ARM vs. X86. Lets vote

[Askarus]

Such a vote actualy needs more infos. What ARM chips are possible, what X86 chips are possible? An more detailed facts sheet would be also a good idea to compare the ARM and x86 leagues. pro's/con's etc... Otherwise it would be just an question of your personal preference instead of the choosing what's the BEST for Pandora 2.

That's what I intended.

 

[MrConfusion]

Such a vote actualy needs more infos. What ARM chips are possible, what X86 chips are possible? An more detailed facts sheet would be also a good idea to compare the ARM and x86 leagues. pro's/con's etc... Otherwise it would be just an question of your personal preference instead of the choosing what's the BEST for Pandora 2.

That's what I intended.

Worry not, judging from the tongue in cheek answers, that's what I think most presumed.

After all, when you ignore the little geek inside you shouting "Intel outside" with leftover RISCvs.CISC rage from the 90's, the rational answer to the question is automatically "As long as it works, frankly my damn I don't give a deer".

(OK, that's enough bad jokes from me, I'll stop posting about this matter now too, let's just try to let it die in peace .

 

[Binky]

I don't give a deer

Special offer!

Four deer for the price of three!

 

[moxie]

That's too deer for me.

 

[Linux-SWAT]

Oh deer !

 

[Klumpen]

Deer god!

 

[3_bit]

Dear god. Why hasn't anyone else ever thought of one of these polls?

 

[Linux-SWAT]

Because ARM is the de facto choice.

 

[3_bit]

I think the poll is unfair, as when you make it one way or the other, as stated by previous users, ARM automatically wins because, I hypothesize, most of the active users here actually own/use an OP, and may have compiled stuff for it, thus meaning the past has a say. It should be given, in the poll, that we can assume difficulties in using either architecture will be the same.

 

[Linux-SWAT]

What's unfair for x86 is simply that it can't win the comparison with ARM from an architecture point of view.

 

[Binky]

Given that the performances of the two "alternatives" seem to be within the stonking great error margin associated with processor benchmarks, we have to take those as equal.

To my mind, a library of stuff designed for pocket devices (PNDs + Android apps) is better than one designed for desktop PCs (steam / windows games), especially as the latter would be a nightmare to get working at all.

 

[ShleeDragon]

I remember telling you chuckleheads to chill out and stop this wank!

might aswell start a vote for the keyboard layout and screen size again... maybe we should vote to include a tiny motor to use the Pandora 2 as a massage tool

 

[Hồng Thất Công]

^ A Pandora vibrator? Awesome! Sell like hot cakes for sure.

 

[Exophase]

What's unfair for x86 is simply that it can't win the comparison with ARM from an architecture point of view.

I generally find ARM assembly less of a pain to write than x86 assembly (x86-64 is better), but when it comes to the parts of architecture outside of the instruction set I'd say ARM still has some warts x86 doesn't. It's gotten better over the years but that makes some of the current omissions more odd. For instance, ARMv7 provides coherency for data caches between separate cores, but it still has no coherency between instruction and data caches on the same core. Instead you have to call explicit invalidation instructions, but they only work in supervisor mode so that usually means an expensive syscall. This is an annoying performance overhead for recompilers/JITs which are only becoming more and more common. What makes it even more ridiculous is that these invalidation instructions have to be broadcast to every core. Adding support in hardware would have negligible impact on current implementations but they won't change it, even for ARMv8 (although that may at least have user mode methods of doing it).

Given that the performances of the two "alternatives" seem to be within the stonking great error margin associated with processor benchmarks, we have to take those as equal.

I don't think performance is within margin of error, and perf/W most certainly isn't.

Yeah this is something that'll change with time, but Intel's manufacturing advantage isn't going away any time soon. Meanwhile a lot of the remaining top talent in CPU design is getting sucked into Apple, which is making really impressive ARM chips that we'll certainly never have a chance to use in a Pandora.

 

[Hồng Thất Công]

^ Damn Apple!

 

[doragasu]

x86 sucks! Long live ARM!

I don't think Pandora would have been able to reach 10+ gameplay hour with an x86 SoC. Lately, Intel is making great progress on the low power race, but x86 CPUs have a lot of bizarre instructions and "features" (like internal instruction translation from CISC to pseudo RISC) that will always drain a lot of power.

Sure you can install Windows + PC software on an x86 CPU, but I don't want Windows at all, if I have to sacrifice battery life!

x86 has lasted way too long. I'm taking each opportunity I can trying to bury it. E.g.: the last laptop I bought was an ARM Chromebook (I thrown away ChromeOS and installed Arch Linux, and now it's a nice machine).

Abandon "Wintel" and embrace Linux+ARM 

 

[Linux-SWAT]

We can always found defaults, but anyway ARM > x86.

And btw, x86_64 is better, because it was made by AMD.

 

[ekianjo]

Err.. I'd like to remind to you guys that product design is (thanksfully!) not a democratic/voting process.

You don't want to end up with "the car built for Homer" with discussions like this.

 

[Askarus]

That's the reason why we vote and let not design it by EvilDragon alone .

 

[Linux-SWAT]

That's why i requested a little moderation.