On 2017-10-16, at 10:23:50, David Cole wrote: > > But all these techniques for efficiency that IBM has created are not human > compatible. They are too complex, they are too messy[!] and they are not even > the same techniques from one machine to the next. In fact, sometimes code > written to be efficient on one machine will be actually inefficient on > another! > Long ago, a co-worker studied instruction timings on the 370/148. He discovered that for certain operands LM; STM outperformed MVC and hand-optimized his Assembler code accordingly.
Then we got our own Magnuson M80s, advertised as 148 performance or better. In fact, LM; STM was faster than on the 148. But Magnuson had optimized MVC so it was faster still. True Blue, he was furious at Magnuson for invalidating his painstaking optimization by providing a better alternative. -- gil