larryr@sdsu.UUCP (Larry Riedel) (08/23/88)
[Take this line... Please!] I am thinking about getting that 14MHz 68000 thing from CMI, but I am wondering if the extra clock speed will be wasted. I have one of those Micron 2MB boards, and I don't know how fast the memory is. Will the fast processor just be hurry-up-and-waiting? Larry ...!ucsd!sdsu!larryr
grr@cbmvax.UUCP (George Robbins) (08/24/88)
In article <3127@sdsu.UUCP> larryr@sdsu.UCSD.EDU (Larry Riedel) writes: > [Take this line... Please!] > > I am thinking about getting that 14MHz 68000 thing from CMI, but I am > wondering if the extra clock speed will be wasted. I have one of those > Micron 2MB boards, and I don't know how fast the memory is. Will the > fast processor just be hurry-up-and-waiting? Unless you have some local high-speed memory, most likely on the speed-up board, the extra clock speed will be mostly, but not quite completely wasted... Of course, some people get excited about a 5-10% speed up, so cases do differ. 8-) -- George Robbins - now working for, uucp: {uunet|ihnp4|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@uunet.uu.net Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
daveh@cbmvax.UUCP (Dave Haynie) (08/25/88)
in article <3127@sdsu.UUCP>, larryr@sdsu.UUCP (Larry Riedel) says: > Keywords: volatile > I am thinking about getting that 14MHz 68000 thing from CMI, but I am > wondering if the extra clock speed will be wasted. I have one of those > Micron 2MB boards, and I don't know how fast the memory is. Will the > fast processor just be hurry-up-and-waiting? All of the Amiga speedup boards run the Expansion Bus memory at it's normal speed. They do speed the system up, but they only run faster in the following cases: - Internal CPU operation - Cache hit - Access to special fast memory If you're using a Commodore-Amiga A2620 or a CSA or Hurricane board with their fast memory options, you get all three benefits. A plain fast 68000 with no on-board memory designed specifically for it will only get the first benefit. So you'll get some additional speed, but it's had to say just how much. With a decent amount of memory that'll keep up with that 68000 (it would go somewhere on the 68000's board), you might approach a factor of 2 speedup (you'll never really get there, since you have to access CHIP ram or I/O registers at least some of the time). > Larry ...!ucsd!sdsu!larryr -- Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy "I can't relax, 'cause I'm a Boinger!"
richard@gryphon.CTS.COM (Richard Sexton) (08/26/88)
In article <4553@cbmvax.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >In article <3127@sdsu.UUCP> larryr@sdsu.UCSD.EDU (Larry Riedel) writes: >> >> I am thinking about getting that 14MHz 68000 thing from CMI, but I am >> wondering if the extra clock speed will be wasted. I have one of those >> Micron 2MB boards, and I don't know how fast the memory is. Will the >> fast processor just be hurry-up-and-waiting? > >Unless you have some local high-speed memory, most likely on the speed-up >board, the extra clock speed will be mostly, but not quite completely >wasted... Of course, some people get excited about a 5-10% speed up, >so cases do differ. 8-) Yes. Anything the processor does completely internally will happen at 14 Mhz. As soon as it has to go to 8 Mhz vanilla memory the whole thing comes to a grinding wait (state). As far as I can see, the only real gain is in the area of math operations. I seem to recall that this is the board with the socket for a 68881. This is a big gain for floating point. A 68000 running at 14 Mhz still does math instructions at a full 14Mhz (one you get the operands INTO the 68000), and given that a 68010 does a multiply in 2/3 the cycles of a 68000, this board may be a gain to somebody doing a lot of math, assuming you can get a 14 Mhz 68010. Translation: it wont make your compiles go faster, but if you do a lot of number crunching it's probably well worth the money. -- How can you believe anything labeled: <8086@cup.portal.com> ? richard@gryphon.CTS.COM {backbone}!gryphon!richard
dale@boing.UUCP (Dale Luck) (08/26/88)
In article <4572@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >in article <3127@sdsu.UUCP>, larryr@sdsu.UUCP (Larry Riedel) says: > >> I am thinking about getting that 14MHz 68000 thing from CMI, but I am >> wondering if the extra clock speed will be wasted. > >All of the Amiga speedup boards run the Expansion Bus memory at it's >normal speed. They do speed the system up, but they only run faster >in the following cases: > > - Internal CPU operation >-- >Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" Floating point routines will benefit from this. In terms of cpu cycles, these routines have the highest fraction of cpu cycles/memory access. In V1.3 ieee libraries this is even more apparent. The 68000 spends more time in the Mul and Div routines relative to the rest of the code then the older V1.2 code. For example a 68010 running at same speed will improve Mul/Div by up to 5% from just having a little smarter Mul and Div microcode. -- Dale Luck Boing, Inc. {cbmvax|oliveb|pyramid}|!amiga!boing!dale Although I do contract work for Amiga-LosGatos, my opinions probably don't represent those of Commodore or its management or its engineers, but I think the world would be a better place if they did.
