Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!bloom-beacon!mit-eddie!uw-beaver!tektronix!orca!tekecs!frip!andrew From: andrew@frip.gwd.tek.com (Andrew Klossner) Newsgroups: comp.arch Subject: 88k speed on FP add/neg/mov (was Re: RISC machines and scoreboarding) Message-ID: <10136@tekecs.TEK.COM> Date: 5 Jul 88 20:42:10 GMT References: <1082@nud.UUCP> <2438@winchester.mips.COM> <1098@nud.UUCP> <2459@gumby.mips.COM> <1110@nud.UUCP> <2483@gumby.mips.COM> Sender: andrew@tekecs.TEK.COM Organization: Tektronix, Wilsonville, Oregon Lines: 20 In article <2483@gumby.mips.COM>, earl@mips.COM (Earl Killian) presents a table of floating point times for the 88k and R3010, with this entry: > 88100 R3010 cycle > Operation cycles cycles ratio > ====================================== > abs/neg/mov 5? 1 To clear up the question mark: on the 88k, floating point values are kept in general registers, so "abs" is done by ANDing to 0 the sign bit, "neg" is done by XORing the sign bit, and "mov" is done with conventional register-to-register move instructions. Thus: abs 1 neg 1 sp mov 1 dp mov 2 -=- Andrew Klossner (decvax!tektronix!tekecs!andrew) [UUCP] (andrew%tekecs.tek.com@relay.cs.net) [ARPA]