Announcement

Collapse
No announcement yet.

Performance Predictions

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

  • #46
    I wonder if the people at Intel who decided on the x86 instruction set had any idea how that would affect the future of computing. Such a small-seeming decision at the time turned out to have a huge impact. Kinda like how America chose English over German as our official language by just one vote. Think of the impact that would have had if one vote had been different.

    It really is amazing how far the X86 language has been pushed, and it's still being pushed into the 64 bit world. If Intel had gotten behind a simpler architecture like Alpha or PA-RISC, they might finally have gotten us out of this mess. As it is, Itanium is so complex and arcane that very few people want to program for it; they would rather carry on with X86-64. Go figure. That's what happens when you just let everything be decided by who's the richest and most aggressive at marketing.

    Comment


    • #47
      Back to topic:
      I will buy it as long as it is faster than a GF2 Pro and thats almost certine
      If there's artificial intelligence, there's bound to be some artificial stupidity.

      Jeremy Clarkson "806 brake horsepower..and that on that limp wrist faerie liquid the Americans call petrol, if you run it on the more explosive jungle juice we have in Europe you'd be getting 850 brake horsepower..."

      Comment


      • #48
        Originally posted by superfly




        Now where does that quote come from?....Duke perhaps?....



        Which reminds me....Where's DNF?...
        The movie..

        They Live starring Rowdy Roddy Piper.

        amish
        Despite my nickname causing confusion, I have no religious affiliations.

        Comment


        • #49
          LMAO....
          note to self...

          Assumption is the mother of all f***ups....

          Primary system :
          P4 2.8 ghz,1 gig DDR pc 2700(kingston),Radeon 9700(stock clock),audigy platinum and scsi all the way...

          Comment


          • #50
            Originally posted by Technoid
            Back to topic:
            I will buy it as long as it is faster than a GF2 Pro and thats almost certine
            Faster@4800x1200 ?

            Comment


            • #51
              Originally posted by KvHagedorn
              I wonder if the people at Intel who decided on the x86 instruction set had any idea how that would affect the future of computing. Such a small-seeming decision at the time turned out to have a huge impact. Kinda like how America chose English over German as our official language by just one vote. Think of the impact that would have had if one vote had been different.

              It really is amazing how far the X86 language has been pushed, and it's still being pushed into the 64 bit world. If Intel had gotten behind a simpler architecture like Alpha or PA-RISC, they might finally have gotten us out of this mess. As it is, Itanium is so complex and arcane that very few people want to program for it; they would rather carry on with X86-64. Go figure. That's what happens when you just let everything be decided by who's the richest and most aggressive at marketing.
              I don't think intel had any idea how successful the PC would be I am sure if they could rewrite history, they would.

              Well, the legacy of X86 isn't that bad really.

              Firstly, the CPUs for X86 are still getting faster. All current (K5, K6, P6, K7, P7) X86 CPUs are simply a nice RISC cpu cores with X86 instruction decoder. I do realize that the X86 instruction set does limit absolute performance somewhat, but it doesn't do it by a massive margin.

              Secondly, you don't need to deal with the x86's weird instruction set most of the time, since assembly programming isn't very useful in most cases. In most tasks, I can't actually beat the new GCC compilers in code generation on the x86. (not that I am actually very good at assembly, in fact I don't think I am very good at it at all )
              The compiler hides all the hard stuff and people mostly program in some other language. Unfortunatly, those writing OS's and compilers end up with a few (lot) more grey hairs.

              Intel and X64(Itanic and co). Should I comment? Yeah

              While I still beleive the design principles behind these X64 processors are cool, it seems that in reality, on real silicon, they simply don't do that well. Unless McKinley is damned good, Intel may have another stillborn archetecture on its hands.

              If that happens, I can imagine Intel falling from grace and going back down to desktop and low-end server processors for a while, while Compaq-HP would suddenly find their new Alpha and HP-RISC designs and pretend they were never cancelled. Sun would laugh at everyone because even though their SPARC processors are slow, poeple are still buying and using them. And life will continue
              80% of people think I should be in a Mental Institute

              Comment


              • #52
                Very immediate benefits of 64bit.

                1. Double your memory bandwidth (and all data busses)

                2. with 64bit, some problema requiring floats can be done with fixed point integers instead. (a lot faster)

                And a 32/64 bit hybrid is quite nice, because most of the time 64bit is overkill for most procesing and by using 32 bit stuff you avoid huge machine instruction sizes..etc

                Comment


                • #53
                  Let's not forget much faster/stronger encryption..
                  Distributed Computing just for fun...
                  www.team-tnt.net

                  Check us out.

                  Comment


                  • #54
                    Originally posted by Marshmallowman
                    Very immediate benefits of 64bit.

                    1. Double your memory bandwidth (and all data busses)
                    No, intel cpu's already have 64-bit datapaths from which the cpu intel caches
                    are served.

                    Originally posted by Marshmallowman

                    2. with 64bit, some problema requiring floats can be done with fixed point integers instead. (a lot faster)

                    [/B]
                    Contrary to popular belief, most floating point operations arn't slower than integer operations. In fact, people often consider doing integer operations on the floating point unit.

                    Originally posted by Marshmallowman

                    And a 32/64 bit hybrid is quite nice, because most of the time 64bit is overkill for most procesing and by using 32 bit stuff you avoid huge machine instruction sizes..etc [/B]
                    True
                    80% of people think I should be in a Mental Institute

                    Comment


                    • #55
                      Yep

                      Originally posted by otlg22
                      Let's not forget much faster/stronger encryption..
                      Having worked on encryption algorithms, public key encryption systems are greatly improved using 64-bit operations.

                      I beleve you could get between a 2 and 4 times improvement in basic big number operations by using 64-bit variables instead of 32-bit variables. Do 4096 bit RSA at the cost of 1024 bit RSA.
                      80% of people think I should be in a Mental Institute

                      Comment


                      • #56
                        You need to decouple the bus from future processors.. the 'bus' is going to become less and less of a traditional bus.. (witness hypertransport).
                        Distributed Computing just for fun...
                        www.team-tnt.net

                        Check us out.

                        Comment


                        • #57
                          Rugger,

                          I'm an encryption guy as well (well ex-encryption, but you get the idea)
                          Distributed Computing just for fun...
                          www.team-tnt.net

                          Check us out.

                          Comment


                          • #58
                            right...If you keep repeating that enough times,you might end up believing it yourself.....
                            Or know it to be true. It's a lot slower than the things I see on a day-to-day basis and make happen. I think people that try to use something they spent $$$ on to win pissing contests had better find something to give real meaning to their life.


                            About the future of architectures: I really think you guys are way off for the most part, but I can't comment much more than that. Keep in mind that IA-64 runs x86 and PA-RISC code though.

                            IA-64 does seem a little complicated at the assembly level, but there are definitely reasons for it. All that hardware prediction and speculation pays off.
                            Gigabyte P35-DS3L with a Q6600, 2GB Kingston HyperX (after *3* bad pairs of Crucial Ballistix 1066), Galaxy 8800GT 512MB, SB X-Fi, some drives, and a Dell 2005fpw. Running WinXP.

                            Comment


                            • #59
                              64bit will also extends the Unix timestamp past 2038. You wouldn't want to be stuck with a 32 bit CPU when 2038 rolls around.

                              Hi otlg#, nice to see a fellow TA/TIMO/TNT DCer around here.
                              I should have bought an ATI.

                              Comment


                              • #60
                                Wombat,

                                IA64 runs x86 poorly, and I'm not sure how well it would do at PA-RISC (although probably much better, as PA-RISC is a tad closer to IA-64)

                                Intel's EPIC (which I'm sorry is really just VLIW) is like every other VLIW processor I have seen so far.. great for DSP like functions and miserable for general purpose computing. Of course I haven't had the chance to see much of McKinnley, and things may change (a first generation of a new architecture usually sucks rocks hard), but I have visions of the IA432, which while a techincally brilliant chip was a market disaster (although they did learn some interesting things in the process, so all wasn't for nought).

                                Regards,

                                Steve
                                Distributed Computing just for fun...
                                www.team-tnt.net

                                Check us out.

                                Comment

                                Working...
                                X