jwhitman@ALMSA-1.ARPA (Jerry Whitman) (12/18/86)
There is a mail-order outfit in Texas advertising the Amiga a 68010 CPU, 1080 monitor, clock w/battery,and (the attention getter) 1.5 Meg of memory *INTERNAL* for $1595. This *sounds* like a goodly deal, but....I have questions! a. I have seen past articles discussing the 68010 in the AMIGA and they left me with these questions: 1 - Is the 68010 fully upward compatible with the 68000. If not does it negatively affect existing software products? 2 - Can/will the existing C/A software (V1.2 OS, Basic, utilites etc) take advantage of 68010 features? 3 - Can/will existing commercial software take advantage of 68010 features? b. I have been laboring under the impression that the AMIGA could only see 500K of internal RAM. This appears not to be so, but... 1 - Can the special function chips (graphics, audio, etc) see and utilize any of the additional 1Meg of internal memory? 2 - Can/will existing software (C/A or commercial) use it? If so how does any software know it is available? (Maybe Autoconfig?) 3 - Can it be used for RAM:? c. Doesn't this modification void the C?A warranty? These are issues that come to mind. If you have the answers to any of them or recognize other issues please let me know. Thanks for listening. Jerry Whitman JWHITMAN@ALMSA-1.ARPA
grr@cbmvax.cbm.UUCP (George Robbins) (12/20/86)
In article <877@ulowell.UUCP> jwhitman@ALMSA-1.ARPA (Jerry Whitman) writes: >There is a mail-order outfit in Texas advertising the Amiga a 68010 >CPU, 1080 monitor, clock w/battery,and (the attention getter) 1.5 Meg >of memory *INTERNAL* for $1595. This *sounds* like a goodly deal, >but....I have questions! >a. I have seen past articles discussing the 68010 in the AMIGA and >they left me with these questions: > 1 - Is the 68010 fully upward compatible with the 68000. If not > does it negatively affect existing software products? > 2 - Can/will the existing C/A software (V1.2 OS, Basic, utilites > etc) take advantage of 68010 features? > 3 - Can/will existing commercial software take advantage of 68010 > features? A) The 68010 is not completly compatible with the 68000, but a little help from the trap routines in the OS can smooth things over. Unfortunatly, the added features don't help much in an Amiga - it runs a few percent faster, but that's it. Most of the features that were added to the 68010 were to make it easy to do virtual memory systems with a some kind of mmu. >b. I have been laboring under the impression that the AMIGA could only >see 500K of internal RAM. This appears not to be so, but... > 1 - Can the special function chips (graphics, audio, etc) see and > utilize any of the additional 1Meg of internal memory? > 2 - Can/will existing software (C/A or commercial) use it? If so > how does any software know it is available? (Maybe Autoconfig?) > 3 - Can it be used for RAM:? B) There are various places in the memory map where you can stick RAM and have it automatically detected, or add it to the available memory list later. These places may/will change in future machines and may require internal decoding to work reliably, so we make no big fuss about it. The current generation of custom chips can only access the first 512K of internal RAM. Various paging schemes are conceivable, but without support from AmigaDOS and the graphics libraries would be of limited utility. Most software can use any RAM that the system has been told about. I don't see why it couldn't be used for RAM: although I understand that the current C/A RAM: device wouldn't limit itself to it. >c. Doesn't this modification void the C?A warranty? Unless they have received some kind of special authorization from C/A then these modifications would void the warranty. It would then be up to this Texas company to offer their own warranty. -- George Robbins - now working for, uucp: {ihnp4|seismo|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@seismo.css.GOV Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
wtm@neoucom.UUCP (Bill Mayhew) (12/21/86)
Hi, Yes, the 68010 will work with Workbnech 1.1 and 1.2 software 95% or greater of the time. It seems like most of the commercial writers have avoided things to work differently a-la 68010. Don't expect amazing things to happen when you throw in (or buy it with it already thrown .. er.. plugged in) a 68010. In my case, I saw about a 10 to 15% gain in speed (yawn). Of course, that is what I expected. Alas, there is one thing that I know for sure does break when you have a 68010-- Transformer to name names. We all know that transformer is s-l-o-w, but I do have times when it is useful since I don't have a pee-cee at home, so I ditched my 68010 until the (hopefully affordable) sidecar shows up around here. As far as the 1.5 meg internal memory goes, it may or may not be useful. I'll bet that it doesn't follow the "Zoro" standard, and thus will require a special driver in order to be recognized by the operating system. Make sure you ask them if the memory can autoconfigure, and if it can't if they provide the necessary software to make it useful to the operating system. That aside, the price you mention seems at least reasonable. --Bill Bill Mayhew Division of Basic Medical Sciences Northeastern Ohio Universities' College of Medicine Rootstown, OH 44272 USA phone: 216-325-2511 (wtm@neoucom.UUCP ...!cbatt!neoucom!wtm) m As far as tgh
wagner@utcs.UUCP (12/23/86)
In article <1154@cbmvax.cbmvax.cbm.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >A) The 68010 is not completly compatible with the 68000, but a little >help from the trap routines in the OS can smooth things over. >Unfortunatly, the added features don't help much in an Amiga - it runs >a few percent faster, but that's it. Most of the features that were >added to the 68010 were to make it easy to do virtual memory systems >with a some kind of mmu. I thought the speedup was on the order of 10%. Did I mis-remember? Two features in the 68010 make for a faster chip. Loop mode, and a general stream-lining of the multiply and divide algorithms. According to the upgrate notes on fish disk 18, one can expect a speedup of from 4-50% (8-50% claimed in Motorola Documentation). Multiply/divide performance can be expected to be 10% better (without recoding). Loop mode performance (needs recoding, I expect) can be from 20-80% better. Loop mode, for those who haven't dived into the hardware manuals yet, is this (amazing) scheme for keeping a few instructions in an on-chip 'cache'. When the loop runs 'in' the cache, the only storage fetches are for operands. Tight loops such as looking for blanks could potentially use such a feature and speed up considerably. So, question: Did anyone go to the trouble of trying to exploit loop mode in the Amiga EXEC? Does BCPL try to exploit it? Does the Lattice C code generator? How about the Lattice library? How about MANX? In any case, if the cost of a 68010 was low enough, it might be worth it for the 4-8% worst case speedup. After all, it does buy back some of the 17-20% of the CPU that seems to be leaking away all the time. And who knows, maybe the little loops that it's running when it's leaking are the sort that can be speeded up by loop mode! :-) Michael