Proprietary versus Open Instruction Sets

......An instruction set architecture (ISA) is one of the most important interfaces in a computer system because it divides software from hardware. Most widely used ISAs were developed decades ago and are proprietary. This may make sense because hardware implementations were and are mostly proprietary and most software was proprietary when these ISAs were developed. Today, however, we appreciate how open source software (such as LAMP stack) and open standards (such as TCP/IP) can unleash competition and creativity. Might open ISAs accelerate the innovation of computer hardware, or are the lessons of open software not pertinent? That was the subject of the debate Mark D. Hill moderated at the 4th Workshop on Computer Architecture Research Directions in June 2015. Normally, a moderator is conflict free. In this case, Hill is conflict full, because David Patterson coadvised Hill’s PhD with Alan Smith, and Hill currently consults for Advanced Micro Devices, where Dave Christie has worked for decades. The first panelist was Dave Christie, a senior fellow at AMD. He first worked for the Control Data Corp. before moving to AMD in the late 1980s. He spent two decades contributing to virtually all of AMD’s x86 processors, including the K5, Athlon, and Opteron. He developed the microcode for the K5, the first superscalar x86 processor designed independently from Intel. He is also known for having co-designed the x86-64 instruction set (the 64-bit extensions to the x86 instruction set). Currently, he is serving as AMD’s ARM architecture liaison. He argued in favor of proprietary ISAs. The second panelist was David Patterson, who holds the E.H. and M.E. Pardee Chair of Computer Science at the University of California, Berkeley. Of all his accomplishments and honors, most pertinent to this debate was that he led the design and implementation of RISC I, an early VLSI reduced-instruction-set computer. This research served as the foundation of the Sparc architecture. He has consulted for 25 years at various microprocessor companies, including spending a sabbatical at Digital Equipment Corp. in 1979 working on the VAX minicomputer and consulting for Intel for a few years on microprocessors, and he was the first person hired by Sun Microsystems to develop the Sparc architecture. He is a member of both the National Academy of Engineering and the National Academy of Sciences, and a Fellow of both IEEE and Mark D. Hill