[net.lan] Proteon Ring Performance?

athey@ungol.UUCP (Charles L. Athey III) (05/23/85)

We have need of performance numbers for proteon rings.  We are currently
running a number of ethernets and need to make some long range plans.
Currently ethernet just doesn't give us good enough performance - we
need in the 5-10 Mbit range, ethernet is only giving 100-800 Kbits.

We are currently using Vaxen, Suns, etc., mostly running 4.2 UNIX, but
not exclusively.  We need such high transfer rates to move graphics files
from Crays to workstations.  Please respond directly to me & I will
post a summary to the net.   Thanks in advance.

			Chuck Athey
			ARPA: athey@lll-crg
			UUCP: {seismo|dual}!lll-crg!athey

hans@kuling.UUCP (Hans Eriksson) (06/01/85)

In article <304@ungol.UUCP> athey@ungol.UUCP (Charles L. Athey III) writes:
>We have need of performance numbers for proteon rings.  We are currently
>running a number of ethernets and need to make some long range plans.
>Currently ethernet just doesn't give us good enough performance - we
>need in the 5-10 Mbit range, ethernet is only giving 100-800 Kbits.
>
>We are currently using Vaxen, Suns, etc., mostly running 4.2 UNIX, but
>not exclusively.  We need such high transfer rates to move graphics files
>from Crays to workstations.  Please respond directly to me & I will
>post a summary to the net.   Thanks in advance.

Could you give some numbers on how many machines and what kind is connected
to how many ethers? "100-800 Kbits", is it per machine or a net-maximum? We
are in the process of building a net of dozen Lispmachines, a couple of
Vaxen, some ten of Sun and a bunch of MacIntosh. Ethernet and some hardwork
will tie them together (I hope). The big question is: Is one Ethernet enough?