On 10/25/18 11:40 AM, Paul Koning via cctalk wrote:
>
>
>> On Oct 25, 2018, at 1:45 PM, Chuck Guzis via cctalk <cctalk@classiccmp.org>
>> wrote:
>>
>> On this subject, is there no interest in serial ALU designs? At one
>> time, if you wanted a low-cast implementation, that was the way to do
>> it. Also gives you a leg up on variable word-length designs.
>>
>> Didn't at least one of the more popular MPU designs employ a serial ALU?
>> TMS9900?
>
> Serial ALU as a cost saver for low end makes sense in discrete transistor or
> early SSI machines, like the PDP-8/L.
PDP-8/S
Saul Dinman's machine, who founded GRI
https://books.google.com/books?id=1jTUBwAAQBAJ&pg=PA52&lpg=PA52&dq=gri+computer+saul&source=bl&ots=Qqd_woK8KO&sig=Ii8W_8Stxwuo2kHY4uJU6DJXYzc&hl=en&sa=X&ved=2ahUKEwiF0NeLoKLeAhWTCjQIHbIFB84Q6AEwAHoECAYQAQ#v=onepage&q=gri%20computer%20saul&f=false
Arrays of single-bit processors have been tried.
CHM has one, I'm blanking on the company name.
We had one in Apple ATG.
