ARM diversity might be an advantage

Article: Microservers must Specialize to Survive
By: Patrick Chase (patrickjchase.delete@this.gmail.com), February 3, 2013 4:29 pm
Room: Moderated Discussions
Paul A. Clayton (paaronclayton.delete@this.gmail.com) on February 2, 2013 11:10 am wrote:
> Patrick Chase (patrickjchase.delete@this.gmail.com) on February 1, 2013 10:11 pm wrote:
> > David suggested posting this to the forum. I think he has a few remarks of his own to add on this topic...
> >
> > I think that the statement that x86 takes 5-15% more area than RISC is a bit simplistic,
> > because the penalty is highly variable depending on what performance level you're
> > targeting and what sort of microarchitecture you have to use to get there.
>
> x86 also has a steeper learning curve as one needs to learn the tricks to handle various odds
> and ends. Intel and AMD already have institutional knowledge about implementation (including
> validation tools), but a third party is less likely to find implementing a variant or an original
> design worthwhile (even if Intel provided the appropriate licensing). It has also been argued
> that a "necessity is the mother of invention" factor drove x86 implementers to innovate.

The P6 and its successors are not terribly innovative. P6 itself is basically a straightforward Tomasulo machine, typical of its generation. The x86-ness was mostly contained in the decoders and in "frontend tricks" like eliding FXCHs via RAT/ROB machinations (arguably a predecessor of uop fusion). If you look at a modern x86 like Haswell it's also a fairly typical OoO microprocessor in the PRF style (schedulers instead of reservation stations, dataless ROB). The principal factors that distinguish it from non-x86 designs are the decoders (of course), the high load/store bandwidth to L1 (doubtless to deal with all of the stack accesses), and frontend tricks like micro- and macro-op fusion. You could argue that the L0 uop cache is also an x86 optimization in that it mitigates the x86 decode penalty, but OoO RISCs have also used predecoded I-caches (for example PA-8000).

What I think really happened is not so much that Intel and AMD got more innovative as that the field of play (specifically transistor/power budgets and performance requirements) moved to a space where they could be competitive *using the same architectural techniques as everybody else*. In other words, if everybody is doing OoO machines then the penalty for x86 is inherently much lower than if everybody [else] is doing classic 5-stage RISC pipelines. If true, then that has huge ramifications for the present discussion, given the direction the ARM camp are headed right now in their effort to steal server share from Intel.

> A clean RISC like Alpha (or--from what I have read--AArch64) would be much more friendly to fast bring-up
> of a decent microarchitecture. (Classic ARM seems to be somewhere in the middle--not as complex as x86 but
> not as simple as Alpha--, but even with Thumb2+classic ARM it might be closer to Alpha than to x86.)

AArch64 is indeed a nice, classic RISC architecture. In particular they fixed the biggest single limitation of classic ARM (spending instruction encoding bits on condition-based predication at the expense of GPRs). I'd put it somewhere between MIPS and Alpha on the "architectural purity scale", and that's a pretty good place to be.

That said, we've seen several nice, classic RISC architectures go up against x86 in an application space that required wide OoO designs, and they ended up not having much of an advantage when all was said and done. That was really the whole point of my post: You can't look at the x86 penalty versus the "classic 5-stage RISC pipe" or even in the "static superscalar" regime, and assume that will hold when everybody is forced by performance requirements to do 3+-wide OoO microarchitectures.

> [snip]
> > My own take is that for ARM-based microservers to survive they need to stay down in the "many weak cores"
> > regime and focus on massively parallel workloads that can tolerate the latency penalty. If they try to
> > move up into higher performance brackets then they'll be playing directly into Intel's hand.
>
> I agree that trying to compete with Intel x86 at the high performance end will be excessively difficult,
> but I think the ARM brigade may have a flexibility advantage. Even though Intel has been demonstrating some
> willingness to try new things and develop concurrent multiple microarchitectures, Intel seems to be too conservative
> to try radical designs. It is not clear that ARM will take advantage of its greater tolerance of diversity
> (while learning to provide a coherent interface to software) to introduce some weird and wonderful architectural
> features. ARM has been very quiet about transactional memory and multithreading; features along the lines
> of Intel's TSX and MIPS' MT-ASE could be significant in the server market.

