Path: utzoo!utgpu!water!watmath!clyde!att!ihnp4!amdcad!tim
From: tim@amdcad.AMD.COM (Tim Olson)
Newsgroups: comp.arch
Subject: Re: Japanese 32-bit micro can be a 6802
Message-ID: <21690@amdcad.AMD.COM>
Date: 20 May 88 15:53:33 GMT
References: <2006@sugar.UUCP> <76700024@uiucdcsp>
Reply-To: tim@amdcad.UUCP (Tim Olson)
Organization: Advanced Micro Devices
Lines: 33

In article <76700024@uiucdcsp> gillies@uiucdcsp.cs.uiuc.edu writes:
| 
| For those of you who enjoy "diddling microcode", the AMD 2900 has been
| around almost forever (well, 10+ years -- it was used in the Xerox
| DLion, which DOES switch instruction sets, although only during
| world-swap).  And there's the AMD29000 in case you want 16 or 32-bit
| performance.  So I don't think the japanese chip is any great
| breakthrough.

There seems to be a lot of confusion out there regarding the Am29xxx
part numbers (probably due to the desire to keep the name recognition
for our new processor ;-), so to clear things up:

Am2900	- family of microprogrammable, bit-slice devices (4-bit alus, 4
	  and 12-bit sequencers, etc.) implemented in TTL and CMOS

Am29100	- family of microprogrammable devices (16-bit alus, 12-bit
	  sequencers) implemented in TTL and CMOS

Am29300	- family of "functionally partitioned" microprogrammable devices
	  (not bit-slice -- 32-bit alus, 18-bit register files [two in
	  parallel give you 32 bits + parity], 16-bit sequencers)
	  implemented in ECL internal, TTL I/O, as well as CMOS

Am29000	- family of fixed-instruction-set CMOS RISC processors (also
	  includes 29027 Arithmetic Accelerator, 29062 Integrated Cache
	  Unit and 29041 Data Transfer Controller)

There. Clear as mud now, right? ;-)

	-- Tim Olson
	Advanced Micro Devices
	(tim@amdcad.amd.com)