SPARC? I was shocked when I found out that the failure of a sing processor could bring Solaris down.
-- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 ________________________________________ From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of Phil Smith III <li...@akphs.com> Sent: Sunday, May 12, 2019 2:25 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: mainframe hacking "success stories"? Charles Mills wrote, in part: >The mainframe seems to me to have also some "architectural" advantages. It >seems to support a denser "clustering." It does not seem to me that there is >anything in the Windows/Linux world that duplicates the advantages of 100 or >so very-closely-coupled (sharing all main storage potentially) CPUs. Sure, >you can link a thousand Windows or Linux 8-way servers on a super-fast net, >and it is fine for some things -- incredibly powerful for some of them, but >it seems there are some things the mainframe architecture is inherently >better at. Actually, various RISC servers such as SPARC have servers with hundreds of processors sharing memory; even Intel machines do. Xeon Phi, for example: https://secure-web.cisco.com/1-JGZXfo-qWk5_L8pUhWZ-gT8ZNUIpk2EWltvIjRVztYIcr80ud7r4wu9yANKu5-EOPXPOE1cH0ThDuC28ZiHwQX2Ytu0InGGKy4F4idP1vEE8K8Sv0vjy5torsAHRQUHebvWFXdi5Lqe5TC5CzzvXvujXmxITcdCpsl8WHPma30lJIT6UeXdna5Ptp0HsoeJDjA-_FTPR1cuHi36KCkrcWIebP3xBLdVjqNDtPiQVJE4nIaA5sKjgKr5aQy0loVxdYC03SxCk-ZTLJ61R_qft9Va4GYmlSHmPHk5l8MF64I8a6DKgbqFKyzTo51JHQi-z0t94sRX1_sm3FvZ1kDLBEP0AdjmHppLXV5tG1nLqPftMjKOX0UoWrsGcunLX0Xe0dhHyJY-9t9-CBVj0FgVSv9B_XkExdM1KA0nRzgniOekEzGanMCzDaRHfsiTMJj2/https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FXeon_Phi Up to 72 cores per chip, so up to 144 threads per socket. On an eight-socket motherboard, that's, um, a lot. And I'm not sure I buy the value of this on IBM Z anyway: what's the largest single image you know of in terms of CPUs? Yes, a z14 can have 170, but nobody runs that way. It's like owning a Bugatti and driving it downtown: you've got the biggest on the block, but you can't use most of its capabilities. Friend's comment along these lines: It's a mug's game to obsess about the biggest possible model in a product line, as most people don't need or buy them (it isn't to Z advantage anyway). There is a price premium for the "biggest box" and it's less flexible, and more of a single point of failure or single planned outage. General practice is to buy boxes that are big enough with room to spare of course, add more as needed, and run modern applications that don't need a monolithic single system in the first place. And Bill Johnson wrote: >I'm a huge fan of the mainframe. And security is not the ONLY reason for >staying on it. But is a major reason large companies do. Assertion without evidence, easy to ignore. I tend to share this bias but have no compelling evidence that IBM Z in general, or even z/OS specifically, is inherently more secure than another platform. It may tend to be, by tradition and history-that is, typical mainframe security posture and change methodology leads to greater security-but that's not inherent in the platform. Can you support this claim? Your article <https://secure-web.cisco.com/1rMJVojybxJj3F93_rg5B2rTcXI15RsC0ir227kmz7YejMhMztyiWCXO9etAj4AnQP-WKxneSDCBuIk6Dziep-SIrRf7dA04R1tMrYTTKf9X73oYaJmPYlRvqEINSWZbWrn5LAG9iQOtug79v8SAooxTVz4uXMjptTHA0vri6OSg0_UudbVqFSqX2wSb-NR9mkPGJRZ5yzcT-dGN6cw1MGuVYcdm7TjkyUM9Y9seu1LwhtGl--3qSYPwcwZ5dG2tBoQSbllHwP-eNdjuRFRVr42IP9cWKDb_XCKj6k5S9Wmr6dKA8nyPMTqyLYP_2kJ9jAxaNXm-Wfu4GojkFIXD_k-qexEaq0QMdGmtJTCi-05OwKWfoUJsgedgSUZMxLokCEynPdxPUlbILN9TPm3eX8gF1nora269tSyl81sskyUMLMCanWAGGn9sT7eeXfang/https%3A%2F%2Fwww.ibm.com%2Fsupport%2Fknowledgecenter%2Fzosbasics%2Fcom.ibm.zos.zmainframe%2Fzconc_whousesmf.htm> , BTW, is (a) from IBM ("never ask the barber if you need a shave") and (b) proves my point more than yours: The mainframe owes much of its popularity and longevity to its inherent reliability and stability, a result of careful and steady technological advances that have been made since the introduction of the System/360T in 1964. No other computer architecture can claim as much continuous, evolutionary improvement, while maintaining compatibility with previous releases. The first sentence says "This has a long history and is therefore good", which of course makes no sense. The second says "compatibility [legacy!] is what makes this good". It goes on to make laughable assertions: Many of today's busiest Web sites store their production databases on a mainframe host. Seriously? "busiest"? Like, say, Google, Instagram, Facebook, Pinterest, Amazon, Wal-Mart? I don't think so. (Yes, Wal-Mart uses IBM Z but their backend is Teradata.) IBM really needs to stop saying dumb things like this, as it just makes the rest of the world snicker. Or: Corporations use mainframes for applications that depend on scalability and reliability. For example, a banking institution could use a mainframe to host the database of its customer accounts, for which transactions can be submitted from any of thousands of ATM locations worldwide. Nothing inherent to IBM Z there. Those ATMs aren't running on IBM Z, either, eh? Again, apply some critical thinking to the claims. They just don't stand up. 20 years ago, perhaps. Today, not so much. And this makes me sad, both for the platform and for IBM, who seem to be denying the reality. .phsiii ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