sneakers@heimat.UUCP (Dan "Sneakers" Schein) (08/26/88)
[ Tommorow is international line eater day. Don't forget to feed yours! ] In Message <4572@cbmvax.UUCP>, daveh@cbmvax.UUCP (Dave Haynie) writes: >in article <3127@sdsu.UUCP>, larryr@sdsu.UUCP (Larry Riedel) says: >> I am thinking about getting that 14MHz 68000 thing from CMI, but I am >> wondering if the extra clock speed will be wasted. I have one of those >> Micron 2MB boards, and I don't know how fast the memory is. Will the >> fast processor just be hurry-up-and-waiting? > >All of the Amiga speedup boards run the Expansion Bus memory at it's >normal speed. They do speed the system up, but they only run faster >in the following cases: > > - Internal CPU operation > - Cache hit > - Access to special fast memory > >If you're using a Commodore-Amiga A2620 or a CSA or Hurricane board with >their fast memory options, you get all three benefits. A plain fast >68000 with no on-board memory designed specifically for it will only >get the first benefit. So you'll get some additional speed, but it's >had to say just how much. With a decent amount of memory that'll keep >up with that 68000 (it would go somewhere on the 68000's board), you >might approach a factor of 2 speedup (you'll never really get there, >since you have to access CHIP ram or I/O registers at least some of >the time). > Hmmmm this sounds like you should be able to obtain very similar results by just upgrading to a MC68010. After all most internal commands of the 010 are roughly 1 to 2 times faster than a 68000 and apperars to be the only benefit of this product. Hazey any comments? Heres a thought - how about using one of these speed boards(?) with a 68010 in place of the 68000. Hmmmm..... now where did I put that 010 :-) >> Larry ...!ucsd!sdsu!larryr >Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" -- Dan "Sneakers" Schein {alegra|amiga|rutgers|uunet}!cbmvax!heimat!sneakers Sneakers Computing 2455 McKinley Ave West Lawn PA 19609 Call: BERKS AMIGA BBS 24 Hrs - 3/12/2400 Baud Disclaimer: Any opinions expressed are 40 Meg -=- 215/678-7691 those of Sneakers Computing Of course heimat is an Amiga, doesn't everyone run UUCP & UseNet on an Amiga?
perley@caesar.steinmetz (Donald P Perley) (08/26/88)
In article <4572@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >in article <3127@sdsu.UUCP>, larryr@sdsu.UUCP (Larry Riedel) says: >> I am thinking about getting that 14MHz 68000 thing from CMI, but I am >> wondering if the extra clock speed will be wasted. >All of the Amiga speedup boards run the Expansion Bus memory at it's >normal speed. They do speed the system up, but they only run faster >in the following cases: > > - Internal CPU operation > - Cache hit > - Access to special fast memory > >. A plain fast >68000 with no on-board memory designed specifically for it will only >get the first benefit. I don't know about the CMI board, but some 68000 speedup boards have floating point processor sockets. If you use it, and are running programs with a lot of FP computation that support it, you should get a dramatic speed increase, yes/no? If you do a lot of floating point stuff (and a lot of graphics programs do), it sounds like a lot more bang for the buck compared to a 68020 board with coprocessor at ~$1000 more (plus 32 bit memory to get a real advantage over the 68000). -don perley
collins@pnet02.cts.com (Steven Collins) (08/29/88)
In principle it is possible to put a 68010 in one of the 14 Mhz boards, but no one (to my knowledge at least) makes a 14 Mhz 68010. The folks at CMI said that one might be able to find a 12 Mhz 68010 that would run at 14 Mhz, by trying out a few of them, but an 8Mhz (the cheap ones) would have little chance of working. I had an 8 Mhz 68010 in my A1000 which I tried in my CMI board and it didn't run. The CMI board is somewhat faster than the 68010 however. I feel that for the sort of stuff I normally do, I got about a 10% speedup from a 68000 using the 68010 and about that much more using the CMI over the 68010. I also dropped the bux for a 68881 but I still don't have the 1.3 IEEE librarys that are required to use the FPU. The CMI board uses the 68881 as a peripheral device( this is supported under 1.3...) but you can't just compile stuff under the MANX 68881 option and get it to run because MANX assumes the 68881 is actually hung on the bus along with a 68020. This latter method of using the 68881 is faster, but somewhat more expensive... steve collins (at least that's how I heard it....) When you're as dumb as me, you never even feel the flames... UUCP: {ames!elroy, <backbone>}!gryphon!pnet02!collins INET: collins@pnet02.cts.com
daveh@cbmvax.UUCP (Dave Haynie) (08/31/88)
