Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.arch    |    Apparently more than just beeps & boops    |    131,241 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 129,334 of 131,241    |
|    Peter Flass to Terje Mathisen    |
|    Re: 64 bits, was VAX    |
|    07 Aug 25 07:34:28    |
      XPost: alt.folklore.computers       From: Peter@Iron-Spring.com              On 8/7/25 06:44, Terje Mathisen wrote:       >       > Bit addressing, presumably combined with an easy way to mask the       > results/pick an arbitrary number of bits less or equal to register       > width, makes it easier to impement compression/decompression/codecs.       >       > However, since the only thing needed to do the same on current CPUs is a       > single shift after an aligned load, this feature costs far too much in       > reduced address space compared to what you gain.       >              Bit addressing *as an option* (Bit Load, Bit store instructions, etc.)       is a great idea, for example it greatly simplifies BitBlt logic. The       432's use of bit addressing for everything, especially instructions,       seems just too cute. I forget the details, it's been a while since I       looked, but it forced extremely small code segments which, combined with       the segmentation logic, etc. really impacted performance.              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca