Is it time to stop research in Computer Architecture ?
|
From: nmm1 on 21 Oct 2009 04:17 In article <obednRQFVp4oBkPXnZ2dnUVZ8tpi4p2d(a)lyse.net>, Terje Mathisen <Terje.Mathisen(a)tmsw.no> wrote: >EricP wrote: >> Robert Myers wrote: >>> People in the sixties had many wild and dangerous ideas that have been >>> proven to be dead ends, at best. Many of the ideas about computers >>> are in a category with Werner von Braun's rotating donuts, which would >>> have been unstable. >> >> The rotating donut space stations are unstable? >> Why? > >At least if they get big enough, see the various articles about Niven's >Ringworld, a donut space station big enough to surround a star at >Earthlike (150e9 m) distance. There's a solution to that which I thought of when I heard about the debate - quite trivial - just tether the ring to a Klemperer rosette. But that's a different problem. >I don't believe a regular-size one without an untethered mass in the >center would suffer the same problem. No, Robert is right. The reason is that the moment of inertia is largest in the symmetric plane, and disturbed rotating objects tend to change to rotating about the axis with the smallest one. Regards, Nick Maclaren.
From: ChrisQ on 21 Oct 2009 05:41 nmm1(a)cam.ac.uk wrote: > > No, Robert is right. The reason is that the moment of inertia is > largest in the symmetric plane, and disturbed rotating objects > tend to change to rotating about the axis with the smallest one. > I think you'll find google gyroscopic precession describes the effect... But anyway, getting back on topic, how about von neumann is dead ?, and what could replace it ?. Regards, Chris
From: Noob on 21 Oct 2009 06:04 Andrew Reilly wrote: > Isn't it the case, though, that for most of that "popular > software" speed is a non-issue? Either a given operation is "fast > enough" (and that's not hard to achieve, when the limitations are those > of hands, eyes and brain of the user), or "not fixable by software": > network and server latencies, disk access latencies, and so on. NB: Replacing hard-disk drives with solid-state drives improves random read latency by two orders of magnitude, which might push this particular bottle-neck to a different sub-system. Regards.
From: Bill Todd on 21 Oct 2009 06:08 nmm1(a)cam.ac.uk wrote: > In article <D7idnZckOPiTI0DXnZ2dnUVZ_vSdnZ2d(a)metrocastcablevision.com>, > Bill Todd <billtodd(a)metrocast.net> wrote: >> Terje Mathisen wrote: >> >>> OTOH, I really do believe Intel intended to start deliver in 1997, in >>> which case it _would_ have been, by far, the fastest cpu on the planet. >> I don't care enough at this late date to do more than question the above >> assertion from what I can remember off the top of my head without >> attempting to quantify my reservations, but remember that the product >> that they allegedly expected to deliver in 1997 (or perhaps 1998, >> depending on how you read their early claims) was Merced, not McKinley. >> Do you really think that Merced in a then-current process "_would_ >> have been, by far, the fastest cpu on the planet" - especially for >> general (rather than heavily FP-specific) code? ... > > Oh, yes, indeed, it would have been - if they had delivered in 1997 > what they were promising in 1995-6. That's irrelevant: the question was how an actual (shipped-in-mid-2001) Merced would have performed if delivered in 1997 (possibly 1998) using a then-current process, not how some vaporware being talked about earlier might have performed. One might get a clue from how Merced actually did perform in 1999 when the first samples went out: not nearly well enough to satisfy anyone, which was a large part of why it didn't ship for another two years. .... > If they had delivered just the hardware, it would have been by far > the fastest cpu on the planet for suitable codes (not necessarily > floating-point, but definitely only a small subset). Once again that's irrelevant to the question under discussion here: whether Terje's statement that Merced "_would_ have been, by far, the fastest cpu on the planet" (i.e., in some general sense rather than for a small cherry-picked volume of manually-optimized code) stands up under any real scrutiny. - bill
From: nmm1 on 21 Oct 2009 06:20
In article <kKudnelkCfoGQEPXnZ2dnUVZ_t2dnZ2d(a)metrocastcablevision.com>, Bill Todd <billtodd(a)metrocast.net> wrote: > >>> Do you really think that Merced in a then-current process "_would_ >>> have been, by far, the fastest cpu on the planet" - especially for >>> general (rather than heavily FP-specific) code? ... >> >> Oh, yes, indeed, it would have been - if they had delivered in 1997 >> what they were promising in 1995-6. > >That's irrelevant: the question was how an actual (shipped-in-mid-2001) >Merced would have performed if delivered in 1997 (possibly 1998) using a >then-current process, not how some vaporware being talked about earlier >might have performed. You may have gathered that I agree with you from the rest of the paragraph :-) But the insoluble problems were software and not hardware - they could, at least if they had handled the project excellently, have delivered what they claimed in terms of hardware. As it was, they didn't handle it well, and wasted a lot of effort trying to solve the software problems by adding massive caches and other hardware tweaks. Regards, Nick Maclaren. |