in article <11988@steinmetz.ge.com>, perley@caesar.steinmetz (Donald P Perley) says: > I don't know about the CMI board, but some 68000 speedup boards have > floating point processor sockets. If you use it, and are running > programs with a lot of FP computation that support it, you should > get a dramatic speed increase, yes/no? You'll get a speedup over no 68881 at all (assuming 1.3 math libraries). What you don't get is a 32 bit interface to the 68881, or a true coprocessor interface to the 68881, so this is still considerably slower than a similarly clocked 68020. Given the speedups you can get with any kind of math chip, it's probably a good idea for someone into image rendering or other heavy duty math operations who doesn't have a 68020 type budget. > If you do a lot of floating point stuff (and a lot of graphics programs do), > it sounds like a lot more bang for the buck compared to a 68020 board > with coprocessor at ~$1000 more (plus 32 bit memory to get a real advantage > over the 68000). Note that for math operations, a 68020 with 68881 is several times faster than a 68000 with 68881, even on a 16 bit bus. The difference is of course the CPU interface that the 68020 uses, plus the wide data bus. The 68000 or 68020 have to fetch data for the 68881 in either case. But while the 68000 will fetch at 16 bits, then transfer at 16 bits, the 68020 will do it's transfers at 32 bits. Of course, if you can only afford a fast 68000 board at $200-$300, it's going to speed up your math more than that cash sitting in the bank waiting for a 68020 board will... > -don perley -- Dave Haynie "The 32 Bit Guy" Commodore-Amiga "The Crew That Never Rests" {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy "I can't relax, 'cause I'm a Boinger!"
mp1u+@andrew.cmu.edu (Michael Portuesi) (09/10/88)
grr@cbmvax.UUCP (George Robbins) writes: > Unless you have some local high-speed memory, most likely on the speed-up > board, the extra clock speed will be mostly, but not quite completely > wasted... Of course, some people get excited about a 5-10% speed up, > so cases do differ. 8-) Perhaps someone who has one of these CMI speed-up boards would like to run the standard set of benchmarks (dhrystones, whetstones, khornerstones, flyntstones :-) on an accelerated and a stock Amiga and post the results -- especially results that depend on the optional 68881. Also, some tests with real programs, such as time to load/scroll in WordPerfect, recalc a Maxiplan sheet, or perform a complex fill with DeluxePaint would be appreciated too. --M Michael Portuesi / Information Technology Center / Carnegie Mellon University ARPA/UUCP: mp1u+@andrew.cmu.edu BITNET: rainwalker@drycas "if you ain't ill it'll fix your car"
billc@percival.UUCP (William Coldwell) (09/13/88)
In article <0X-=awF28k-0QninBy@andrew.cmu.edu> mp1u+@andrew.cmu.edu (Michael Portuesi) writes: >grr@cbmvax.UUCP (George Robbins) writes: >> Unless you have some local high-speed memory, most likely on the speed-up >> board, the extra clock speed will be mostly, but not quite completely >> wasted... Of course, some people get excited about a 5-10% speed up, >> so cases do differ. 8-) > >Perhaps someone who has one of these CMI speed-up boards would like >to run the standard set of benchmarks (dhrystones, whetstones, >khornerstones, flyntstones :-) on an accelerated and a stock Amiga >and post the results -- especially results that depend on the >optional 68881. Well, I suppose that I could do something like that. I did write an article for the local Amiga Users Group (NAG), and it's in the NAGRAG. It was also distributed at AmiExpo. Though things have changed, most of the timings are still fairly accurate. >Also, some tests with real programs, such as time to load/scroll in >WordPerfect, recalc a Maxiplan sheet, or perform a complex fill with >DeluxePaint would be appreciated too. > >Michael Portuesi / Information Technology Center / Carnegie Mellon University Now, let's keep our heads screwed on straight here... timings can be - well - uh - biased. Compilers also rake havok on trying to get accurate results. Another thing, is that if you set a program up to use IEEE and show off the 68881, then without it the program will run like a dead horse (or my ex-wife...).. I suppose that I can upload a semi-reprint of the article and let you see what I came up with (also remember that the 1000 will generally show a difference in speed on things with interlace screens, like (gasp!) Microbotics' FFUZ program. Bill -- William J. Coldwell - Amiga Attitude Adjuster {aka Software Developer} (503) 684-9300 ...tektronix!reed!percival!billc "Blame the hardware." [I break for 14MHz Processor Accelerators with 68881's.] Creative Microsystems Inc. / 10110 SW Nimbus Suite B1 / Portland, OR 97223