How fast is the Apple A5X compared to desktop CPUs? I would say it is much slower than every Intel Core processor, but how about a 1GHz PowerPC G4 or Pentium III? Dual Core 1GHz doesn't say much about its performance...
However, Arm chips will get to the desktop within about five years. . .Battery life will then be in the order of 10-20 hours.
What is it about ARM processors that shines brighter than Intel processors?I hope you're right. I want to see this brighter future.
What is it about ARM processors that shines brighter than Intel processors?
They are a RISC CPU like the PowerPC CPU's and they are very effienct when it comes to power requirements.
What is it about ARM processors that shines brighter than Intel processors?
I'm not interested in the processors; I'm interested in the battery life.
Well you need one for the other.
Embedded operating systems are what make tablets and phones seem fast. They are built for the specific processor in mind and will likely forgo many of the libraries and applications commonly found in more generic OS builds.
The reality is that a desktop ARM systems would be slow when it has to run a generically optimised operating system. Which is why it could be (worryingly) likely that desktop ARM systems are locked down, highly proprietary systems.
Great post.This is a great point to make and there's a lot of possible responses.
For example, there was once a belief that open ecosystems for hardware and software were healthy and indeed necessary, because that's what created the PC revolution and the Internet. But there's been a lot of examples where such systems have fallen flat on their faces, especially in the embedded field. Take the opening up of Symbian, for example, which utterly failed to get traction. Then there was Maemo, which was supposed to save Nokia, but turned into a massive white elephant. Opening up a software system isn't a magic wand.
Taking a locked down approach seems to work better. Apple does it, for example, and even Google Android is far from an open source project in spirit. Google engineers do their stuff in secret and then release the source when they have to, usually when a binary release of a point uograde is made public. And sometimes not even then. aren't we still waiting for the source of the latest Android release?
I agree that Apple on ARM desktop will probably be locked down but that's down to anal retentiveness at Infinite Loop, rather than a general trend. OS X is open because it had to be, and we all forget how dire it was at Apple back in the late 1990s and early 2000s. Jobs himself said that if there's no longer any space to innovate then you must follow, which is what OS X did at that point in time. OPen source was the fashion.
This is a great point to make and there's a lot of possible responses.
For example, there was once a belief that open ecosystems for hardware and software were healthy and indeed necessary, because that's what created the PC revolution and the Internet. But there's been a lot of examples where such systems have fallen flat on their faces, especially in the embedded field. Take the opening up of Symbian, for example, which utterly failed to get traction. Then there was Maemo, which was supposed to save Nokia, but turned into a massive white elephant. Opening up a software system isn't a magic wand.
Taking a locked down approach seems to work better. Apple does it, for example, and even Google Android is far from an open source project in spirit. Google engineers do their stuff in secret and then release the source when they have to, usually when a binary release of a point uograde is made public. And sometimes not even then. aren't we still waiting for the source of the latest Android release?
I agree that Apple on ARM desktop will probably be locked down but that's down to anal retentiveness at Infinite Loop, rather than a general trend. OS X is open because it had to be, and we all forget how dire it was at Apple back in the late 1990s and early 2000s. Jobs himself said that if there's no longer any space to innovate then you must follow, which is what OS X did at that point in time. OPen source was the fashion.
Goodness me that's an annoying comment.
In actual fact... the computer industry spent the 70s and 80s closed. For profit, chiefly by Bell Labs. You weren't allowed to do anything with the system. UNIX was not allowed to be ported.
...
I appreciate this level of cynicism, but I don't think that a company's mandate to make money excludes extensive research from taking place. Apple's enormous value helps fuel its R&D, and then the 2% of that research, tied with the value of the Apple brand, goes and finds its way into stores which perpetuate the company's growing wealth. I can see tonnes of choices they've made on grounds of cost, but I also have a little schoolboy naiveté that says they make other choices that are technological and aesthetic, and the products we get are the best compromise between the two forces.That is... unless they're university staff. Most of the good open source or free software work, eg: BSD, has been undertaken and delivered by universities. At the moment we have a huge brain drain out of higher education and academic research, into walled multi-billion-dollar guilds called giant technology companies. The problem is, a company's job is to make money for the shareholders, not to conduct research, to make mistakes and learn from them.
Goodness me that's an annoying comment.
In actual fact... the computer industry spent the 70s and 80s closed. For profit, chiefly by Bell Labs. You weren't allowed to do anything with the system. UNIX was not allowed to be ported.
So, a bunch of people at Berklee hijacked the source code, and took to illegally porting it to other, smaller, more affordable platforms. These guys believed that computing was a universal good which should be available to everybody, not just banks and the military.
AT&T Bell Labs pursued these guys as though they were thieves, and some of them went on the run chased by the FBI across the USA. The original hackers really believed that the architecture of a computing system would be subject to severe dysfunction and abuse if people were not allowed in, or allowed to know how it works. It would severely restrict the possibilities.
The story goes back earlier... to the 70s. When Bell Labs poached and brain drained the Cambridge University computing department, along with the guy who developed C++, it was those guys that specified UNICS.
At Cambridge, they weren't particularly well paid, and their work wasn't patented. It was publicly available to the academic and commercial community as a public good. Once their work was carried over to Bell, it became entombed in legal protectionism. This is the AT&T that refused to let people use their phone lines for modems, or plug a 3rd party phone into the wall socket, that was broken up for monopolies and antitrust.
Open Source is not the Free Software movement. I have noticed that when proprietary and non-technical commercial shareholder forces get involved, the whole system becomes very corrupt and user unfriendly. You can look at Adobe, or Microsoft, or Apple's iPad/iPhone app store behaviour. It leads to rent seeking behaviour. With a publicly traded company, a closed system attracts Private Equity firms. Blackstone are currently LBO'd into apple, and it doesn't bode well for the future. Marketing departments don't design good computers.
Software is a language, it should be mutual to its users, and it should be possible for anybody to contribute.
The failure of the Open Source software movement is simple. Our current economic systems don't support programmers writing complex systems and being able to eat.
That is... unless they're university staff. Most of the good open source or free software work, eg: BSD, has been undertaken and delivered by universities. At the moment we have a huge brain drain out of higher education and academic research, into walled multi-billion-dollar guilds called giant technology companies. The problem is, a company's job is to make money for the shareholders, not to conduct research, to make mistakes and learn from them.
Knowledge, and knowledge systems should *not* be commercial monopolies. The university system has become a training camp for industry paid for on loans at the future worker's expense, rather than a federal organisation of researchers for the common benefit of all humanity and the advancement of knowledge.
I'm find the unix thing pretty interesting - do you have any articles that I could read to find out more?
That is... unless they're university staff. Most of the good open source or free software work, eg: BSD, has been undertaken and delivered by universities. At the moment we have a huge brain drain out of higher education and academic research, into walled multi-billion-dollar guilds called giant technology companies. The problem is, a company's job is to make money...
Well all modern chips are RISC inside. It's just that Intel and Amd chips have a built-in translation layer to convert x86 instructions into RISC code. Wasn't Itanium Intel's attempt at pure RISC? And that obviously bombed badly.
The key thing about ARM chips if I recall is that they have various ultra low powers saving modes so can go into standby mode and still do stuff like communicate with cellphone towers yet only use milliwatts of power. ARM have a major lead in this technology compared to Intel and probably enough patents to keep Intel at its heels.
Intel have recently hit back with a new x86 chip which, in practice, is almost as efficient as ARM chips. I forget the name of the chip but it was reviewed on Anandtech and you can probably google for it.
I'm find the unix thing pretty interesting - do you have any articles that I could read to find out more?