With the Google, Amazon, and Microsoft/Bing search engines programming FPGAs with their algorythyms, aren't at that stage? I mean FPGAs are like a level below RISC instruction sets. https://www.forbes.com/sites/moorinsights/2017/08/28/microsoft-fpga-wins-versus-google-tpus-for-ai/#444247e39045
Has any IBM engineers considered experimenting with z/CPUs & z/Channels & Crypto cards built with FPGAs? Might be ready for z/16, z/15 would probably be too soon. Maybe a Micro z/15 on a 1U or 2U blade server with 16/32 FICON on the back? 2-4 LPARS max? Maybe even fits into a slot in a DS8xxx storage box? On Wed, Jan 24, 2018 at 7:51 AM, John McKown <[email protected]> wrote: > On Wed, Jan 24, 2018 at 7:41 AM, Elardus Engelbrecht < > [email protected]> wrote: > >> John McKown wrote: >> >> >Opinion article on Vulture Central about the end of an era for Moore's >> Law. >> >http://www.theregister.co.uk/2018/01/24/death_notice_for_moores_law/ >> >> Hmmmm, I usually read those register articles with a pinch of salt, but it >> seemed they have something today ... >> >> Shift is now apparently on software algoritms instead of hardware >> optimization and improvements ... >> > > I have a vague memory from long ago. When RISC machines were first coming > out (IIRC), it was said that the days of CISC machines were over because > the compilers on the RISC would do all the optimization needed for high > performance. I.e. a lot of the "out of order" sequencing (to avoid AGI & > optimize register usage) would be done by the compilers instead of the > hardware. That is one argument used by IBM for people to use Metal C > instead of HLASM. > > >> >> Ok, here is an interesting article describing that law 'number of >> transistors doubles in about 2 years'. >> >> https://en.wikipedia.org/wiki/Moore's_law >> >> Enjoy... >> >> Groete / Greetings >> Elardus Engelbrecht >> > > > -- > I have a theory that it's impossible to prove anything, but I can't prove > it. > > Maranatha! <>< > John McKown > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN -- Mike A Schwab, Springfield IL USA Where do Forest Rangers go to get away from it all? ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
