Re: DMA successes with Verilog

From: silverdr_at_wfmh.org.pl
Date: Thu, 14 Jun 2018 21:42:30 +0200
Message-Id: <B08328E9-98F3-41E5-95D7-E1DD4F7CC3F0@wfmh.org.pl>
> On 2018-06-14, at 20:35, smf <smf@null.net> wrote:
> 
>> Probably yes. It's a matter of wording but to my understanding it's also a form of DMA.
> 
> DMA means that hardware accesses memory without the help of a CPU.

That's what I mean. What I am looking at is a piece of hardware that accesses memory without CPU aid.

> The alternative is PIO where the CPU fetches the data from memory and stores it in the registers, a good example of this is the Atari 2600.
> 
> The Atari 2600 doesn't use DMA, so as the 6507 is running the game code it has to load the graphics data and store it in the output shift registers.

Roger.

> Two CPU's accessing the same memory is not DMA.

It may be hard to define precisely what is and what is not a CPU in those designs where one cat basically put a whole computer inside a chip (or use a uC as a quasi-logic/glue chip but my understanding here is that a circuit (whether integrated, discrete or PLD) here is not a CPU in the sense/case I am talking about. It is more a custom _data_ processor. An example: audio/video encoder chip: the CPU leaves raw data for it to encode and sets the params up but then doesn't do anything with this data / memory until the encoding by a specialised and much more efficient circuit is finished. The circuit fetches the raw data, processes it and stores the results all by itself, directly accessing memory, without any further help from the CPU.

-- 
SD! - http://e4aws.silverdr.com/
Received on 2018-06-14 22:00:05

Archive generated by hypermail 2.2.0.