That is indeed ARM's greatest historical strength, and IMO the reason why they succeeded where some RISC players with significantly cleaner ISAs didn't. As you may have gathered I'm not a fan of the pre-AArch64 ARM ISA, and yet I have a LOT of experience developing for it. The reason why is because the overall ARM solution (including ecosystem, IP cost, and ease of integration) was compelling for the applications in question, despite any architectural foibles.

I think that business model will be under increasing pressure going forward, though. Fixed development costs (design time, masks, bringup and FW) for SoCs are increasing by a factor of about sqrt(2) per generation, and that's gradually squeezing smaller players out and forcing them to buy standardized parts, whether from TI, NVidia, Qualcomm, or Intel. At that point the fact that your *supplier* had a lot of flexibility and an easy time integrating the core is of rather secondary importance. From the end developer's perspective ARM still benefits from a strong ecosystem, but so does x86.

> Even if ARM does not innovate much architecturally, I think the implementers may feel
> much more free to try different accelerators and microarchitectural tweaks. With
> an Architecture license, non-ARM implementers could even add new instructions.

People with ARM architecture licenses have added custom instructions to ARM, generally for application-specific DSP-like functionality. I don't know how much of what I know is public, so I'll leave it at that. I would observe that ISA changes at the "OS/compiler visible" level (beyond adding an intrinsic here or there) are less common, probably because those would tend to erode the ecosystem benefits. I would also observe that as microarchitectures become more complex people tend to get priced out of that game as well. I expect it's vastly easier to create and verify, say, an ARM9 or R4 derivative than to a customized A15.

Please don't get me wrong: I'm not a proponent of x86 as an ISA. Looking at the assembly from even the best compilers (icc) is a discouraging experience. With that said, I personally don't buy the hype around ARM microservers. We've seen this play out before, and I personally think this time won't be different.

The point you made in a separate post about big.LITTLE is probably the most compelling argument I've heard to date, but I think the evidence is very thin. In particular, a lot will depend on whether "LITTLE" ends up being small enough to put x86 into the noncompetitive regime. We shall see :-).
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/28 01:47 PM
  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 04:42 AM
    Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 06:09 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 07:01 AM
        Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 01:35 PM
          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 02:42 PM
            Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/30 12:12 AM
              Article Online: Microservers must Specialize to SurviveMichael S2013/01/30 02:06 AM
              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 09:58 AM
              Article Online: Microservers must Specialize to SurviveDoug S2013/01/30 12:38 PM
                FRU type and granularity selectionPaul A. Clayton2013/01/30 02:29 PM
                  FRU type and granularity selectionrwessel2013/01/30 04:00 PM
                    FRU type and granularity selectionEtienne2013/01/31 03:26 AM
                Article Online: Microservers must Specialize to SurviveMaynard Handley2013/02/13 02:46 PM
                  Article Online: Microservers must Specialize to SurviveDavid Kanter2013/02/16 03:16 PM
                    Article Online: Microservers must Specialize to SurviveDoug S2013/02/17 12:24 PM
        Article Online: Microservers must Specialize to SurviveJohan2013/02/01 01:40 AM
          Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/02/01 03:06 AM
            Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/02/01 03:13 AM
    Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 06:49 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 07:19 AM
        Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 09:47 AM
          Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 02:18 PM
            Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 06:14 PM
              Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 11:51 PM
                Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 08:34 AM
                  Article Online: Microservers must Specialize to Surviveanon2013/01/31 06:45 AM
                    Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 12:25 PM
