[HN Gopher] 128-bit RISC-V assembler
       ___________________________________________________________________
        
       128-bit RISC-V assembler
        
       Author : ingve
       Score  : 23 points
       Date   : 2021-09-19 11:43 UTC (11 hours ago)
        
 (HTM) web link (github.com)
 (TXT) w3m dump (github.com)
        
       | b0tzzzzzzman wrote:
       | Looking for insight from people who work at low level optimising
       | of code, instruction sets, and hardware.
       | 
       | 1. Do you see 128bit processors/systems becoming mainstream
       | anytime in the near future?
       | 
       | 2. What type of applications would benefit consumers and general
       | computing at scale?
       | 
       | 3. Do we know of any of the main players working on these?
        
       | mrlonglong wrote:
       | Are there any 128-bit core done yet in FPGA?
        
       | baybal2 wrote:
       | Who needs 2^128 integers?
        
         | sedatk wrote:
         | Evergreen comment, "2^128" being an arbitrary upper limit.
        
         | hdjjhhvvhga wrote:
         | It opens up some very interesting possibilities and
         | optimization options that you wouldn't normally ever consider.
         | In any case, I think programmers of most languages wouldn't
         | even need to care as all optimizations would be done by the
         | teams working on the compilers/interpreters/VMs.
         | 
         | Note that it's not just about integers, floating point would
         | also benefit from this.
        
         | mrlonglong wrote:
         | You can use tagged pointers for memory protection.
        
       | ruslan wrote:
       | SCALL instruction, used in his example, has been renamed to
       | ECALL. Not a big deal though.
        
         | fwsgonzo wrote:
         | Good catch. I will rename it to ecall and provide an alias for
         | syscall instead.
        
       | nynx wrote:
       | When you need a pointer that can address any byte that has or
       | will ever exist.
        
       | phkahler wrote:
       | And here I was thinking 48 bits would be optimal for a lot of
       | things. From address space to posits and instruction lengths
       | (actually 24 is probably better there).
        
         | ruslan wrote:
         | Although I completely agree with you, I cannot help but
         | remember "640K ought to be enough for anyone" by W.H. Gates.
        
       ___________________________________________________________________
       (page generated 2021-09-19 23:02 UTC)