home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!think.com!news!jcallen
- From: jcallen@marley.think.com (Jerry Callen)
- Newsgroups: comp.arch
- Subject: MIPS R4400
- Date: 6 Nov 92 13:52:00
- Organization: Thinking Machines Corporation, Cambridge MA, USA
- Lines: 41
- Distribution: comp
- Message-ID: <JCALLEN.92Nov6135200@marley.think.com>
- NNTP-Posting-Host: marley.think.com
-
-
- The Nov 2 issue of EE Times contains an article on the new MIPS R4400
- chip. Quick summary:
-
- - 2.3 million transistors (vs 1.1 million(?) for the R4000)
- - .6 vs .8 micron process
- - 75MHz external clock, 150MHz internal
- - 3.3V and 5V versions available
- - I and D caches doubled to 16K each, still direct mapped.
- - MIPS has sample silicon from at least one source, and
- expects sampling this quarter and production chips in
- early 1993.
-
- The die photo looks a lot like the R4000 die photo, but with
- noticably larger caches. The article notes that much of the R4000
- design was reused (no surprise).
-
- The article mentions that a "large write buffer" was added. Could
- someone from MIPS comment on the size of the buffer, and the implications
- for the memory system?
-
- Also, where are all the extra transistors going? I believe I've read in
- posts by John Mashey that the R4000 uses about 600k transistors for
- everything except the cache, which suggests that the two 8K caches on the
- R4000 use about 500K transistors, so I might (naively) expect the two 16K
- caches on the R4400 to use about a million. Did the write buffer consume
- 700K transistors?!?
-
- The article also says that MIPS is predicting performance at 75MHZ,
- based on simulations, to be:
-
- SPECint89 = 95
- SPECfp89 = 126
-
- [I was surprised to see MIPS use SPEC 89 numbers...]
-
- --
- -- Jerry Callen
- jcallen@world.std.com (preferred)
- jcallen@think.com (OK, too)
- {uunet,harvard}!think!jcallen (if you must)
-