Announcement

Collapse
No announcement yet.

NVIDIA Says It Will Deliver ARM CPUs Spanning PCs to SCs

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #41
    I like how this thread title has Every First Letter Written In Capitals.
    Nvidia must be planning own ARM-based system.
    Then they will port 95% of games from x86 to ARM.
    Then they take Linux and mix it throughly.
    And finally, they will close everything down. >

    Comment


    • #42
      Originally posted by crazycheese View Post
      I like how this thread title has Every First Letter Written In Capitals.
      The word "to" isn't :P

      Comment


      • #43
        Originally posted by Thatguy View Post
        Really ? This is news to me since they have half the mips performance per cycle of a x86 cpu. Which is why we are all still using x86 based designs.
        have you ever owned and played with anything other than x86 Thatguy ?

        well, in you're limited view "we are all still using x86 based designs" may be true, however ARM have been around a very long time, the latest ARM cortex quad core A9 have OC only just arrived and these are what your prototype server farms etc will be using to start with for instance...

        while its not really representative of what's coming speed wise, it gives the reader an idea of the ARM cortex potential.

        but this real life Benchmark test came from the old x264-dev logs if anyone's interested in the real life Number's.

        and StippenG's number's came from an older Quad A9/NEON developer board at his Uni apparently, so a current Marvell ARM v7 A9 quad /SIMD at 1.6 GHz for instance would produce a Far better result today OC.

        640x360 at Ultrafast: 38.59 seems like a very good real life start for encoding on ARM cortex even without those extra SIMD patches being written yet

        "2010-08-24 15:39:19 < StippenG> Some X264 Benchmarks (Rush Hour 640x360, preset=medium, crf=24): 4-core Cortex-A9 @ 400 MHz gives 5.55 fps, Beagleboard (A8 @ 720MHz) gives 1,65. Really nice speedup, considering the much higher frequency of the A8

        2010-08-24 15:39:56 < Dark_Shikari> It'd go a lot faster if you used a faster preset.
        2010-08-24 15:40:01 < Dark_Shikari> Or if you wrote some of the asm we're missing
        2010-08-24 15:40:26 < Dark_Shikari> But yeah, that scales surprisingly well. about 3.5x faster

        2010-08-24 15:40:27 < StippenG> Yes. Superfast gives 22.07. Ultrafast: 38.59
        2010-08-24 15:41:23 < StippenG> Guess the out-of-order execution and shorter pipeline is really quite a bit better for performance

        2010-08-24 15:41:35 < Dark_Shikari> Well, the A9 is known to be a lot faster
        "

        Comment


        • #44
          Originally posted by deanjo View Post
          The instruction count for a particular program is completely dependent on the nature of the program and the instruction set architecture. What one architecture may require 20 instructions to do may only take 10 on another architecture. It's an apples and oranges comparison.
          So does the ARM architecture have any instructions that x86 doesn't have? I had the impression that x86 packed everything that ARM has, and then some, at the expense of running at a higher wattage.
          One thing, at least, that I could find on Wikipedia was Jazelle, execution of Java bytecode in hardware. That seems pretty usable, especially for mobile devices.

          Comment


          • #45
            Originally posted by popper View Post
            have you ever owned and played with anything other than x86 Thatguy ?

            well, in you're limited view "we are all still using x86 based designs" may be true, however ARM have been around a very long time, the latest ARM cortex quad core A9 have OC only just arrived and these are what your prototype server farms etc will be using to start with for instance...

            while its not really representative of what's coming speed wise, it gives the reader an idea of the ARM cortex potential.

            but this real life Benchmark test came from the old x264-dev logs if anyone's interested in the real life Number's.

            and StippenG's number's came from an older Quad A9/NEON developer board at his Uni apparently, so a current Marvell ARM v7 A9 quad /SIMD at 1.6 GHz for instance would produce a Far better result today OC.

            640x360 at Ultrafast: 38.59 seems like a very good real life start for encoding on ARM cortex even without those extra SIMD patches being written yet

            "2010-08-24 15:39:19 < StippenG> Some X264 Benchmarks (Rush Hour 640x360, preset=medium, crf=24): 4-core Cortex-A9 @ 400 MHz gives 5.55 fps, Beagleboard (A8 @ 720MHz) gives 1,65. Really nice speedup, considering the much higher frequency of the A8

            2010-08-24 15:39:56 < Dark_Shikari> It'd go a lot faster if you used a faster preset.
            2010-08-24 15:40:01 < Dark_Shikari> Or if you wrote some of the asm we're missing
            2010-08-24 15:40:26 < Dark_Shikari> But yeah, that scales surprisingly well. about 3.5x faster

            2010-08-24 15:40:27 < StippenG> Yes. Superfast gives 22.07. Ultrafast: 38.59
            2010-08-24 15:41:23 < StippenG> Guess the out-of-order execution and shorter pipeline is really quite a bit better for performance

            2010-08-24 15:41:35 < Dark_Shikari> Well, the A9 is known to be a lot faster
            "

            Its like this, by the time ARM catchs up on mips per cycle"by adding instructions", it'll be just like x86 in terms of power consumption.

