Editorial x86 Lacks Innovation, Arm is Catching up. Enough to Replace the Giant?

[bug]

Yes, we are speaking about consumer apps, he shifts to offtopic workstations...

It's not offtopic. You want consumer apps for ARM, somebody has to compile them. So there has to be ARM-based workstations. Because cross-compilation is tedious, slow and pretty error prone.

 

[ARF]

They have already been compiled. 2.87 million apps currently in Google Play Store alone.

 

[notb]

Yes, we are speaking about consumer apps, he shifts to offtopic workstations...

You're making the fundamental logical error here. You're saying: people spend most time on smartphones, so they should be fine with an ARM laptop.
For some reason you can't grasp the fact that they use laptops precisely because they can't do something on their smartphones.

IMO you're a little confused when it comes to how consumers use their PCs. You probably think it's just browsing web, watching movies, listening to music and using communicators.

Sooner or later probably every consumer will run into a problem, because a program he uses on x86 has no ARM version.
ARM is not a usable consumer platform. It could be, but it isn't.

 

[bug]

They have already been compiled. 2.87 million apps currently in Google Play Store alone.

So workstations when talking ARM computers are off topic, but Android/Google Play Store isn't. Noted.

 

[trparky]

They have already been compiled. 2.87 million apps currently in Google Play Store alone.

There is a big damn difference between an app and what most people, even regular users, expect full blown programs to be like on a desktop platform. This is essentially why Windows 8.x failed so badly on the desktop, Microsoft underestimated what people would be doing with the desktops and so are you. People don't want apps; they want full programs with full functionality on a desktop.

By the way, did you work at Microsoft during the creation of Windows 8.x? Sure sounds like you did.

 

[bug]

There is a big damn difference between an app and what most people, even regular users, expect full blown programs to be like on a desktop platform. This is essentially why Windows 8.x failed so badly on the desktop, Microsoft underestimated what people would be doing with the desktops and so are you. People don't want apps; they want full programs with full functionality on a desktop.

By the way, did you work at Microsoft during the creation of Windows 8.x? Sure sounds like you did.

I doubt he worked anywhere so far. He seems to be a kid that is just learning about computers. Or a student.

 

[trparky]

I doubt he worked anywhere so far. He seems to be a kid that is just learning about computers. Or a student.

It was meant as a joke that poked fun at Microsoft's stupidity.

 

[bug]

It was meant as a joke that poked fun at Microsoft's stupidity.

I got that. I was just saying, don't make fun of him. Try to help him instead. Not much is getting through, though

 

[trparky]

I that. I was just saying, don't make fun of him. Try to help him instead. Not much is getting through, though

No, not at all.

 

[ARF]

So many trolls here, a moderator should come and clean all the nonsense.

 

[trparky]

So many trolls here, a moderator should come and clean all the nonsense.

No, you're the damn troll here for not listening to people who obviously have far more experience in the PC ecosystem than you are clearly showing. I've been around computers for nearly 20 years and I will tell you one damn thing, even normal people won't accept mini-apps as you seem to keep claiming that people will accept. Look no further than the failure of Windows UWP apps and the UWP platform. It died because Microsoft thought just like you are now and they lost billions.

Run along child, let the grownups have the floor.

 

[mtcn77]

Whenever someone hypes arm, I remember when jim keller told they thought they found a bug in the arm processor(that turned out there was a bug in their verification tool)@3:00.

 

[bug]

So many trolls here, a moderator should come and clean all the nonsense.

Feel free to bring this to a moderator's attention.

 

[ARF]

No, you're the damn troll here for not listening to people who obviously have far more experience in the PC ecosystem than you are clearly showing. I've been around computers for nearly 20 years and I will tell you one damn thing, even normal people won't accept mini-apps as you seem to keep claiming that people will accept. Look no further than the failure of Windows UWP apps and the UWP platform. It died because Microsoft thought just like you are now and they lost billions.

Run along child, let the grownups have the floor.

You violate the forums' rules. Address the opinion, not the member. You are reported to the moderators for trolling.

 

[trparky]

I don't feel like I'm violating the rules at all, I'm just telling you how wrong you really are. If you call me a troll, I'm going to call you a troll right back. Facts are facts man, you may not like the facts but that doesn't change the fact that the facts are facts. You have opinions, they aren't facts based upon 20 years of PC industry growth knowledge like many of us here have that you don't have. Many of us here have more knowledge about how the industry works in our pinky finger than you have in your whole brain. Why? Because many of us here either work in the industry, have been around the industry for a long time, or just are plain enthusiasts.

And not only that but come on dude, you sign up two months ago and then you waltz into this place and tell us who have been here for years that we're wrong and expect us not to react like this. Really? Go home, you're drunk.

But I'm done with you. Keep spouting off the same bullshit that Microsoft lost billons over. They too thought the same way you are now, they lost big money. Now Microsoft is embracing the desktop yet again because even now they learned that the PC is king, it will always be king. Long live the desktop! PC MASTER RACE!!!

