Reports & Publications
Proteon ProNET CNX 600 1994 Industry Benchmark Local Token Ring Bridge Performance
Login or create an account to download this report
Abstract
Proteon commissioned Tolly, as part of the 1994 Industry Benchmark for Local Token Ring Bridges, to evaluate the ProNET CNX 600, with the main focus on measuring local Token Ring bridge performance across source-route bridging and optional IP and IPX routing scenarios. The project examined frame-handling efficiency, maximum effective unidirectional throughput, and the platform’s ability to deliver near-wire-speed forwarding on 16Mbit/s Token Ring networks.
The December 1994 Technology Spotlight describes the ProNET CNX 600 as a five-slot multi-protocol bridge/router that supports Token Ring, Ethernet, FDDI, and serial line topologies. The tested unit used dual Token Ring modules based on the Texas Instruments TMS380 chipset and supported Source Route, Adaptive Source Route Transparent, Translational, and Spanning Tree bridging. Routed protocol support listed on page 3 includes IPX, TCP/IP, OSI, DECnet Phase IV and V, XNS, AppleTalk, Apollo DOMAIN, and Banyan VINES. The product-specification panel also notes support for 4 or 16Mbit/s Token Ring, STP and UTP media, maximum Token Ring frame size of 17,997 bytes, and management through SNMP, MIB II, OneView, and NetView.
Tolly’s benchmark measured maximum effective unidirectional throughput across a range of Token Ring frame sizes. The bar chart on page 1 shows source-route bridging throughput of 3.06Mbit/s at 30-byte frames, 6.77Mbit/s at 64-byte frames, 12.29Mbit/s at 256-byte frames, 12.82Mbit/s at 512-byte frames, 13.03Mbit/s at 1,024-byte frames, 13.04Mbit/s at 2,048-byte frames, and 13.01Mbit/s at 4,096-byte frames. Tolly states on page 2 that this corresponds to about 80% of wire speed for 512-byte frames and larger. The same page explains that smaller frames produce lower effective throughput because Token Ring bridges forward traffic on a frame-by-frame basis and are limited by the frame-processing capability of the Token Ring chipset.
The report also converts the measured throughput into frame-per-second values. On page 2, Tolly lists 13,673fps at 30-byte frames, 13,221fps at 64-byte frames, 6,000fps at 256-byte frames, 3,130fps at 512-byte frames, 1,590fps at 1,024-byte frames, 796fps at 2,048-byte frames, and 397fps at 4,096-byte frames. Tolly further notes that optional IP and IPX routing throughput tracked closely with the source-route bridging results. Testing used two-port, four-port, and six-port configurations built from multiple 16Mbit/s Token Ring segments, PC-based frame generators on the source rings, and analyzers on the destination rings. Overall, the report presents the Proteon ProNET CNX 600 as a high-performance local Token Ring bridge/router delivering strong frame handling, about 80% of wire-speed throughput on larger frames, and broad multi-protocol network support.