Loading...
Please wait, while we are loading the content...
Similar Documents
A panel session-Computer structure-Past , present and future Possibilities for Computer Structures 1971
| Content Provider | Semantic Scholar |
|---|---|
| Author | Gordon, Cameron Newell, Allen |
| Abstract | What computer structures come into existence in a given epoch depends on the confluence of several factors: The underlying technology-its speed, cost, reliability , etc. The structures that have actually been conceived. The demand for computer systems (in terms of both economics and user influence). One ignores any of these factors a t one's peril. In particular, with technology moving rapidly, a real limitation exists on our ability as designers to discover appropriate structures that exploit the new trade-offs between the various aspects of a computer system. The design of computer structures is not a systematic art. So new is it, in fact, that in a recent book (Bell and Kewell, 1971) we found ourselves dealing with basic issues of notation. We are still a long may from Concern nyith the sort of synthesis procedures that characterize, say, linear circuit design. However, the immaturity is dictated, not so much by youth (after all we have been designing computers for almost 30 Years), as by the shifts in technology that continually ' The idem expressed in this presentation have emerged from a umber of overlapping design efforts, mostly around CMU and the ARPA list ~rocessing machine effort, and the effort at the Stanford A1 ~roject). Consistent with this being a short note, fc have attempted to indicate the individuals involved in these efforts at appropriate places in the text. But we wish here to '~knowledge more generally the contribution of all these individuals. throw us into previously uninhabited parts of the space of all computer structures. Whatever systematic techniques start to emerge are left behind. This note comments on several possibilities for computer structures in the next halfdecade. Given the un-familiarity that we all have with the region of computer space into which we are now moving, there can be no systematic coverage. Neither is it appropriate simply to reiterate what would be nice to have. Such an esercise is not responsive to the new constraints that will limit the new designs. Such constraints will certainly continue to exist, no matter how rapidly logic speed rises and logic costs fall. In fact, it is useful to view any prognostication of new computer structures (such as this paper) as an attempt to reveal the nature of the design constraints that will characterize a new epoch of technology. We d l discuss five aspects of computer structures. Mostly, these represent design features that … |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://research.microsoft.com/en-us/um/people/gbell/CGB%20Files/Possibilities%20for%20Computer%20Structures%20ComConference%201971%20ng%20c.pdf |
| Alternate Webpage(s) | http://research.microsoft.com/~GBell/CGB%20Files/Possibilities%20for%20Computer%20Structures%20ComConference%201971%20ng%20c.pdf |
| Alternate Webpage(s) | http://www.research.microsoft.com/en-us/um/people/gbell/CGB%20Files/Possibilities%20for%20Computer%20Structures%20ComConference%201971%20ng%20c.pdf |
| Alternate Webpage(s) | http://gordonbell.azurewebsites.net/CGB%20Files/Possibilities%20for%20Computer%20Structures%20ComConference%201971%20ng%20c.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
