Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!ucbvax!decwrl!amdcad!crackle!tim
From: tim@crackle.amd.com (Tim Olson)
Newsgroups: comp.sys.amiga
Subject: Re: RISC coprocessor for Amiga?
Message-ID: <26165@amdcad.AMD.COM>
Date: 29 Jun 89 02:50:28 GMT
References: <18689@louie.udel.EDU>
Sender: news@amdcad.AMD.COM
Reply-To: tim@amd.com (Tim Olson)
Organization: Advanced Micro Devices, Inc. Sunnyvale CA
Lines: 44
Summary:
Expires:
Sender:
Followup-To:

In article <18689@louie.udel.EDU> 451061@uottawa.bitnet writes:
| When comparing Drystones, indeed RISC technology seems fantastic, but the real
| world out there, including the simulated ray-traced world talks floating point.
| And on floating point benchmarks, the RISC and CISC architectures are on par.

What data do you have to back up your claim?  A look at the MIPS
Performance Brief, issue 3.6, shows RISC processors consistently
outperforming CISC processors on both integer and FP applications,
including the Livermore Loops, LINPACK, Whetstone, and SPICE. [well,
some of the high-end IBMs and AMDAHLs creep in there...]

As far as ray-tracing is concerned, here are some of the statistics from
a simulation of a ray-tracing program which are very interesting:

---------- Instruction Mix ----------
	  1.11% Calls
	 10.54% Jumps
	 17.36% Loads
	  5.26% Stores
	  8.10% FP ops
		  0.12% Convert
		  0.04% Sqrt
		  0.00% Feq
		  0.09% Deq
		  0.00% Fgt
		  0.57% Dgt
		  0.00% Fge
		  2.37% Dge
		  0.00% Fadd
		  0.95% Dadd
		  0.00% Fsub
		  1.20% Dsub
		  0.00% Fmul
		  2.51% Dmul
		  0.00% Fdiv
		  0.25% Ddiv


i.e. the FP-intensive ray-tracing program is really only 8.1% FP
instructions; the remaining are the standard ALU ops, loads, stores,
jumps, etc.
	-- Tim Olson
	Advanced Micro Devices
	(tim@amd.com)