                      Article Online: Microservers must Specialize to Surviveanon2013/01/31 12:47 PM
                        Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 01:36 PM
                          Article Online: Microservers must Specialize to Surviveanon2013/01/31 04:59 PM
                            Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 05:37 PM
                              Article Online: Microservers must Specialize to Surviveanon2013/01/31 06:47 PM
                                Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 07:16 PM
                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 07:36 PM
                                    Article Online: Microservers must Specialize to Surviveanon2013/02/01 01:47 AM
                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 04:26 AM
                                        Article Online: Microservers must Specialize to Surviveanon2013/02/01 05:56 AM
                                          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 06:35 AM
                                            Article Online: Microservers must Specialize to SurviveMichael S2013/02/01 07:07 AM
                                            Article Online: Microservers must Specialize to Surviveanon2013/02/01 08:08 AM
                                              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 08:49 AM
                                                Article Online: Microservers must Specialize to Surviveanon2013/02/01 09:36 AM
                                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 10:11 AM
                                                    Article Online: Microservers must Specialize to Surviveanon2013/02/01 06:57 PM
                                                      Article Online: Microservers must Specialize to Surviverwessel2013/02/02 01:42 AM
                                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 06:41 AM
                                                        Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 02:14 PM
                                                          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 02:50 PM
                                                            Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 04:05 PM
                                                              Article Online: Microservers must Specialize to SurviveMichael S2013/02/02 04:20 PM
                                                              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 05:05 PM
                                                                ASP estimatesRichard Cownie2013/02/02 06:07 PM
                                                                  ASP estimatesRicardo B2013/02/02 06:48 PM
                                                                    ASP estimatesRichard Cownie2013/02/02 07:04 PM
                                                                      ASP estimatesanon2013/02/02 10:19 PM
                                                                        ASP estimatesRichard Cownie2013/02/03 06:06 AM
                                                                          ASP estimates...cache?Per Hesselgren2013/02/03 12:08 PM
                                                                      ASP estimatesMichael S2013/02/03 02:51 AM
                                                                        ASP estimatesRichard Cownie2013/02/03 06:10 AM
                                                                          ASP estimatesDaniel Bizo2013/02/04 03:45 PM
                                                                            ASP estimatesRichard Cownie2013/02/04 05:01 PM
                                                                              ASP estimatesJS2013/02/05 04:25 AM
                                                                              ASP estimatesDaniel Bizo2013/02/05 06:31 AM
                                                                                ASP estimatesRichard Cownie2013/02/05 06:46 AM
                                                                  ARM OEMs get free silicon?someone2013/02/03 01:44 PM
                                                                    ARM OEMs get free silicon?Richard Cownie2013/02/03 08:28 PM
                                                                      ARM OEMs get free silicon?Doug S2013/02/04 11:08 AM
                                                                        ARM OEMs get free silicon?Richard Cownie2013/02/04 12:17 PM
                                                                          ARM OEMs get free silicon?geeker2013/02/06 12:16 PM
                                                                            ARM OEMs get free silicon?Gabriele Svelto2013/02/07 04:18 AM
                                                                              ARM OEMs get free silicon?izotop2013/02/07 06:11 AM
                                                                            ARM OEMs get free silicon?Richard Cownie2013/02/07 06:48 AM
                                                                Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 06:31 PM
                                                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 06:55 PM
                                                                    Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 07:25 PM
                                                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 07:50 PM
                                                                        Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 08:22 PM
                                                                        Article Online: Microservers must Specialize to SurviveBrendan2013/02/03 12:48 AM
                                                                        Article Online: Microservers must Specialize to SurviveMichael S2013/02/03 02:22 AM
                                                                    Article Online: Microservers must Specialize to Surviveanon2013/02/02 10:21 PM
                                                                      Article Online: Microservers must Specialize to SurviveMichael S2013/02/03 02:40 AM
                                                                        Article Online: Microservers must Specialize to Surviveanon2013/02/03 03:54 AM
                                                                      Overlooking the advantage (or not) of x86 in different marketsDoug S2013/02/03 01:54 PM
                                                                Article Online: Microservers must Specialize to Surviveanon2013/02/02 10:14 PM
                                                                  Article Online: Microservers must Specialize to Surviveanon2013/02/02 10:32 PM
                                                                    Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/03 06:24 AM
                                                        Article Online: Microservers must Specialize to Surviveanon2013/02/02 10:08 PM
                                          Article Online: Microservers must Specialize to SurviveMichael S2013/02/01 06:41 AM
                                            Article Online: Microservers must Specialize to Surviveanon2013/02/01 07:41 AM
                            Article Online: Microservers must Specialize to SurviveDoug S2013/02/01 04:41 PM
                              Article Online: Microservers must Specialize to Surviveanon2013/02/01 07:08 PM
                                Article Online: Microservers must Specialize to SurviveDoug S2013/02/02 02:38 PM
                                  Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 03:03 PM
              Article Online: Microservers must Specialize to Survivebakaneko2013/01/30 01:16 AM
        Article Online: Microservers must Specialize to Surviveanon2013/01/29 11:34 PM
    Article Online: Microservers must Specialize to Survivesomeone2013/01/29 08:58 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 09:28 AM
      Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 09:42 AM
      Article Online: Microservers must Specialize to SurviveMark Roulo2013/01/29 06:08 PM
        Article Online: Microservers must Specialize to Surviverwessel2013/01/29 10:15 PM
          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 02:39 PM
            Article Online: Microservers must Specialize to Surviverwessel2013/01/30 03:54 PM
          Or compare to *desktop* ASPsMark Roulo2013/01/30 03:51 PM
            Or compare to *desktop* ASPsrwessel2013/01/30 04:07 PM
            Or compare to *desktop* ASPsCeleron2013/01/30 05:49 PM
            Or compare to *desktop* ASPsDoug S2013/01/30 09:50 PM
              Or compare to *desktop* ASPsrwessel2013/01/30 11:04 PM
              when thoughput matters, ultra-cheap Intel CPUs are not attractiveMichael S2013/01/31 02:59 AM
                when thoughput matters, ultra-cheap Intel CPUs are not attractiveDoug S2013/01/31 12:00 PM
                  when thoughput matters, ultra-cheap Intel CPUs are not attractiveMichael S2013/01/31 02:04 PM
                    Price/perf curve on intel.Jouni Osmala2013/01/31 11:28 PM
                    when thoughput matters, ultra-cheap Intel CPUs are not attractiveDoug S2013/02/01 04:45 PM