            I fully expect x86 to refine power useage and x86 will proliferate in more devices at the next few node shrinks.

            Its how it is, ARM has been out of the running for a while.


            I know motorla,powerpc,intel8051 etc.

            your ill formed opinion of performance is that, a ill formed opinion. To get ARm to reach x86 performance will basically require building a x86 cpu.

            Comment


            • #46
              Originally posted by runeks View Post
              So does the ARM architecture have any instructions that x86 doesn't have? I had the impression that x86 packed everything that ARM has, and then some, at the expense of running at a higher wattage.
              One thing, at least, that I could find on Wikipedia was Jazelle, execution of Java bytecode in hardware. That seems pretty usable, especially for mobile devices.
              its all abut IPC and x86 has it. even if you dicount the small efficiency difference between x86 and ARM x86 has raw horsepower.

              I haven't followed ARM for a long time but I don't think ARM really offer anything but Less then x86 in terms of computing power. ARM is just a very optimized idea of x86 with a much lower power footprint.

              A distinction that both intel and AMD intend to remove soon enough at the next big node jump. I think x86 will start migrating heavily upstream into mobile devices soon enough. window7 on a tablet, phone etc.

              think of the software development savings and the interoprability advantages.

              Its gonna happen. the question is will linux expand in the deksotp arean and client side enough to get competitive with MAC and MS ?? or will MS and MAC keep up comming and replacing current linux arm variants?

              Comment


              • #47
                Originally posted by Thatguy View Post
                Its like this, by the time ARM catchs up on mips per cycle"by adding instructions", it'll be just like x86 in terms of power consumption.

                I fully expect x86 to refine power useage and x86 will proliferate in more devices at the next few node shrinks.

                Its how it is, ARM has been out of the running for a while.


                I know motorla,powerpc,intel8051 etc.

                your ill formed opinion of performance is that, a ill formed opinion. To get ARm to reach x86 performance will basically require building a x86 cpu.
                LOL you are you're own worst enemy sometimes Thatguy, you have no idea who or what i know and you need to look in the mirror and have a hard look at yourself before throwing out that kind of thought to/about others.

                you remind me of the mac kid with a new toy and goes around telling everyone that 'the app store' is our saviour and so your current toy haiku will be fine if we only had one too and lets without or ignore the missing video capability years after it began, i guess You wont be writing or compiling the haiku ARM port on your super 20Mbit home download link even after someone tells you how to actually do it ROTFL

                perhaps you should start a bounty and buy that haiku developer a real life ARM A9 dual or quad or even a cheap $200 Genesi efika A8 with slow memory bus box rather than bitch and moan about linux OSS

                Comment


                • #48
                  Originally posted by runeks View Post
                  So does the ARM architecture have any instructions that x86 doesn't have? I had the impression that x86 packed everything that ARM has, and then some, at the expense of running at a higher wattage.
                  One thing, at least, that I could find on Wikipedia was Jazelle, execution of Java bytecode in hardware. That seems pretty usable, especially for mobile devices.
                  Quite a few, read from here down:

                  Comment


                  • #49
                    Originally posted by runeks View Post
                    So does the ARM architecture have any instructions that x86 doesn't have? I had the impression that x86 packed everything that ARM has, and then some, at the expense of running at a higher wattage.
                    One thing, at least, that I could find on Wikipedia was Jazelle, execution of Java bytecode in hardware. That seems pretty usable, especially for mobile devices.
                    your best advice runeks is to just forget anything Thatguy said to you in reply to your question, especially that classic horse shit "remove soon enough at the next big node jump"

                    as by the time ARM gets around to making even today's so called x86 node/die size they will be far faster and even lower power than Thatguy and his x86 will be by then.

                    apparently he doesn't even know that ARM cortex are hardware RISC not CISC with software micro code to try and make them hybrid RISC with massive legacy overlay's or apparently that there is even a difference LOL

                    to actually point you in the right direction as regards your "does the ARM architecture have any instructions that x86 doesn't have?"
                    see http://www.keil.com/support/man/docs...m_Cjafgdih.htm

                    Comment

                    Working...
                    X