Some 'reverse-engineering' of Zen

By: juanrga (nospam.delete@this.juanrga.com), October 31, 2015 11:50 am
Room: Moderated Discussions
lurker (lurker9000.delete@this.realemail.mail) on October 31, 2015 8:07 am wrote:
> juanrga (nospam.delete@this.juanrga.com) on October 31, 2015 6:58 am wrote:
> >
> > I tried to explain that what matters is not the total number of memory units
> > but the ratio between different execution units, the ISA, and the workloads.
> >
> > As explained before 2AGU+4ALU looks fine for ARM mobile SoC. 2AGU+4ALU looks unbalanced for x86 server SoC.
> >
> > Same stuff about HPC. A third AGU is not obligatory for HPC. Intel's Borkar just gave a recent
> > talk about his vision for a future supercomputer much faster than actual. The HPC core that he
> > proposes has only one AGU, and that is fine because his core has a 1:1 ratio for FP:mem.
> >
> > Fujitsu presented this year the new CPU that will replace the old CPU in the successful
> > K-computer. This new CPU, the XIfx, is exclusively optimized for HPC and each
> > core has 2 AGUs. Again fine because it has a 1:1 ratio for FP:mem...
> >
> > But this discussion about HPC is rather irrelevant, because we know
> > that AMD will not use CPUs for HPC. AMD will propose HSA APUs.
> >
> > About markets Zen can win back easier. I know none. I am rather convinced that HPC Zen-based APUs
> > will fail. Zen on AM4 APUs for desktop/mobile will be too late and competing against Kabylake/Cannonlake.
> > Zen AM4 for CPUs will be competing with Skylake-E CPUs for a niche market: HEDT. And I guess Zen on
> > servers will be crushed by both sides: ARM (Qualcomm, Broadcom, APM,...) and x86 (Intel).
>
> I wouldn't be so negative before it's even released(or before we have some sort of performance indicators
> for it). There's a bunch of engineers with a lot of experience at AMD after all, I'm sure there's a reason
> why it's 2AGU+4ALU and not 3+3 or 2+3. What's the point of 4ALUs if one or two of them will remain idle?
> I'd imagine that extra ALUs, even when not fully utilized, would suck up a lot of power and generate more
> heat. Wouldn't getting rid of it, or at least disabling it, give better performance/watt?

I think I am being simply realist. We know the claimed IPC for Zen; we know Papermaster admitted @ FAD 2015, that it is not enough to match Intel; we know AMD will use a 14nm node that will not hit clocks so high as Intel 14nm; we know AMD lose the moar core advantage on HEDT...

I am sure the reason for 4ALU+2AGU and 128bit FP pipes is not because it is the best possible configuration. The real reason? We only can speculate at this time. Maybe a cache bottleneck did make adding a third AGU useless, maybe the fourth ALU is here for symmetry reasons, maybe...
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Some 'reverse-engineering' of Zenjuanrga2015/10/03 03:48 AM
  Some 'reverse-engineering' of ZenHeikki Kultala2015/10/03 07:01 AM
    Some 'reverse-engineering' of Zenitsmydamnation2015/10/03 02:47 PM
      Some 'reverse-engineering' of ZenAnders2015/10/03 10:04 PM
        Some 'reverse-engineering' of ZenContrarian2015/10/04 10:53 AM
          Some 'reverse-engineering' of ZenJukka Larja2015/10/05 08:15 AM
            Some 'reverse-engineering' of Zenmpx2015/10/05 08:37 AM
              Some 'reverse-engineering' of ZenContrarian2015/10/25 11:25 AM
                Some 'reverse-engineering' of Zenitsmydamnation2015/10/26 12:49 AM
                Some 'reverse-engineering' of Zenjuanrga2015/10/29 01:16 AM
                  Some 'reverse-engineering' of Zenanon2015/10/30 05:54 AM
                    Some 'reverse-engineering' of Zenjuanrga2015/10/30 09:37 AM
                    POWER8 load/store unitsGabriele Svelto2015/10/30 05:34 PM
                Some 'reverse-engineering' of Zenlurker2015/10/29 02:12 PM
                  Some 'reverse-engineering' of ZenDavid Kanter2015/10/29 11:06 PM
                    Some 'reverse-engineering' of Zenlurker2015/10/30 01:39 AM
                      Some 'reverse-engineering' of ZenGabriele Svelto2015/10/30 02:14 AM
                        Some 'reverse-engineering' of Zenlurker2015/10/30 02:41 AM
                      Some 'reverse-engineering' of Zendmcq2015/10/30 04:12 AM
                        Some 'reverse-engineering' of Zenlurker2015/10/30 04:56 AM
                          Some 'reverse-engineering' of Zendmcq2015/10/30 05:09 AM
                            Some 'reverse-engineering' of Zennone2015/10/30 05:16 AM
                          Some 'reverse-engineering' of ZenDavid Kanter2015/10/30 07:16 AM
                            Some 'reverse-engineering' of Zenlurker2015/10/30 07:44 AM
                          Some 'reverse-engineering' of Zenjuanrga2015/10/30 12:00 PM
                            AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)Heikki Kultala2015/10/30 02:45 PM
                              AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)juanrga2015/10/31 05:20 AM
                                AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)Heikki Kultala2015/10/31 01:19 PM
                                  AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)juanrga2015/11/01 06:41 AM
                                    AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)Anne O. Nymous2015/11/01 08:39 AM
                                AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)Matthias Waldhauer2015/11/02 12:26 PM
                                  AVX and 128-bit FPUs in Zen(Re:Some 'reverse-engineering' of Zen)juanrga2015/11/03 02:42 AM
                            Some 'reverse-engineering' of ZenMatthias Waldhauer2015/11/02 10:39 AM
                        Some 'reverse-engineering' of Zenbakaneko2015/10/31 06:28 AM
                          Some 'reverse-engineering' of Zendmcq2015/10/31 07:19 AM
                            Some 'reverse-engineering' of ZenJukka Larja2015/10/31 08:21 AM
                            Some 'reverse-engineering' of Zenbakaneko2015/10/31 09:23 AM
                              Some 'reverse-engineering' of Zendmcq2015/10/31 03:12 PM
                                Some 'reverse-engineering' of Zenbakaneko2015/10/31 04:25 PM
                                  Some 'reverse-engineering' of Zendmcq2015/11/01 06:36 AM
                                    Some 'reverse-engineering' of Zenbakaneko2015/11/01 09:11 AM
                                      Some 'reverse-engineering' of Zendmcq2015/11/01 09:27 AM
                                        Some 'reverse-engineering' of Zenbakaneko2015/11/01 02:35 PM
                                          Some 'reverse-engineering' of Zendmcq2015/11/01 03:52 PM
                                            Some 'reverse-engineering' of Zenbakaneko2015/11/03 02:17 AM
                                              Some 'reverse-engineering' of Zendmcq2015/11/03 03:17 AM
                  Some 'reverse-engineering' of Zenjuanrga2015/10/30 11:43 AM
                    Some 'reverse-engineering' of Zenlurker2015/10/30 01:09 PM
                      Some 'reverse-engineering' of Zenjuanrga2015/10/31 05:58 AM
                        Some 'reverse-engineering' of Zenlurker2015/10/31 07:07 AM
                          Some 'reverse-engineering' of Zenjuanrga2015/10/31 11:50 AM
                            Some 'reverse-engineering' of ZenPoindexter2015/10/31 01:47 PM
                              Some 'reverse-engineering' of Zenlurker2015/10/31 03:06 PM