                      when thoughput matters, ultra-cheap Intel CPUs are not attractiverwessel2013/02/02 01:54 AM
              Or compare to *desktop* ASPsGabriele Svelto2013/01/31 03:33 AM
    Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 10:36 AM
      Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 01:52 PM
        Article Online: Microservers must Specialize to SurviveARMandLeg2013/01/29 03:31 PM
        Article Online: Microservers must Specialize to SurviveNo2013/01/29 03:32 PM
          Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 04:36 PM
            Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 07:25 PM
        Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 07:20 PM
          Article Online: Microservers must Specialize to SurviveEduardoS2013/01/30 03:48 AM
  Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/01/29 09:08 AM
  Article Online: Microservers must Specialize to Survivecarop2013/01/29 02:55 PM
    Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 11:33 PM
      Article Online: Microservers must Specialize to SurviveGabriele Svelto2013/01/31 03:51 AM
    Article Online: Microservers must Specialize to Survivebakaneko2013/01/30 01:32 AM
      Cache capacity benefits (babbling a bit)Paul A. Clayton2013/01/30 01:42 PM
        Cache size experiments may have been done on ARMsMark Roulo2013/01/30 02:14 PM
          Way and line locking and associativityPaul A. Clayton2013/01/30 03:08 PM
            Way and line locking and associativityMark Roulo2013/01/30 03:37 PM
        Cache capacity benefits may have been done in simulationPatrick Chase2013/01/30 07:25 PM
          Simulation has benefits but is SLOWPaul A. Clayton2013/01/30 09:08 PM
            Simulation has benefits but is SLOWPatrick Chase2013/01/30 10:36 PM
              Simulation has benefits but is SLOWRichard Cownie2013/02/01 12:32 PM
                Simulation has benefits but is SLOWPatrick Chase2013/02/01 01:06 PM
                Simulation has benefits but is SLOWnone2013/02/01 01:13 PM
      Article Online: Microservers must Specialize to Survivesoupdragon2013/01/31 12:02 PM
    Article Online: Microservers must Specialize to SurvivePatrick Chase2013/01/30 12:26 PM
  4x or 10x advantageS. Rao2013/01/30 05:35 PM
    4x or 10x advantagerwessel2013/01/30 06:32 PM
      4x or 10x advantageS. Rao2013/02/03 10:32 PM
        4x or 10x advantageDavid Kanter2013/02/03 11:09 PM
        4x or 10x advantagerwessel2013/02/04 12:08 AM
          x86_64Richard Cownie2013/02/04 05:52 PM
            x86_64Michael S2013/02/05 02:51 AM
              x86_64Richard Cownie2013/02/05 04:33 AM
                x86_64EduardoS2013/02/05 04:41 AM
                  x86_64sJ2013/02/05 05:52 AM
                  x86_64Richard Cownie2013/02/05 06:47 AM
                x86_64someone2013/02/05 09:22 AM
                  x86_64Richard Cownie2013/02/05 09:44 AM
              "Just recompile"Mark Roulo2013/02/05 08:40 AM
                "Just recompile"Michael S2013/02/05 10:35 AM
                  "Just recompile"Mark Roulo2013/02/05 10:55 AM
  Virtualization and custom interconnectsGabriele Svelto2013/01/31 05:05 AM
  x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/01 10:11 PM
    x86 penalty depends on performance and microarchitecturempx2013/02/02 04:31 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 09:27 AM
        Heterogenous multicore in serversPaul A. Clayton2013/02/02 10:25 AM
    x86 penalty depends on performance and microarchitectureanon2013/02/02 05:04 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 09:42 AM
        x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 09:58 AM
        x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 10:26 AM
        x86 penalty depends on performance and microarchitectureanon2013/02/02 05:49 PM
          x86 penalty depends on performance and microarchitecturePer Hesselgren2013/02/03 01:44 AM
            x86 penalty depends on performance and microarchitectureMichael S2013/02/03 02:59 AM
          x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 02:27 PM
            x86 penalty depends on performance and microarchitectureanon2013/02/03 02:58 PM
              x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 03:35 PM
                x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 03:39 PM
                x86 penalty depends on performance and microarchitectureanon2013/02/03 06:08 PM
                  x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 08:05 PM
                    x86 penalty depends on performance and microarchitectureanon2013/02/03 09:52 PM
                      x86 penalty depends on performance and microarchitectureDavid Kanter2013/02/03 11:23 PM
                        U shape curve on performance...Jouni Osmala2013/02/04 01:07 AM
