because per-core perf has stagnated

By: chester lam (, May 19, 2019 5:32 pm
Room: Moderated Discussions
> > . For gamers, it's certainly worth it upgrading from Sandy Bridge. For non-gamers, a SSD and/or RAM
> > will provide a much bigger bang/buck.
That's certainly true. I understated perf improvement from SNB to SKL, especially the later ones clocking close to 5 GHz. However, I don't like that Anandtech left out plenty of games that don't perform well even with the lastest CPUs. Civ/Total War late game turn times can leave you waiting for minutes. MMO games like Elder Scrolls Online quickly become CPU bound and dip under 20 fps, even with a quad core Skylake at 4.5 GHz (and no, not all cores are being used). Badly optimized indie games like PUBG stutter badly on the latest CPUs too.

> For floating-point tasks, Haswell has doubled the throughput, then Skylake
> Server has doubled it again. These are not minor improvements.
Your code needs to use FMA to get Haswell's doubled FMA throughput. On Skylake server, your code needs to use 512-bit vectors to get that advantage. That doesn't apply to a lot of existing programs.

> Moreover, integer multi-precision computations have also been accelerated a lot since Broadwell.
Hmm, I'll need to read up on that. But that sounds like another area that requires recompilation/rewriting code, so adoption will also be slow and limited.
> So the speed of these common computer tasks has not increased since Sandy Bridge, but at least you
> can do them now on a 15 W CPU at the same speed as they were done on a 95 W Sandy Bridge CPU (I know
> for sure that on an Intel NUC with Coffee Lake that I am using now the program compilation speed is
> greater than on my old Sandy Bridge workstation, while the SSD speed was the same, so there is no
> influence from that), so it cannot be said that CPU progress has stagnated in all directions.

That's good. We got a good perf jump from 2011 to 2019. However, I don't think that's comparable to the jump from 2003 to 2011. The difference between a 3 GHz Prescott Pentium 4 and a 3.5 GHz Sandy Bridge is a lot bigger.

But sorry about bringing up the SNB -> SKL point - that was a bit of a distraction, and I did underestimate the perf improvement. Back to the point of 'why do we need faster CPUs' (aka 'why does anyone need more than 640k of RAM' argument in 2019):

The idea that if a task benefits from more CPU power, it's a niche thing - hasn't that been the case for the past two decades? We've always had the 'niches' of content creation, data analysis, and gaming. And we've had the things common to everyone like email that were fast enough on a 5 MHz 8086 and 640K of RAM. Should we have stopped working on CPUs decades ago? Maybe.

But wait, AMD's share price has more than doubled from 2017 (just before Ryzen's launch) to 2019. From, AMD's increased revenue comes from selling chips with more cores/threads than similarly priced intel chips. Apps that require more CPU power and are well threaded might be a niche, but it's a pretty freaking big and profitable niche :)

Now remember that improving pre-core performance covers an even bigger area, because better cores obviously help well threaded programs as well. How big do you think that niche would be?
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Intel's roadmapLaurent2019/05/13 07:37 AM
  Intel's roadmapAlberto2019/05/13 08:44 AM
    Intel's roadmapblue2019/05/13 09:26 AM
    Intel's roadmapMaynard Handley2019/05/13 10:04 AM
      Intel's roadmapAdrian2019/05/13 12:15 PM
      Actually not bad for IntelChester Lam2019/05/14 04:26 PM
        Actually not bad for IntelMaynard Handley2019/05/14 05:33 PM
          Actually not bad for IntelChester Lam2019/05/14 07:52 PM
            Easily, just grab the LN2... (NT)blue2019/05/14 09:41 PM
            Actually not bad for IntelMaynard Handley2019/05/14 10:32 PM
              Application mattersChester Lam2019/05/15 02:15 AM
      Intel's roadmapAlberto2019/05/15 06:58 AM
        Intel's roadmapnone2019/05/15 07:25 AM
        Intel's roadmapChester Lam2019/05/15 07:32 AM
          Sh*** is Sh** foreverAlberto2019/05/15 07:47 AM
            Sh*** is Sh** forevernone2019/05/15 08:05 AM
              benchmarks...Chester Lam2019/05/15 08:33 AM
                benchmarks...none2019/05/15 09:09 AM
                  benchmarks...Chester Lam2019/05/15 03:51 PM
                    benchmarks...Doug S2019/05/16 12:10 PM
                      benchmarks...chester lam2019/05/16 02:20 PM
                        benchmarks...Doug S2019/05/16 02:28 PM
                          benchmarks...chester lam2019/05/16 03:00 PM
                            benchmarks...Doug S2019/05/17 02:39 AM
                              benchmarks...Chester Lam2019/05/17 03:54 AM
                                benchmarks...Doug S2019/05/17 10:52 AM
                                  because per-core perf has stagnatedchester lam2019/05/17 12:55 PM
                                    because per-core perf has stagnatedNathan2019/05/17 01:54 PM
                                      because per-core perf has stagnatedAdrian2019/05/17 09:39 PM
                                        because per-core perf has stagnatedchester lam2019/05/19 05:32 PM
                                          because per-core perf has stagnatedAdrian2019/05/19 08:09 PM
                                            because per-core perf has stagnatedFoo_2019/05/19 11:58 PM
                                              because per-core perf has stagnatedMichael S2019/05/20 12:48 AM
                                                because per-core perf has stagnatedAdrian2019/05/20 02:12 AM
                                                  because per-core perf has stagnatedMichael S2019/05/20 03:23 AM
                                                    because per-core perf has stagnatedMichael S2019/05/20 06:07 AM
                                                  because per-core perf has stagnatedAdrian2019/05/20 03:28 AM
                                                    because per-core perf has stagnatedMichael S2019/05/20 06:15 AM
                                                  because per-core perf has stagnatednone2019/05/20 03:41 AM
                      benchmarks...Maynard Handley2019/05/16 03:35 PM
                        benchmarks...dmcq2019/05/20 07:15 AM
                    benchmarks...Maxwell2019/05/16 09:47 PM
            Sh*** is Sh** foreverAnon2019/05/15 08:43 AM
    Intel's roadmapDoug S2019/05/13 12:24 PM
      Intel's roadmapwumpus2019/05/14 06:52 AM
      Intel's roadmapAlberto2019/05/15 07:10 AM
Reply to this Topic
Body: No Text
How do you spell avocado?