Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!ogicse!unmvax!uokmax!drtiller From: drtiller@uokmax.ecn.uoknor.edu (Donald Richard Tillery Jr) Newsgroups: comp.sys.amiga.graphics Subject: Re: ColorBurst NTSC @ LA VideoExpo Message-ID: <1991May31.041515.14091@uokmax.ecn.uoknor.edu> Date: 31 May 91 04:15:15 GMT References: <91145.032250ACPS1072@Ryerson.CA> Organization: Engineering Computer Network, University of Oklahoma, Norman, OK Lines: 60 In a message From: mwm@pa.dec.com (Mike (My Watch Has Windows) Meyer) >I've seen the '040 quoted at anywhere between 12 and 20 MIPS. The Motorola literature indicates the 68040 @ 25Mhz runs 21 MIPS. >previous generation of Sun chips were quoted at 14 MIPS. The current >generation is quoted at 25 MIPS(*). Ergo, the '040 isn't faster than >Sun's processor, at least based on the numbers I've seen. > > and that a GVP 3050 board is faster than an '040 > >What clock speed '040? And running what application? And in the what >system? There's no way a GVP 3050 will be faster than a (hypothetical) >50MHz '040 on a daughterboard on the 3050 card. It's unlikely that the >3050 is faster than a 25MHz or faster 040 in the same situation. On >the other hand, the 3050 should blow the doors off an '040 card (any >speed) in a 2000 CPU slot with no 32 bit memory. > > so a 3050 equipped Amiga is faster than a Sun. WRONG! Even if you have a linear speedup from 25 to 50 MHZ (which you don't) the 50 MHz would only run at twice my 25 (now 30)Mhz '030 which achieves 7.5 MIPS - that's 15 MIPS and that is less than 75% of the 21 MIPS of the '040. Adding in the 10X speedup of the FPU (2X for transcendental functions) and the 68040 "should blow the doors off an" '030. > >Well, it's certainly faster than the last one I used (a 3/50). Then >again, so is my Amiga 3000. Of course, on the right applications, my >Amiga 3000 is faster than the 3050. On the other hand, I'd be Under WHAT applications is your 3000 faster than a 3050? Even if everything happens in chip RAM, the extra 16 bits doesn't make enough difference to out run a 50 MHz '030. I'd like to see one that did! Please upload such an application for me to test myself. >surprised if any of the above was faster than a SparcStation 2. > > I didn't read this in AmigaWhirled. > >I'll agree with that - while they may publish fluff, they usually get >their fact straight. > By the way, the comment in Amiga World (and elsewhere) was misquoted - as usual! What GVP said was that at present their 50 MHz 68030 out runs THEIR ALPHA 68040 card (which runs about 14 MIPS right now). They aren't satisfied with that speed, and they are working on it. But that does NOT mean that the eventual production version won't hit the 20MIPS mark (which it should). What I'm wondering is whether I can push the 68040 like I have my 68030 - right now I've stretched my 25 MHz 68030 to 30 MHz. 20% is a rather large push, but it seems to work fine so far. Rick Tillery (drtiller@uokmax.ecn.uoknor.edu)