                          U shape curve on performance...none2013/02/04 03:03 AM
                            U shape curve on performance...Michael S2013/02/04 07:55 AM
                              U shape curve on performance...Patrick Chase2013/02/04 10:14 AM
                                U shape curve on performance...Michael S2013/02/04 02:43 PM
                                  U shape curve on performance...Patrick Chase2013/02/04 04:06 PM
                            Microcode can be usefulPaul A. Clayton2013/02/04 08:08 AM
                          U shape curve on performance...Michael S2013/02/04 07:48 AM
                          U shape curve on performance...Patrick Chase2013/02/04 10:05 AM
                            It only works if ... ??Paul A. Clayton2013/02/04 12:20 PM
                              It only works if ... ??Patrick Chase2013/02/04 12:48 PM
                                It only works if ... ??Patrick Chase2013/02/04 01:02 PM
                                It only works if ... ??Patrick Chase2013/02/04 01:08 PM
                                  It only works if ... ??Patrick Chase2013/02/04 01:15 PM
                                  It only works if ... ??anon2013/02/04 05:45 PM
                            I don't follow...Megol2013/02/05 10:16 AM
                      x86 penalty depends on performance and microarchitecturerwessel2013/02/04 12:28 AM
            spec95 vs Spec2kMichael S2013/02/04 02:27 AM
              spec95 vs Spec2kPer Hesselgren2013/02/04 04:42 AM
              spec95 vs Spec2kbakaneko2013/02/04 04:48 AM
                correct linksMichael S2013/02/04 06:10 AM
              spec95 vs Spec2kPatrick Chase2013/02/04 11:47 AM
                spec95 vs Spec2kPatrick Chase2013/02/04 11:55 AM
                  spec95 vs Spec2kMichael S2013/02/04 01:39 PM
                    spec95 vs Spec2kPatrick Chase2013/02/04 02:07 PM
                    spec95 vs Spec2knone2013/02/04 04:28 PM
                      spec95 vs Spec2kMichael S2013/02/05 02:35 AM
                spec95 vs Spec2kMichael S2013/02/04 01:50 PM
                spec95 vs Spec2k (hair-splitting)Kira2013/02/04 03:26 PM
    x86 penalty depends on performance and microarchitecturenone2013/02/02 05:33 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 10:11 AM
        x86 penalty depends on performance and microarchitecturenone2013/02/02 10:43 AM
          Cortex A9 ROBPatrick Chase2013/02/02 01:13 PM
            Cortex A9 ROBPatrick Chase2013/02/02 02:01 PM
              Cortex A9 ROBnone2013/02/04 04:12 PM
                Cortex A9 ROBPatrick Chase2013/02/04 04:43 PM
                  Cortex A9 ROBnone2013/02/04 04:58 PM
    ARM diversity might be an advantagePaul A. Clayton2013/02/02 11:10 AM
      ARM diversity might be an advantagePatrick Chase2013/02/03 04:29 PM
        ARMv8 cleanlinessanon2013/02/03 06:11 PM
          ARMv8 cleanlinessPatrick Chase2013/02/03 07:40 PM
            Does AArch64 not support table node page sizes?Paul A. Clayton2013/02/03 09:40 PM
              Does AArch64 not support table node page sizes?Patrick Chase2013/02/03 10:14 PM
                Does AArch64 not support table node page sizes?none2013/02/04 02:55 AM
              LPAE supports 2MiB and 1 GiB blocksPaul A. Clayton2013/02/03 10:52 PM
            ARMv8 cleanlinessanon2013/02/03 10:06 PM
              ARMv8 cleanlinessrwessel2013/02/04 12:37 AM
                ARMv8 cleanlinessMichael S2013/02/04 01:50 AM
              ARMv8 cleanlinessEtienne2013/02/04 08:01 AM
                Similar to Thumb If-Then instructionPaul A. Clayton2013/02/04 08:43 AM
                  Similar to Thumb If-Then instructionEtienne2013/02/05 07:22 AM
                ARMv8 cleanlinessJouni Osmala2013/02/04 11:15 PM
                  ARMv8 cleanlinessEtienne2013/02/05 06:57 AM
                    ARMv8 cleanlinessJouni Osmala2013/02/05 07:59 AM
                      ARMv8 cleanlinessEtienne2013/02/05 08:55 AM
              ARMv8 cleanlinessMaynard Handley2013/02/13 03:53 PM
                ARMv8 cleanlinessGabriele Svelto2013/02/14 01:37 AM
                ARMv8 cleanlinessEtienne2013/02/14 06:00 AM
                  ARMv8 cleanlinessrwessel2013/02/14 04:03 PM
                ARMv8 cleanlinessSeni2013/02/14 07:45 AM
                  ARMv8 cleanlinessrwessel2013/02/14 04:14 PM
                    ARMv8 cleanlinessLinus Torvalds2013/02/14 04:44 PM
                      ARMv8 cleanlinessGabriele Svelto2013/02/15 02:28 AM
                        ARMv8 cleanlinessEduardoS2013/02/15 10:45 AM
                        ARMv8 cleanlinessMax2013/02/15 01:24 PM
                          ARMv8 cleanlinessWilco2013/02/15 05:46 PM
                            ARMv8 cleanlinessLinus Torvalds2013/02/15 06:34 PM
                              ARMv8 cleanlinessWilco2013/02/17 01:47 PM
