Re: FPGA implementation of (bits of) the CBM2 8088 coprocessor in VHDL

Date: Mon, 30 Apr 2012 11:44:44 +0200
Message-ID: <10250203.7351335779084375.JavaMail.www@wwinf3711>
My apologies Mark.

I did receive your e-mail. I copied it to an e-mail folder and then got sidetracked and failed to answer it. I will reply without fail 

My current VHDL expects the use of 'real' chips (especially the 8088 and 8259). I have looked at a few 'soft cores' for the INTEL x86 
family over the years and find them all lacking in one department or another. I have had a commercial interest in the past - but we have 
decided for our multi-million pound refurbishment project at work to go for real Intel chips on the boards rather than an FPGA 
implementation because then I can guarantee it works and there are no nasty surprises around the corner for me! I have access to a company 
that bought all of Intel's IPR for Multibus 1 products and all of their old stock of chips and wafers.

I am, however, looking at CPU86 at the moment (which is supposedly an 8088 core) for some private work I am doing. The first thing I 
notice is a separate data input and output bus rather than a bidirectional one. This clearly simplifies the core implementation but is 
clearly not 100% compatible with a real 8088. The other thing is that it has no concept of MIN/MAX mode. This may (however) be a good 
thing for VHDLing the Commodore 8088 coprocessor card as we may be able to incorporate the 8288 chip in the process for free!

My concern is (as always) how thoroughly tested is the core and (therefore) will there be any nasty surprises down the line when you try 
to run software on it that has never been tried before. If there is a problem, you either have to ditch it and try another core (a costly 
issue even if you can find one) or try and fix the problem yourself (and since what I have seen is not as well commented as I would have 
liked...) - or contact the company who provided the core (albeit free) who would probably be glad to have you pay to fix their errors. I 
always like to see lots of good documentation, good commenting and structure and exact details of how they have tested the core before I 
even start to look at using it. For home use, however, I would drop my standards...


> Message Received: Apr 29 2012, 11:58 PM
> From: "Mark McDougall" <>
> To:
> Cc: 
> Subject: Re: FPGA implementation of (bits of) the CBM2 8088 coprocessor in VHDL
> On 30/04/2012 5:26 AM, wrote:
> > I have a "first cut" of some VHDL for the CBM2 8088 coprocessor card if
> > anyone is interested in a copy.
> I sent you a PM a week or so ago about the Vectrex. Did you receive it?
> Are you going to use a soft core 8088? Because I've just started using one 
> myself...
> Regards,
> -- 
> |              Mark McDougall                | "Electrical Engineers do it
> |  <>   |   with less resistance!"
>        Message was sent through the cbm-hackers mailing list

       Message was sent through the cbm-hackers mailing list
Received on 2012-04-30 10:00:29

Archive generated by hypermail 2.2.0.