Announcement

Collapse
No announcement yet.

MIPS Loongson 3A Benchmarks On Debian

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

  • #16
    Originally posted by linux5850 View Post
    If the central committee tells every comrade you have to use it in your system (so they can have a hardware backdoor into your system) you have to use it or you go to the gulag (or whatever the Chinese version is called).
    The back-door from the CIA is called: "Intel-Trusted Execution Technology" and there gulag is called : "Guantanamo Bay Naval Base"
    I can not find any difference between the USA and China on that point but China is in fact much more open-source friendly!

    Comment


    • #17
      Originally posted by uid313 View Post
      Does the MIPS architecture have any advantage over ARM or x86?
      Does the Loongson have any advantage over a x86 or ARM?

      If it is not open source, and have no advantages, why would I want a MIPS?
      The Loongson in particular has instruction extensions what allow it run x86 via QUEMU at about 75-80% of the native speed.

      I also think it's a bit of a simpler design so massively multicore (tilera) is more feasible.

      I would also guess the IP is a little looser. MIPS will license both designs and patents essential to any MIPS compatible design, wheras ARM only licenses design blocks.

      Comment


      • #18
        Hmm i should have done benchmarks last time i had one of those in my hands, no, not the 3a, the older one in the yeeloong netbook...
        This is a fine 64 bit processor. When you install the OS, you need the mipsel architecture, not mips.

        Just like arm, it should find its way into the datacenter eventually. The Yeeloong netbook has no hardware blobs (even the bios is open), which is why Stallman has one. I wasn't able to install gnewsense but Debian worked fine on it.

        I wonder if that deal nvidia lost in China was to put gpus in loongson based machines? They should be fine with the radeon driver i guess.

        Comment


        • #19
          Originally posted by maldorordiscord View Post
          Do ARM systems have: "Hardware-assisted x86 emulation" ?
          and 64bit right now?
          And the GS464 core do have a high complex superscalar, out-of-order execution architecture with a complex microcode working internals for high speed optimizations.
          In the definition of the FSF its 100% open-source software hardware because no closed source firmware is needed.
          Richard Stallman use a Longson2 based system because of this.

          This means its more trustfully than Intel: "Trusted Execution Technology"
          Hi there maldorordiscord

          Comment


          • #20
            Originally posted by maldorordiscord View Post
            In the definition of the FSF its 100% open-source software hardware because no closed source firmware is needed.
            Richard Stallman use a Longson2 based system because of this.
            Not exactly correct. There is a lot of closed source firmware (microcode really, not firmware), it's just built into the chip and doesn't have a published update mechanism. The x86 processors have built-in microcode plus a patch mechanism, which somehow makes them "less open-source".

            Comment


            • #21
              Originally posted by bridgman View Post
              (microcode really, not firmware), it's just built into the chip and doesn't have a published update mechanism. The x86 processors have built-in microcode plus a patch mechanism, which somehow makes them "less open-source".
              Arm still has no microcode right? As for having a microcode update mechanism, in some ways that does open a potential security threat, however that is assuming there were no security threats in the design to begin with (either accidental through a bug or purposefully placed there), and being able to patch a potential vulnerability is of course important, I can understand arguments on both sides.

              When it comes to China's 'government', I don't trust them any less than I trust the US government (stuxnet, flame) which is to say, 'not at all'.

              Comment


              • #22
                Originally posted by bridgman View Post
                Not exactly correct. There is a lot of closed source firmware (microcode really, not firmware), it's just built into the chip and doesn't have a published update mechanism. The x86 processors have built-in microcode plus a patch mechanism, which somehow makes them "less open-source".
                In the definition of the FSF a microcode built into the chip without a update mechanism is HARDWARE!
                Because of this you can use 100% open-source software on Loongson hardware.

                A CPU with a patch mechanism for the microcode prove that this is in fact a part of "Software" because you can not chance the hardware but you can change the software because of this the microcode in a X86 CPU is Closed Source software!

                Now you just attack the definition of the FSF but you don't have a better definition.

                Comment


                • #23
                  Originally posted by maldorordiscord View Post
                  Now you just attack the definition of the FSF but you don't have a better definition.
                  Sure I do, and I've said the same thing multiple times although the exact wording is probably a bit different each time.

                  Microcode is microcode, whether burned into the hardware, loaded by hardware from EPROM, loaded by BIOS or loaded by the driver.

                  The FSF definitions (along with some of the GPL terms) are intended to prevent people from moving what would otherwise be open source software functionality into firmware or (in the case of GPL) binary software. They make sense in principle but become misleading and counter-productive if you separate the rule from the rationale and blindly apply it to microcode which really *is* part of the hardware despite being externally loaded or patched.

                  It's like having overly broad definitions of what constitutes criminal activity in order to help catch criminals who are skirting the edge of the law -- OK in principle if used correctly, but problematic if you blindly apply it to non-criminals whose activities fall within the "guard band" of the expanded definitions.

                  Comment


                  • #24
                    Originally posted by bridgman View Post
                    Sure I do, and I've said the same thing multiple times although the exact wording is probably a bit different each time.

                    Microcode is microcode, whether burned into the hardware, loaded by hardware from EPROM, loaded by BIOS or loaded by the driver.

                    The FSF definitions (along with some of the GPL terms) are intended to prevent people from moving what would otherwise be open source software functionality into firmware or (in the case of GPL) binary software. They make sense in principle but become misleading and counter-productive if you separate the rule from the rationale and blindly apply it to microcode which really *is* part of the hardware despite being externally loaded or patched.