                                ARMv8 cleanlinessLinus Torvalds2013/02/17 04:08 PM
                                  ARMv8 cleanlinessWilco2013/02/18 05:14 PM
                                    ARMv8 cleanlinessanon2013/02/19 01:56 AM
                                      ARMv8 cleanlinessWilco2013/02/19 02:20 PM
                                    ARMv8 cleanlinessLinus Torvalds2013/02/19 10:13 AM
                                      ARMv8 cleanlinessWilco2013/02/19 02:10 PM
                                        ARMv8 cleanlinessLinus Torvalds2013/02/19 06:16 PM
                                          ARMv8 cleanlinessLinus Torvalds2013/02/19 06:41 PM
                                            ARMv8 cleanlinessWilco2013/02/25 06:05 AM
                                              ARMv8 cleanlinessLinus Torvalds2013/02/25 06:22 PM
                                                ARMv8 cleanlinessnone2013/02/26 01:31 AM
                                                  ARMv8 cleanlinessMichael S2013/02/26 05:05 AM
                                                ARMv8 cleanlinessDavid Hess2013/02/26 11:01 PM
                                                  ARMv8 cleanlinessLinus Torvalds2013/02/27 11:22 AM
                                                    ARMv8 cleanlinessDavid Hess2013/03/01 01:58 PM
                                                      What is wrong with ARM?Paul A. Clayton2013/03/01 06:56 PM
                                                        What is wrong with ARM?David Hess2013/03/01 07:49 PM
                                                          What is wrong with ARM?Michael S2013/03/02 05:11 PM
                                                            What is wrong with ARM?David Hess2013/03/03 11:32 PM
                                                              What is wrong with ARM?Michael S2013/03/04 02:05 AM
                                                        What is wrong with ARM?David Hess2013/03/07 06:30 PM
                                                          Thank you very much!Paul A. Clayton2013/03/07 07:59 PM
                                                            Thank you very much!David Hess2013/03/07 11:22 PM
                                                              Thank you very much!gallier22013/03/08 01:43 AM
                                                                Thank you very much!David Hess2013/03/08 11:24 AM
                                                                  Thank you very much!Ungo2013/03/08 04:37 PM
                                                                    Thank you very much!David Hess2013/03/08 08:47 PM
                                                            Thank you very much!Ricardo B2013/03/08 05:07 AM
                                                              ARM lackings a historical accident?Paul A. Clayton2013/03/08 08:54 AM
                                                                ARM lackings a historical accident?Ricardo B2013/03/08 03:33 PM
                                                          What is wrong with ARM?Michael S2013/03/08 06:25 AM
                                                            What is wrong with ARM?David Hess2013/03/08 12:37 PM
                                                            What is wrong with ARM?dmsc2013/03/12 03:47 PM
                                                          What is wrong with ARM?dmsc2013/03/12 03:43 PM
                                                      ARMv8 cleanlinessMichael S2013/03/02 05:06 PM
                                                    ParityAndi Kleen2013/03/03 01:47 PM
                                                      ParityLinus Torvalds2013/03/03 03:06 PM
                                                        ParityAndi Kleen2013/03/03 07:15 PM
                                            Except the borrow bitMatthew D2013/02/26 12:07 AM
                                              Except the borrow bitrwessel2013/02/26 12:42 AM
                                              Except the borrow bitLinus Torvalds2013/02/26 11:13 AM
                                                are ZSCO flags really fundamental?Michael S2013/02/27 04:24 PM
                                                  are ZSCO flags really fundamental?Linus Torvalds2013/02/27 05:41 PM
                                                    are ZSCO flags really fundamental?Michael S2013/02/28 01:39 AM
                                                    are ZSCO flags really fundamental?2013/03/01 10:23 AM
                                                  are ZSCO flags really fundamental?EduardoS2013/02/27 11:07 PM
                                                    are ZSCO flags really fundamental?Michael S2013/02/28 04:58 AM
