ARM for the future?

[Khephren]

Heh, cool, looks very nice. It's inspirations are pretty clear (some heavy rtype inspired bits there), but it looks to have been done quite well.
What's the story behind it though, is it a commercial game ? Looks too nice to be homebrew  

Yeah, I love R-Type. I think they are an indie publishing house, did Rush Rush Rally for DC as well.  http://www.redspotgames.com/shop/

To stay on thread, do any of you think ARM will hit the desktop/notebook heavily?  
I know intel are definitely trying to get heat/wattage down to the smartphone arena (wonder if they regret selling strongARM?) Could be a big battle ahead!

SIDENOTE: I have an ARM netbook, but it runs CE it is a bit shit. The clips in at 144MIPS, but can do less, and feels less responsive than my A1200:030/50. (This is definitely due to the OS though, as my Psion Revo is well nippy).

[Hattig]

ARM is the common sense non-x86 future choice of processor for the Amiga line.

The performance is there, the cost is low, the availability is massive, the choice is high. They're integrated, they're multi-core (for when that is available in AmigaOS).

Sure, we would need another 68k emulator in AmigaOS, and we would need all the PowerPC apps recompiled (or a PowerPC emulator... ugh).

[jj]

No ARM in trying.
 
Sad I know. But I miss Franko..
 
Tag: Typed on a little Netbook(Atom) running XP. No ARM in that either.

Where has Franko gone then ?

[fishy_fiz]

Where has Franko gone then ?

So far as I know he decided to take his bat and ball and go home  
Nah, seriously, I think he's just decided to return to a life of solitude and his amigas.

[r06ue1]

Hopefully this new CPU kicks Intel butt!  I have been sick of x86 since the day I bought my first PC.  
 
More info on the Nvidia ARM:  
http://blogs.nvidia.com/2011/01/project-denver-processor-to-usher-in-new-era-of-computing/

[fishy_fiz]

Nah, it wont. ARM doesnt come close to high end x86 gear. Also x86 of today shares very little in common with x86 of yesteryear, so saying youre sick of x86 since you bought your first pc is like saying youre sick of a different cpu compared to what what is your 1st pc.

I probably shouldnt be surprised to see this kind of garbage on an Amiga site though.

[ElPolloDiabl]

Nah, it wont. ARM doesnt come close to high end x86 gear. Also x86 of today shares very little in common with x86 of yesteryear, so saying youre sick of x86 since you bought your first pc is like saying youre sick of a different cpu compared to what what is your 1st pc.

I probably shouldnt be surprised to see this kind of garbage on an Amiga site though.

He might mean to program on. Or maybe he's had a series of buggy computers. You do have to be careful choosing your parts because some, usually cheap mfgrs parts will give you trouble.

Do you remember the old celeron with it's gimped 66mhz bus? At the same time the Pentium II use to cost a fortune. How about all those overclocked and rebadged CPUs before intel and AMD started locking them. Intel did a share swap with RAMBUS and tried to push their serial memory.
I can see how someone could have a bad opinion x86s.

[Iggy]

ARM is the common sense non-x86 future choice of processor for the Amiga line.

The performance is there, the cost is low, the availability is massive, the choice is high. They're integrated, they're multi-core (for when that is available in AmigaOS).

Sure, we would need another 68k emulator in AmigaOS, and we would need all the PowerPC apps recompiled (or a PowerPC emulator... ugh).

"or a PowerPC emulator... ugh", yeah that definitely does not sound ideal. We've been bantering this around in the MorphOS community and I think many of us have reached the same conclusion, PPC apps are going to have to be recompiled. Many ARM processors would perform fairly poorly when running a PPC JIT compiler anyway.

I don't think ARM is a possibility anymore, I think its inevitable. We'll probably see an AROS port first, but I expect to see other NG OS' move to this as well.

[commodorejohn]

ARM doesnt come close to high end x86 gear. Also x86 of today shares very little in common with x86 of yesteryear, so saying youre sick of x86 since you bought your first pc is like saying youre sick of a different cpu compared to what what is your 1st pc.

x86 is fine for performance, and it's true that it's come a long way since the 8086, it's still ugly as hell under the hood - the "four sort-of-but-not-really general-purpose registers" approach was obnoxious back in the day, and it's downright inexcusable now, with not only modern architectures but even the original 68000 providing a full complement of interchangeable registers. And it hasn't improved, either - they've just kept expanding the register width. Granted, a compiler can work out a good approach for shuffling values into and out of the registers, but for those of us who like assembler, it's pure aggravation.

[fishy_fiz]

He might mean to program on. Or maybe he's had a series of buggy computers. You do have to be careful choosing your parts because some, usually cheap mfgrs parts will give you trouble.

Do you remember the old celeron with it's gimped 66mhz bus? At the same time the Pentium II use to cost a fortune. How about all those overclocked and rebadged CPUs before intel and AMD started locking them. Intel did a share swap with RAMBUS and tried to push their serial memory.
I can see how someone could have a bad opinion x86s.

Im more the sort of person who will judge something on it's merrits rather than dislike something because I chose something else. Sure, there has been some x86 cpu's that have been weak compared to others, but that doesnt mean x86 as a whole is bad (incidently the original celerons (apart from the 266mhz version) werent so bad vs. p2's.... thier l2 cache ran at full clock rate vs 1/2 clock rate of p2 and early p3's,.. it was only once coppermine was released that the 66mhz bus started to be a problem, but who ran thier celerons at original fsb anyway ?).

