If you look at (I think) Appendix "A" of "Linux for the S/390" you will find a comparison of the mainframe's CPs and Intel's x86 architecture... as if comparing them makes much sense.
Remember this from http://www.mainframes.com/whatis.htm ? >>>>>>>> 1) Maximum reliable single-thread performance: Some processes, such as the merge phase of a sort/merge (sorting can be subdivided...) MUST be run single thread. Other operations (balancing b-trees, etc) are single thread and tend to lock out other accesses. Therefore, single thread performance is critical to reasonable operations against a DataBase (especially when adding new rows). <<<<<<<< Compared to an S/390 or zSeries CP, the x86 and brethren architectures are wildly unreliable... but, if you can cope with their foibles, give exceptional performance. (IIRC even a PowerPC can give a CP a run for its money... but, then, even the PPC doesn't run parallel CPUs in order to compare their results.) While I personally believe that IBM was protecting their partner who provided FlexES, a commercial-grade S/390 (and zSeries?) emulator for contractual reasons, at some point it would behoove IBM to find a path to "peaceful coexistence" with Hercules. (Right now IBM seems to be turning as blind and eye towards it though seems to want to blind the eyes of others looking in that direction.) I dunno... maybe there's a desire to avoid seeing it as "MainframeLITE" or the mainframe equivalent of a PCjr... I once had a tiny S/390 -- I forget the name of it, it had a couple of PCI cards with an S/390 on 'em and a byte multiplexor channel, all in a box running OS|2-- in one of my labs in Tampa though I was only working with it to make sure the it was properly configured into my network. I think I gave the VM TCP/IP gateway an IP address of .42 and plugged it into my DNS server as "deepthought". (I forget what I named the OS|2 side of the machine.) I *really* wanted to pull the cards and slap 'em into one of the 7025-F50s I had so that I'd get a better boost using AIX to support I/O for the beastie. (sighs) I kinda miss those days. >>>>>>>> 2) Maximum I/O Connectivity: Mainframes excel at providing a convenient paradigm for HUGE disk farms; While SAN devices kind of weaken this to some degree, SAN devices mimic the model of the Mainframe in connectivity "tricks" (at least internally). 3) Maximum I/O Bandwidth: Despite the huge quantities of drives that may be attached to a mainframe, the drives are connected in such a way that there are very few choke-points in moving data to/from the actual processor complex. <<<<<<<< I have noticed that various new networking mechanisms work around the second two criteria... - soup On Fri, Jan 15, 2010 at 11:24 AM, Chris Cox <[email protected]> wrote: > On Mon, 2010-01-11 at 13:00 -0500, Hall, Kenneth J wrote: >> I'd rather take a 4381 cabinet, gut it, and put a very small PC inside >> running Hercules. Dummy up the control panel and 3279 console terminal, and >> use the empty space for storage. >> > > If only IBM would allow ISV's to put something real on Hercules. > It's like IBM doesn't want anyone to learn stuff anymore... not > that I mind Linux of course. > > ---------------------------------------------------------------------- > For LINUX-390 subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO LINUX-390 or visit > http://www.marist.edu/htbin/wlvindex?LINUX-390 -- John R. Campbell Speaker to Machines souperb at gmail dot com MacOS X proved it was easier to make Unix user-friendly than to fix Windows ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