                                                  are ZSCO flags really fundamental?rwessel2013/02/28 12:29 AM
                                                    are ZSCO flags really fundamental?rwessel2013/02/28 12:32 AM
                                                    Bit numbering on big endian machinesKonrad Schwarz2013/02/28 01:49 AM
                                                      Bit numbering on big endian machinesrwessel2013/02/28 02:32 AM
                                          ARMv8 cleanlinessWilco2013/02/25 05:54 AM
                                            ARMv8 cleanlinessEduardoS2013/02/25 06:24 AM
                                              ARMv8 cleanlinessWilco2013/02/25 07:36 AM
                                                ARMv8 cleanlinessEduardoS2013/02/25 01:06 PM
                                                  ARMv8 cleanlinessWilco2013/02/25 05:24 PM
                                                    ARMv8 cleanlinessEduardoS2013/02/25 05:55 PM
                                                    Utility of data loss exceptionsPaul A. Clayton2013/02/25 08:18 PM
                                                      Utility of data loss exceptionsrwessel2013/02/26 01:00 AM
                                                        Excessively expensive mode changingPaul A. Clayton2013/02/26 06:15 AM
                                                      Utility of data loss exceptionsEduardoS2013/02/26 02:55 AM
                                                        Utility of data loss exceptionsTREZA2013/02/26 04:53 AM
                                                          Utility of data loss exceptionsEduardoS2013/02/26 05:37 AM
                                                      Utility of data loss exceptionsEtienne2013/02/26 06:01 AM
                                            1b vs. 4b not a huge difference?Paul A. Clayton2013/02/25 09:57 AM
                                  Possible benefit of multi-bit CCPaul A. Clayton2013/02/23 08:09 AM
                                ARMv8 cleanlinessrwessel2013/02/18 12:01 AM
                                  ARMv8 cleanlinessWilco2013/02/18 05:33 PM
                                    ARMv8 cleanlinessEduardoS2013/02/18 08:18 PM
                                      ARMv8 cleanlinessWilco2013/02/25 06:19 AM
                      ARMv8 cleanlinessTREZA2013/02/15 04:21 AM
                      ARMv8 cleanliness2013/02/17 10:49 AM
                        ARMv8 cleanlinessLinus Torvalds2013/02/17 11:54 AM
                          ARMv8 cleanlinessLinus Torvalds2013/02/17 11:58 AM
                          ARMv8 cleanlinessnone2013/02/17 04:36 PM
                        ARMv8 cleanlinessTREZA2013/02/17 11:59 AM
                          ARMv8 cleanlinessWilco2013/02/17 12:30 PM
                            and-compare and or-comparePaul A. Clayton2013/02/17 05:28 PM
                              early branch resolutionanon2013/02/18 02:05 AM
                                My ignorance about compilers is vasterPaul A. Clayton2013/02/18 01:30 PM
                                early branch resolutionanon2013/02/18 05:12 PM
                              early branch resolution?EBFE2013/02/19 01:03 AM
                                Early resolution: some but limited potentialPaul A. Clayton2013/02/19 05:52 PM
                                  Early resolution: some but limited potentialhobold2013/02/20 02:02 PM
                                    PPC750 was an inspirationPaul A. Clayton2013/02/23 08:09 AM
                          ARMv8 cleanlinessTREZA2013/02/17 01:53 PM
                            ARMv8 cleanlinessExophase2013/02/17 04:49 PM
                        ARMv8 cleanlinessWilco2013/02/17 12:07 PM
                        ARMv8 cleanlinessbakaneko2013/02/17 12:43 PM
                        ARMv8 cleanlinessJouni Osmala2013/02/17 01:10 PM
                          -1 comparison true seems betterPaul A. Clayton2013/02/17 06:07 PM
                  ARMv8 cleanlinessEtienne2013/07/17 02:58 AM
      ARM server benchmark- why Apache?Per Hesselgren2013/02/05 12:13 AM
        ARM server benchmark- why Apache?anon2013/02/05 12:38 AM
          ARM server benchmark- why Apache?Per Hesselgren2013/02/05 01:25 AM
            ARM server benchmark- why Apache?Michael S2013/02/05 02:13 AM
          ARM server benchmark- why Apache?Michael S2013/02/05 01:51 AM
          I think that was debunked alreadyDavid Kanter2013/02/05 07:36 PM
        ARM server benchmark- why Apache?bakaneko2013/02/06 02:37 AM
          ARM server benchmark- why Apache?Michael S2013/02/06 03:43 AM
            Marketing to PHBs?Paul A. Clayton2013/02/06 06:40 AM
            ARM server benchmark- why Apache?bakaneko2013/02/06 07:08 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?