Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
08/03/07 16:52
Read: times


 
#142717 - I think it meets the spec ...
Responding to: ???'s previous message
The 8x300 used a Harvard Architecture with simple load and store instructions. The only "gotcha", IIRC, and it has been 30 years or so, was that each instruction, while it took only 300 ns to execute each of its 8 instructions, was a load and store, with logic operations and shifts in nearly every instruction. Each instruction had a data source and data destination outside the CPU. The version that I used had already been sped-up to 250 ns per cycle.

It is, in fact, a "Cambrian" DSP, perhaps "pre-Cambrian," but it was the first common RISC that I ever encountered. The technology was, indeed, bipolar, but not TTL, as it was current-mode-logic, like ECL, but TTL-compatible, intgrated injection logic (IIL) requiring an external current pump.

RE


List of 13 messages in thread
TopicAuthorDate
Instruction Set in RISC Microcontroller            01/01/70 00:00      
   the name is misleading...            01/01/70 00:00      
      RISC actually means ...            01/01/70 00:00      
         possible, rather than necessary            01/01/70 00:00      
   Reduced complexity            01/01/70 00:00      
   if you want to see a real RISC ...            01/01/70 00:00      
      RISCness            01/01/70 00:00      
      I wouldn't call this Signetics thing a RISC...            01/01/70 00:00      
         The original RISC            01/01/70 00:00      
            Only 8 Codes!            01/01/70 00:00      
            If I were you, I'd consult one of the PDP8 sites            01/01/70 00:00      
         I think it meets the spec ...            01/01/70 00:00      
         about that website bug ...            01/01/70 00:00      

Back to Subject List