                    It's like having overly broad definitions of what constitutes criminal activity in order to help catch criminals who are skirting the edge of the law -- OK in principle if used correctly, but problematic if you blindly apply it to non-criminals whose activities fall within the "guard band" of the expanded definitions.
                    To refute you is simple with Logic.

                    First of all we define a Function:"virus"
                    a Virus is something that can manipulate all kind of software but not Hardware.

                    Now the "Virus" uses the microcode update function to manipulate the microcode in the x86 CPU this prove its "Software" to be more specific: its closed source software.

                    Now the "Virus" try to manipulate the microcode in the Loongson but there is no update function of the microcode this means the result is the virus can not manipulate the microcode.

                    Now we make a truth table:

                    Software+Virus= manipulation successful
                    X86-microcode+virus= manipulation successful
                    Hardware+Virus = Manipulation failed
                    Loongson-micocode+Virus= Manipulation failed

                    Now your conspiracy kick in and you claim that there is a secret Chinese government only back-door to make the virus attack successful.

                    But you have to prove this and I'm sure you can't!

                    Maybe the banned member: "Qaridarium" can do this with some nasty rhetorical tricks but he is banned this means he can not help you out this time.

                    Comment


                    • #25
                      Q, is that you ?

                      Comment


                      • #26
                        who else

                        Comment


                        • #27
                          Originally posted by bridgman View Post
                          Q, is that you ?
                          I told you he is banned he can not help you this time.
                          You will not prove your anti Loongson conspiracy theories?
                          Is there any evidence for a back-door free microcode in the AMD CPUs? I think the possibility of a back-door free AMD CPU microcode is by ZERO and i can prove this: "hidden-debug-mode-found-in-amd-processors"http://hardware.slashdot.org/story/1...amd-processors and in the future with "TrustedZone" even bigger.
                          Now I get your way of thinking you only claim this because your own company do this!
                          You still do not have any prove to cover your conspiracy against the Loongson CPU but on the other side for AMD CPUs its not a conspiracy its a proved fact and for Intel CPUs its the same.
                          I think a Chinese CPU system is the only way to get a system you can trust if you life in the USA or Europe!
                          One is for sure a company like AMD will betray you in a nanosecond if the government ask.

                          Comment


                          • #28
                            Originally posted by maldorordiscord View Post
                            To refute you is simple with Logic.

                            First of all we define a Function:"virus"
                            a Virus is something that can manipulate all kind of software but not Hardware.

                            Now the "Virus" uses the microcode update function to manipulate the microcode in the x86 CPU this prove its "Software" to be more specific: its closed source software.

                            Now the "Virus" try to manipulate the microcode in the Loongson but there is no update function of the microcode this means the result is the virus can not manipulate the microcode.

                            Now we make a truth table:

                            Software+Virus= manipulation successful
                            X86-microcode+virus= manipulation successful
                            Hardware+Virus = Manipulation failed
                            Loongson-micocode+Virus= Manipulation failed

                            Now your conspiracy kick in and you claim that there is a secret Chinese government only back-door to make the virus attack successful.

                            But you have to prove this and I'm sure you can't!

                            Maybe the banned member: "Qaridarium" can do this with some nasty rhetorical tricks but he is banned this means he can not help you out this time.
                            So according to your logic any software that is burned into a PROM, making it unalterable, automatically becomes hardware. This is just non-sense.

                            By the way, this is Qaridarium, as anyone with a sane mind can see (same writing style, same non-sense logic). He should be banned again to make clear that the forum rulers in fact have the authority over this forum.

                            Comment


                            • #29
                              Originally posted by TobiSGD View Post
                              So according to your logic any software that is burned into a PROM, making it unalterable, automatically becomes hardware. This is just non-sense.

                              By the way, this is Qaridarium, as anyone with a sane mind can see (same writing style, same non-sense logic). He should be banned again to make clear that the forum rulers in fact have the authority over this forum.
                              The FSF do have criterias and only because I follow the criterias of the FSF you insult my writing as "non-sense".
                              And more you spread conspiracy theories about the banned member "Qaridarium" to insult me even more.
                              Also no Richard Stallman do have the same Logic and he does not get a ban just because of the same conclusion.
                              Educate yourself! Software in hardware unchangeable is defined as hardware and its allowed in the definition of the FSF!
                              http://libreplanet.org/wiki/Hardware...ement_criteria
                              And because Loongson meets all criteria and x86 hardware do not meet any criterias Richard Stallman use Loongson hardware.
                              But you prefer to insult people instead of try to understand people.
                              Its the same with Bridgman he prefer to spread conspiracy theories about backdoors in the loongson product instead of prove anything he says.
                              In fact Bridgman should be banned because of abusive criticism on the competitive product based on conspiracy theories!
                              And you should be banned because of insulting people only because they follow FSF criterias.

                              Comment


                              • #30
                                Originally posted by maldorordiscord View Post
                                Its the same with Bridgman he prefer to spread conspiracy theories about backdoors in the loongson product instead of prove anything he says. In fact Bridgman should be banned because of abusive criticism on the competitive product based on conspiracy theories!
                                I don't know who you are but you're starting to exhibit a lot of the same bad habits as Q (ie flat-out lying about what I said). That is probably not the best path to take, particularly for a new user.

                                Comment

                                Working...
                                X