                                Some 'reverse-engineering' of ZenPoindexter2015/10/31 04:37 PM
                                  Some 'reverse-engineering' of Zenlurker2015/11/01 02:46 AM
                                  Some 'reverse-engineering' of Zenjuanrga2015/11/01 07:16 AM
                                    Some 'reverse-engineering' of ZenMaynard Handley2015/11/01 05:33 PM
                                      Some 'reverse-engineering' of Zenjuanrga2015/11/02 04:06 AM
                                        Zen transistor countHeikki Kultala2015/11/04 12:30 AM
                                          Zen transistor countjuanrga2015/11/05 04:34 AM
                                    Some 'reverse-engineering' of ZenSymmetry2015/11/02 05:56 AM
                                      Some 'reverse-engineering' of ZenDavid Hess2015/11/02 06:16 AM
                                      Some 'reverse-engineering' of Zennobody2015/11/02 07:19 AM
                                        Some 'reverse-engineering' of ZenJukka Larja2015/11/02 08:34 PM
                                          Some 'reverse-engineering' of Zennobody2015/11/03 12:35 AM
                                            Some 'reverse-engineering' of ZenJukka Larja2015/11/03 01:41 AM
                                              Some 'reverse-engineering' of Zennobody2015/11/03 04:10 AM
                                                Some 'reverse-engineering' of ZenJukka Larja2015/11/04 01:52 AM
                                            Some 'reverse-engineering' of ZenKlimax2015/11/03 01:47 AM
                                              Some 'reverse-engineering' of Zennobody2015/11/03 03:19 AM
                                                Some 'reverse-engineering' of ZenKlimax2015/11/03 11:09 AM
                                          Some 'reverse-engineering' of ZenBanana_Comedown2015/11/04 10:17 AM
                                            Some 'reverse-engineering' of ZenJukka Larja2015/11/05 02:51 AM
                                              Some 'reverse-engineering' of ZenBanana_Comedown2015/11/05 12:23 PM
                                      Some 'reverse-engineering' of Zenjuanrga2015/11/03 03:13 AM
                                Some 'reverse-engineering' of Zennobody2015/10/31 07:16 PM
                                  Some 'reverse-engineering' of Zenlurker2015/11/01 02:48 AM
                                    Some 'reverse-engineering' of Zensylt2015/11/01 03:07 AM
                                    Some 'reverse-engineering' of Zennobody2015/11/01 10:49 AM
                                  Some 'reverse-engineering' of Zenmd2015/11/01 12:06 PM
                                    Some 'reverse-engineering' of Zenblu2015/11/01 12:59 PM
                                      Some 'reverse-engineering' of Zenjuanrga2015/11/02 04:10 AM
                              Some 'reverse-engineering' of Zenjuanrga2015/11/01 06:00 AM
                                Some 'reverse-engineering' of ZenPoindexter2015/11/01 08:25 AM
                                  Some 'reverse-engineering' of Zenjuanrga2015/11/02 03:47 AM
                                    Some 'reverse-engineering' of ZenPoindexter2015/11/02 04:39 AM
                                      Some 'reverse-engineering' of Zenjuanrga2015/11/02 12:26 PM
                                        Some 'reverse-engineering' of ZenDan Downs2015/11/02 02:12 PM
                                          Some 'reverse-engineering' of Zenjuanrga2015/11/03 02:51 AM
                            Some 'reverse-engineering' of Zenlurker2015/10/31 02:57 PM
                              Some 'reverse-engineering' of Zenjuanrga2015/11/01 06:26 AM
                                Some 'reverse-engineering' of ZenHeikki Kultala2015/11/01 08:57 AM
                                  Some 'reverse-engineering' of ZenMaynard Handley2015/11/01 05:39 PM
                              Some 'reverse-engineering' of ZenUngo2015/11/01 01:05 PM
                  Some 'reverse-engineering' of Zenquzhujian2015/10/31 09:39 AM
                  Some 'reverse-engineering' of ZenPoindexter2015/11/04 05:10 AM
                    Some 'reverse-engineering' of Zenlurker2015/11/05 12:54 PM
                      Some 'reverse-engineering' of Zennobody2015/11/05 02:28 PM
                      The sad status of tech mediajuanrga2015/11/06 04:19 AM
                        The sad status of tech mediaJoel2015/11/06 05:45 PM
                        The sad status of tech mediaMatthias Waldhauer2015/11/07 02:12 PM
                          The sad status of tech mediaJoel2015/11/07 04:40 PM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell tangerine? ūüćä