| PLEX86 | ||
The System360 Model 20 Wasn't As Bad As All That 3860
I kind of doubt the 360-20 was even in the same league as an 8 MHz MC68000. IIRC the 8 MHz MC68000 had a 0.5 usec. memory cycle time accessing 16 bits at a time, the ALU was 16 bits wide, and the general purpose registers were stored in a semiconductor scratchpad memory. Not having a clue what sort of hardware was inside a 360-20, I would have to guess that the MC68000 had to be at least an order of magnitude faster than the 360-20 on integer problems, and probably more given the lackluster performance of the other low end 360s. I could imagine that on the bread and butter decimal jobs the 360-20 might have offered about the same performance as the MC68000 since it had micro code to do decimal arithmetic while the MC68000 didn't. Does anyone have any idea what sort of hardware was inside a 360-20? What did the data paths look like? What was the memory cycle time and access width? Are instruction timings available for the 360-20? There is a lot of this sort of hardware data easily available for the 360-30 and 360-40, but the guts of the 360-20 appear to be pretty much of a mystery, at least to me. The System360 Model 20 Wasn't As Bad As All That 3862 While this example does prove that long file names are supported, it illustrates another major "broken as... Regards, The System360 Model 20 Wasn't As Bad As All That 3861 On Thu, 06 Jul 2006 02:06:23 -0500, Charles Richmond My original defense of the Model 20, such as it was, was... John Byrns
|
||||
The System360 Model 20 Wasn't As Bad As All That 3861 Alt Folklore Computers from Newsgroups The #1 Usenet Provider on the Internet
|
||||