All this is neither here nor there anyway. As I said x86 of today is a far cry from x86 of yesteryear.
Modern x86 cpus have 16 general purpose registers by the way. Sure it's still less than a lot of RISC based chips, but quite a difference to the 4 that has been quoted.

[Iggy]

Nah, it wont. ARM doesnt come close to high end x86 gear. Also x86 of today shares very little in common with x86 of yesteryear, so saying youre sick of x86 since you bought your first pc is like saying youre sick of a different cpu compared to what what is your 1st pc.

I probably shouldnt be surprised to see this kind of garbage on an Amiga site though.

ARM is better suited as a destination for ports of 68K based OS' and software than X86. Yes, there are no ARM processor that match X86 in computing power, but ARM is considerably more powerful than 68K and has closed the gap in performance it had with PPCs.
ARM can be configured to run with the same endian  structure as the 68K (as can most PPCs). X86 can not and suffers at least a 20% performance penalty because of that. The only reason this is not significant, is that X86 processors are already more powerful than needed to run this software.
Which brings up another point. How much CPU power do we need to run our software?
Both AOS4 and MorphOS can run on CPUs that operate below 400Mhz. Right now cheap ARM boards exist (at under $200) that run at 1Ghz. Clearly there is enough CPU power.

"I probably shouldnt be surprised to see this kind of garbage on an Amiga site though."

Why are you here if you don't think much of our opinions? This discussion was prompted by announcements made by Nvidia that they intended to compete with X86s in the desktop and server market with a new ARM design. I don't think Nvidia's ideas are foolish AND if they think they can create an ARM CPU that is powerful enough to compete in this market, well then we are going to see ARM processors that are
MORE powerful than we need (for our purposes).

[orb85750]

Don't you realize that you're confusing the hell out of C=USA with this thread.  Are you trying to get them to change their plans and announce yet another new direction for the Amiga?  :-)

[Iggy]

Don't you realize that you're confusing the hell out of C=USA with this thread.  Are you trying to get them to change their plans and announce yet another new direction for the Amiga?  :-)

I don't think our opinions carry much weight with CUSA. They've got their plans and I rather expect they'll stick to them.
They're not interested in the hobbyist market and we already upset them with some of our negative posts.
I don't think they see us a a major part of their potential market.

[Fats]

 x86 is being attempted by AROS ...

Currently a  nightly build of AROS is set up for the following cpu/archs:
- m68k: amiga native
- i386: pc native, linux hosted, windows hosted, darwin(OSX) hosted
- AMD64: pc native, linux hosted
- PPC: efika native, sam440 native, linux hosted, darwin(OSX) hosted
- ARM: linux hosted

Only i386 has currently a somewhat stable ABI through the Icaros and Broadway distros but people should stop equalling AROS with i386.

greets,
Staf.

[Iggy]

Im more the sort of person who will judge something on it's merrits rather than dislike something because I chose something else. Sure, there has been some x86 cpu's that have been weak compared to others, but that doesnt mean x86 as a whole is bad (incidently the original celerons (apart from the 266mhz version) werent so bad vs. p2's.... thier l2 cache ran at full clock rate vs 1/2 clock rate of p2 and early p3's,.. it was only once coppermine was released that the 66mhz bus started to be a problem, but who ran thier celerons at original fsb anyway ?).

All this is neither here nor there anyway. As I said x86 of today is a far cry from x86 of yesteryear.

I'm not anti-X86. I've been using them for years, have several, and have had some great results with them. And your right about Celeron overclocking. It had some great moments. Early Celerons with full speed cache easily dusted P2s. I overclocked every generation successfully right up to the Tualatin core (1.6Ghz - faster than any PIII). The P4 based Celeron were kind of disappointing. You could overclock a 2,4 Ghz Celeron to 3.2 Ghz, but the performance was still significantly below a P4.

And I don't expect X86 OS' like AROS to fade away. Rather I think ARM will just be another branch of the Amiga OS tree.

[fishy_fiz]

ARM is better suited as a destination for ports of 68K based OS' and software than X86. Yes, there are no ARM processor that match X86 in computing power, but ARM is considerably more powerful than 68K and has closed the gap in performance it had with PPCs.
ARM can be configured to run with the same endian  structure as the 68K (as can most PPCs). X86 can not and suffers at least a 20% performance penalty because of that. The only reason this is not significant, is that X86 processors are already more powerful than needed to run this software.
Which brings up another point. How much CPU power do we need to run our software?
Both AOS4 and MorphOS can run on CPUs that operate below 400Mhz. Right now cheap ARM boards exist (at under $200) that run at 1Ghz. Clearly there is enough CPU power.

"I probably shouldnt be surprised to see this kind of garbage on an Amiga site though."

Why are you here if you don't think much of our opinions? This discussion was prompted by announcements made by Nvidia that they intended to compete with X86s in the desktop and server market with a new ARM design. I don't think Nvidia's ideas are foolish AND if they think they can create an ARM CPU that is powerful enough to compete in this market, well then we are going to see ARM processors that are
MORE powerful than we need (for our purposes).

Im not arguing that ARM is the wrong choice of targets. As I said already, it makes sense as a migration path for a 68k or ppc based system.
Nor am I sying ARM is bad. As I said already in this thread there's some interesting thing happening in the ARM world. All Im saying is that it gets tiring hearing the same old anti x86 garbage that's based on how x86 used to be.