creative tasks.

 

It's safe to say AMD and Intel will both have to work considerably harder than in the past (moreso Intel, of course). They've had to compete against each other, but they've never really had even the hint of an existential threat until now.

[Sarra]

3 hours ago, Commodus said:

I wouldn't be so sure about HEDT and servers. You could be right, but we also haven't seen much effort to tackle that space, and likely won't until Apple offers an ARM-based Mac Pro (for HEDT, clearly). I imagine ARM being helpful for enterprise servers, though — massively parallel computing with modest power budgets is rather appealing.

 

On gaming and the low end you likely have a point, at least for now. However, we're already seeing Apple tackle productivity and creativity.... and you now have a $999 ultraportable that can handle complex image and 4K video edits better than computers (Mac or Windows) costing twice as much. I don't think Apple will necessarily up-end the entire industry, but I do expect ARM-based iMacs and Mac Pros to make a stronger case for creative tasks.

 

It's safe to say AMD and Intel will both have to work considerably harder than in the past (moreso Intel, of course). They've had to compete against each other, but they've never really had even the hint of an existential threat until now.

While Apple is a force to be reckoned with, it has absolutely no influence on the HEDT market. Most people who are Apple fans will use whatever Apple pumps out, while Apple detractors wouldn't use an Apple product even if it had significantly better performance, used less power, and was cheaper. Why? Because moving from a Windows or Linux ecosystem to Apple is really expensive, and because like a few scenes in movies, getting those three things from Apple at the same time would basically be a figment of your imagination.

 

I can see some enterprise systems switching over to ARM, but I am sure there are a lot of cases where ARM doesn't make sense, and unless they can actually compete with high performance chips, which would be impossible, they aren't going to really take over. What I do see is AMD making Epyc look so attractive to Enterprise that ARM becomes a hard sell. If AMD can improve the performance, and reduce the power consumption, especially in a constant load workspace, there's a pretty big chance that ARM won't find a leg to stand on.

 

For gaming... I really just don't see it. Look at Cyberpunk. Look at Flight Simulator. Those games will never perform well on an ARM solution. Not to mention, real FPS games would just never work. If ARM takes over, and Intel and AMD leave the market, competitive FPS gaming won't be a thing anymore.

 

ARM is going to have to figure out a way to push considerable performance from silicon that is by design low power. As manufacturing processes run out of nanometers to trim, efficiency per performance metrics are going to hit a wall, and unless new manufacturing tech can be realized, ARM's total performance potential is really limited. Gaming is notorious for being slow in adopting multi-threading capacity, and even in a large data center, there's a limit to space, and eventually, more performance is going to be required than ARM can deliver, and it's going to (well, I hope) only remain as a niche outside of mobiles.

 

I'm pretty sure that as long as ARM focuses on a single market, they can retain their dominance. Though, it would be nice to see them force Intel and AMD to make better, faster, and more efficient chips for mobile.

[unsorted]

On 1/3/2021 at 12:34 AM, oldSock said:

The bigger question here is, what does this mean for NVIDIA, AMD and INTEL?

 

 

https://nvidianews.nvidia.com/news/nvidia-to-acquire-arm-for-40-billion-creating-worlds-premier-computing-company-for-the-age-of-ai/

 

Nvidia has already made their move. Anyone who wants to license this stuff will go through Nvidia and feed Nvidia.

<content removed by forum staff>

 

Edited January 5, 2021 by SansVarnic
Removed content.

[SansVarnic]

-= Topic Cleaned =-

Keep thread On Topic.

[oldSock]

All Good, Lets say ARM will not get into the server race until a much later date, however a lot of server grade components does contain ARM already like routers switches and so on. Not all of them but some of them.

 

See this is where Soft solutions get into the mix. If a company can buy a box configure it to do whatever then chances are that ARM will be found in "some" of those soft solutions. The fact is that a lot of infrastructure used to use low level hardware. This however is changing as even a switch has a TON of work and components that can now be upgraded. This is not new for many years now. BUT it does open the door to ARM in the server grade market. 

 

Another important thing is the death of the laptop. See Apple made a massive statement with their ARM CPUs and that is that it CAN do desktop work. How i see the future is a phone being plugged into a charging dock that will activate more functionality because it has constant power. That means the phone becomes the laptop. We have seen this happen to a degree. But i think manufacturers may include a second CPU that is dormant when on battery but will get used once plugged into its docking station. 

 

Any company will look at this and nod their head because it will work with cloud on so many levels that it is downright scary. You can even encrypt the phone to a point that some functions will only work when docked at the work place. There is so much potential here. I think it is where all this is pointing towards. 

 

Yes it is early days but i am almost 100% sure this is the future as it makes financial sense "to a degree" Right now many employees are given phones,laptops and desktops. Then there is infrastructure cost. If companies can eliminate 40% of that or even higher you can be sure they will role out something like this. Most jobs don't need very powerful computers it is just fact. So i can see something like this happening.