Wikipedia is not a source?

But your must ludicrous claim is that the fields controlling the number of data 
added are not relevant. The effect of 11 - 25 depends very much on those 
fields. I'm done here: I'm not going to spend any more time trying to cure your 
willful ignorance.


--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3

________________________________________
From: IBM Mainframe Assembler List [ASSEMBLER-LIST@LISTSERV.UGA.EDU] on behalf 
of Robin Vowels [robi...@dodo.com.au]
Sent: Sunday, June 19, 2022 3:44 PM
To: ASSEMBLER-LIST@LISTSERV.UGA.EDU
Subject: Re: When did logical instructions appear?

On 2022-06-20 04:25, Seymour J Metz wrote:
> I see that you're citing sources that you haven't read;

Now you are being absurd.
I cited no sources.  I mentioned three computers.

> what a surprise.

> [[Pilot ACE]] does not mention array instructions, but does
> note that the hardware only did fixed point and they had to write
> software to do it.

Rubbish! The hardware did exactly what I said, namely, I gave
two instructions, one to add 32 values to the accumulator
and another to subtract 32 values from the accumulator.

> Neither that article nor you mentioned the "Timing
> Number",

It wasn't relevant to the examples.

> nor did you seem aware of the requisite values of C, W and T.

I am very aware of those and all the other components of the
Pilot ACE and DEUCE instructions.

> What that entire instruction gives you is a complicated form of the
> REPEATE instruction found on some later machines,

Obviously you have learned nothing.  You write nonsense.

A Pilot ACE and DEUCE instruction specifying a transfer of from
2 to 32 numbers is not a "repeat" instruction. The instruction
is not re-executed for each subsequent word.  The instruction
is not re-entered into Control.

Executing such an instruction causes the hardware to continue the
transfer for a specified number of minor cycles.  The number
of minor cycles is the difference between the wait and timing numbers,
plus 1.
The transfer commences when the wait number counts down to zero, and
terminates
when the timing number counts down to zero.

For the record, there are some 32 × 29 instructions (i.e. 928)
instructions
that can operate on arrays.  I gave examples of only two of them.

> not "an instruction
> set where its instructions are designed to operate efficiently and
> effectively on large one-dimensional arrays of data called vectors".

Oh?  A group of 32 numbers is not a vector?
Go back to school.

The instructions on Pilot ACE and DEUCE were designed to operate
efficiently and effectively on arrays.  They could be vectors
and they could be matrices.

And for the record, at the time, Pilot ACE was the fastest machine
in the world.  Addition time was 32 + 32n microseconds (n = number of
words).
Not efficient?

> ________________________________________
> From: IBM Mainframe Assembler List [ASSEMBLER-LIST@LISTSERV.UGA.EDU]
> on behalf of Robin Vowels [robi...@dodo.com.au]
> Sent: Saturday, June 18, 2022 11:05 PM
> Subject: Re: When did logical instructions appear?
>
> From: "Seymour J Metz" <sme...@gmu.edu>
> To: <ASSEMBLER-LIST@LISTSERV.UGA.EDU>
> Sent: Saturday, June 18, 2022 7:50 AM
>
>
>> That's not enough to make it an array machine, any more than
>> Table lookup made the 650 an array machine.
>
> You still don't know what you are talking about.
>
> I suggest you read up on array processors (maybe
> take a look at wiki) to inform yourself what an array processor does.
>
> Every instruction* on Pilot ACE and DEUCE is capable of
> operating on two or more values.
>
> The examples that I gave were just two, each operating on an array of
> 32 values,
> adding / subtracting to / from the accumulator.
>
> _________________
> * except for about a dozen instructions initiating I/O operations,
> or changing the state of the computer.

Reply via email to