 

[Kursah]

Let's be done with the petty drama here folks. Anyone wants to push it past this post will be met with infraction points.

 

[mtcn77]

I like living dangerously.

 

[ARF]

A new Premiere Pro Beta with AMD and Nvidia hardware encoding.
Intel no longer has a lead over AMD with Premiere and Quick Sync.

 

[notb]

A new Premiere Pro Beta with AMD and Nvidia hardware encoding.
Intel no longer has a lead over AMD with Premiere and Quick Sync.

You don't understand this article or you didn't read it. Which one is it?

Adobe added NVENC support, i.e. hardware-accelerated encoding for Nvidia GPUs. This article is about NVENC.
All red bars are for encoding using Nvidia GPU. Blue are for the CPU. Green are for the Intel IGP.

Any questions?

 

[Chehh984]

I would suggest going back a bit to a time when RISC and CISC were the options. There is a reason one of these won in personal computers. That reason is still valid today, ARM has great advantages which it uses really well in mobile world, but suggesting it could replace x86 in all spheres is just pure speculation. If ARM would be able to do everything x86 does than it would not be ARM anymore....

 

[ARF]

I would suggest going back a bit to a time when RISC and CISC were the options. There is a reason one of these won in personal computers. That reason is still valid today, ARM has great advantages which it uses really well in mobile world, but suggesting it could replace x86 in all spheres is just pure speculation. If ARM would be able to do everything x86 does than it would not be ARM anymore....

Huh, since Intel can't do anything on a more modern process than 14nm, since AMD's tremendous competitive advantages and consequences from that Intel will only have more troubles going forward, I would not claim with such certainty what is a speculation and what not.

ARM didn't exist when Intel made the first x86 processors back 40-50 years ago.
Things change and they change pretty fast.

Do you know what x86 actually is. It's basically the 1978 version, just with added some width here and there, more instructions which take transistor budget, and multiplied by more cores.
There is no innovation there, just pure evolution, and very slow at Intel's pace.

 

[londiste]

x86 and ARM are not that far apart. x86 started with 8086 in 1978. ARM was founded in 1990 but first ARM CPU was in 1985.

ARM isn't that much different from what you describe - some added width here and there, more instructions which take transistor budget and multiplied by more cores.
In practice, that is hugely simplified way of looking at it. There are major changes in microarchitecture of them both even if ISA is largely the same.

 

[Aquinus]

ARM didn't exist when Intel made the first x86 processors back 40-50 years ago.

No, but x86 was definitely not the only ISA out there at the time and RISC ISAs like MIPS and SPARC were being introduced around the same time 32-bit x86 CPUs were showing up in the mid 80s.

Do you know what x86 actually is. It's basically the 1978 version, just with added some width here and there, more instructions which take transistor budget, and multiplied by more cores.
There is no innovation there, just pure evolution, and very slow at Intel's pace.

There have been a lot of changes since the 8080. Just because core instructions haven't changed doesn't mean the rest of the CPU hasn't.

Edit: One of the things that makes x86 a potent option are the extension to x86; dedicated hardware for tasks that would otherwise take a boatload of clock cycles to accomplish otherwise. This is why vector extensions exist and why you don't just call things like add and multiply a bunch of times instead. So no, a modern x86 processor is very different than an 8080. Even extensions like x86_64 added a boatload of registers in addition to increasing their widths. The only thing that's in common is the core ISA.

 

[bug]

I would suggest going back a bit to a time when RISC and CISC were the options. There is a reason one of these won in personal computers. That reason is still valid today, ARM has great advantages which it uses really well in mobile world, but suggesting it could replace x86 in all spheres is just pure speculation. If ARM would be able to do everything x86 does than it would not be ARM anymore....

There is a reason, but that reason may not be valid anymore: compilers.
If your target instruction set is smaller, translating everything is harder. Back in the day this was also hindered by the available computing power, which doesn't seem to be the case anymore.
Which one is better overall (and I'm not confusing x86 with CISC or ARM with RISC here) or even if there will be a one size fits all solution, I couldn't tell you.

 

[mtcn77]

There is a reason, but that reason may not be valid anymore: compilers.
If your target instruction set is smaller, translating everything is harder. Back in the day this was also hindered by the available computing power, which doesn't seem to be the case anymore.
Which one is better overall (and I'm not confusing x86 with CISC or ARM with RISC here) or even if there will be a one size fits all solution, I couldn't tell you.

Since I am unfettered with firsthand knowledge, I think I know an easy shortcut.
A simple instruction set will put more pressure on the data caches, either by running further cycles, or further instructions(i-cache) to do the same amount of work, so will use up more overhead to do the same amount of work(d-cache).
However, its data caches are aligned -there are no divergent flow rates, so control management is simpler.
I'll guess it comes to how much overhead is present from wasted cycles due to the complex vs. simple instruction set difference, in reference to how much transistor - and therefore power - budget is saved from simplifying the instruction flow.