Intel's roadmap

By: Maynard Handley (name99.delete@this.name99.org), May 13, 2019 11:04 am
Room: Moderated Discussions
Alberto (git.delete@this.git.it) on May 13, 2019 9:44 am wrote:
> Laurent (anon.delete@this.anon.com) on May 13, 2019 8:37 am wrote:
> > Intel released their roadmap for the next few years.
> >
> > https://www.nextplatform.com/2019/05/13/intel-gives-moores-law-a-makeover/
> > https://www.anandtech.com/show/14312/intel-process-technology-roadmap-refined-nodes-specialized-technologies
> >
> > How credible is it?
> >
> > Best quote: “And what I have found in discussions with many of you is the perception
> > that Intel’s process innovation has slowed down during this time.”
>
> As lithography slowed down, Intel slowed down.
> As new nodes not allowed higher clocks anymore, Intel slowed down.
>
> Anyway Ice Lake Ep apperently will be online early than expected sometime in the first half
> next year. A nice speedup i have to say.

What do you base this "nice speedup" on? Do numbers exist anywhere for single-threaded speedup (ie how much faster does gcc go)?
I've seen nothing of that form from Intel and the (true? who knows? but they look plausible, not an obvious fake) numbers on GeekBench are not impressive.

https://browser.geekbench.com/v4/cpu/compare/11493813?baseline=11585297
https://browser.geekbench.com/v4/cpu/compare/13056449?baseline=11585297

This is presumably early silicon. The peak frequency (2.3GHz) will surely rise. But the IPC seems unlikely to rise, and it seems to be a very minor improvement on *Lake; not a great leap forward, and nothing close to the similarly clocked A12X.
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Intel's roadmapLaurent2019/05/13 08:37 AM
  Intel's roadmapAlberto2019/05/13 09:44 AM
    Intel's roadmapblue2019/05/13 10:26 AM
    Intel's roadmapMaynard Handley2019/05/13 11:04 AM
      Intel's roadmapAdrian2019/05/13 01:15 PM
      Actually not bad for IntelChester Lam2019/05/14 05:26 PM
        Actually not bad for IntelMaynard Handley2019/05/14 06:33 PM
          Actually not bad for IntelChester Lam2019/05/14 08:52 PM
            Easily, just grab the LN2... (NT)blue2019/05/14 10:41 PM
            Actually not bad for IntelMaynard Handley2019/05/14 11:32 PM
              Application mattersChester Lam2019/05/15 03:15 AM
      Intel's roadmapAlberto2019/05/15 07:58 AM
        Intel's roadmapnone2019/05/15 08:25 AM
        Intel's roadmapChester Lam2019/05/15 08:32 AM
          Sh*** is Sh** foreverAlberto2019/05/15 08:47 AM
            Sh*** is Sh** forevernone2019/05/15 09:05 AM
              benchmarks...Chester Lam2019/05/15 09:33 AM
                benchmarks...none2019/05/15 10:09 AM
                  benchmarks...Chester Lam2019/05/15 04:51 PM
                    benchmarks...Doug S2019/05/16 01:10 PM
                      benchmarks...chester lam2019/05/16 03:20 PM
                        benchmarks...Doug S2019/05/16 03:28 PM
                          benchmarks...chester lam2019/05/16 04:00 PM
                            benchmarks...Doug S2019/05/17 03:39 AM
                              benchmarks...Chester Lam2019/05/17 04:54 AM
                                benchmarks...Doug S2019/05/17 11:52 AM
                                  because per-core perf has stagnatedchester lam2019/05/17 01:55 PM
                                    because per-core perf has stagnatedNathan2019/05/17 02:54 PM
                                      because per-core perf has stagnatedAdrian2019/05/17 10:39 PM
                                        because per-core perf has stagnatedchester lam2019/05/19 06:32 PM
                                          because per-core perf has stagnatedAdrian2019/05/19 09:09 PM
                                            because per-core perf has stagnatedFoo_2019/05/20 12:58 AM
                                              because per-core perf has stagnatedMichael S2019/05/20 01:48 AM
                                                because per-core perf has stagnatedAdrian2019/05/20 03:12 AM
                                                  because per-core perf has stagnatedMichael S2019/05/20 04:23 AM
                                                    because per-core perf has stagnatedMichael S2019/05/20 07:07 AM
                                                  because per-core perf has stagnatedAdrian2019/05/20 04:28 AM
                                                    because per-core perf has stagnatedMichael S2019/05/20 07:15 AM
                                                  because per-core perf has stagnatednone2019/05/20 04:41 AM
                      benchmarks...Maynard Handley2019/05/16 04:35 PM
                        benchmarks...dmcq2019/05/20 08:15 AM
                    benchmarks...Maxwell2019/05/16 10:47 PM
            Sh*** is Sh** foreverAnon2019/05/15 09:43 AM
    Intel's roadmapDoug S2019/05/13 01:24 PM
      Intel's roadmapwumpus2019/05/14 07:52 AM
      Intel's roadmapAlberto2019/05/15 08:10 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?