Date: Tue, 11 Jul 1995 14:16:40 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com> To: dennis@et.htp.com (dennis) Cc: freebsd-hackers@freebsd.org Subject: Re: Wanted: 100bT EISA ethernet recommendation Message-ID: <199507112116.OAA16567@gndrsh.aac.dev.com> In-Reply-To: <199507111840.OAA08675@mail.htp.com> from "dennis" at Jul 11, 95 02:40:34 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > >> > >> Tom's opinion.... > >> > > >> >On Tue, 11 Jul 1995, dennis wrote: > >> > > >> >> The question is, who would build one? EISA cards are too expensive to > build > >> >> and EISA is too slow for a 100mbs medium. If someone is making them then > >> >> I'll bet they have a much bigger marketing dept than engineering. > >> > > >> > That's wrong. EISA is fast enough for 100mbs ethernet. > >> > >> It can't be wrong, because any way you slice it its an opinion. Under light > >> load anything will work, but under heavy load its nice if your bus > >> throughput is greater than the bandwidth. If your EISA card is bus mastering > >> it can take over your machine under heavy load. For a workstation, sure, but > >> not for a server. And EISA is too expensive for a workstation. > >> > >> db > > > >EISA has a bus bandwidth of 33 mega*bytes* per second. > > > OK. I'll bite. Obviously my spec sheet is old/wrong or I don't get the new > math. The orginal EISA spec was 8.3mhz / 32 bits with 4-6 cycle access. This > is 88mbs best case with a real expectation of a little better than 60mbs > actual xfer capability. The knock on EISA has always been that its not that > much faster than ISA so this 32MB/s stuff must be new. You completely ignored bus master DMA which is 1 32 bit word per cycle or 33MB/sec. Your looking at PIO data rates, not DMA rates. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199507112116.OAA16567>