Date: Mon, 26 May 1997 21:14:40 +0200 From: Ollivier Robert <roberto@keltia.freenix.fr> To: hardware@FreeBSD.ORG Subject: Re: Intel Pentium II released Message-ID: <19970526211440.51805@keltia.freenix.fr> In-Reply-To: <199705252226.PAA01640@silvia.HIP.Berkeley.EDU>; from Satoshi Asami on Sun, May 25, 1997 at 03:26:15PM -0700 References: <19970525131757.62904@keltia.freenix.fr> <199705252226.PAA01640@silvia.HIP.Berkeley.EDU>
next in thread | previous in thread | raw e-mail | index | archive | help
According to Satoshi Asami: > Ouch. That's horrible! Even a p P133 with 2.1.5 (i.e., without the > fast copyin/out) can get 80MB/s on both.... The culprit is the bcopy/bzero/copyin FPU code. I suspected it a bit as the FPU in the K6 is slower but the results are interesting. With flags 0x4 or 0x7 (so the real culprit is the copyin/out code) I get this : 202 [21:04] roberto@keltia:~> dd if=/dev/zero of=/dev/null bs=64k count=16000 1048576000 bytes transferred in 9.008729 secs (116395552 bytes/sec) 203 [21:07] roberto@keltia:~> dd if=/dev/zero of=/dev/null bs=1m count=1000 1048576000 bytes transferred in 14.700602 secs (71328779 bytes/sec) *much* better. I still can't overclock it at 2.5x 83 MHz. Only 2.5x 75 MHz works with the K6 whereas it worked at 2x 83 MHz for the P133. Oh well, I can live with that... My kernel compilation time (not GENERIC but my own config) fell from 4:55 down to 3:29 :-) -- Ollivier ROBERT -=- FreeBSD: There are no limits -=- roberto@keltia.freenix.fr FreeBSD keltia.freenix.fr 3.0-CURRENT #10: Fri May 23 22:47:39 CEST 1997
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19970526211440.51805>