QSFP28: 100G, 50G, 40G*, 25G, and 10G* Ethernet
QSFP+: 40G and 10G Ethernet
*Depending on transceiver capabilities
Number of test ports (software configurable)
2x100GE, 4x50GE, 2x50GE PAM4, 2x40GE, 8x25GE, and 8x10GE
1 x 100GBASE-SR4/LR4/CR4 (as per 802.3bj), or
2 x 50GBASE-SR2/LR2/CR2 (as per Consortium), or
1 x 50GBASE-SR/CR (PAM4) (as per 802.3cd), or
1 x 40GBASE-SR4/LR4/CR4 (as per 802.3ba), or
4 x 25GBASE-SR/LR/CR (as per 802.3by/Consortium), or
4 x 10GBASE-SR/LR/CR (as per 02.3ae)
Actual interface options depend on the capabilities of the inserted transceiver.
Both cages must run with the same base interface configuration (e.g. 2 x 100GE)
** As defined by 25/50 GigabitEthernet Consortium
Forward Error Correction (FEC)
RS-FEC (Reed Solomon) 528,514,t=7, IEEE 802.3 Clause 91 (100GE)
RS-FEC (Reed Solomon) 544,514,t=15, IEEE 802.3 Clause 134 (50GE PAM4)
RS-FEC (Reed Solomon) 528,514,t=7, IEEE 802.3 Clause 108 (25GE)
RS-FEC (Reed Solomon) 528,514,t=7, 25/50G Ethernet Consortium (25/50GE)
Number of transceiver module cages
2 x QSFP28/QSFP+
Port statistics (2)
Link state, FCS errors, pause frames, ARP/PING, error injections, training packet
All traffic: RX and TX Mbit/s, packets/s, packets, bytes
Traffic w/o test payload: RX and TX Mbit/s, packets/s, packets, bytes
Adjustable Inter Frame Gap (IFG)
Configurable from 16 to 56 bytes, default is 20B (12B IFG + 8B preamble)
Transmit line rate adjustment
Ability to adjust the effective line rate by forcing idle gaps equivalent to -1000 ppm (increments of 10 ppm)
Transmit line clock adjustment
From -400 to 400 ppm in steps of 0.001 ppm (shared across all ports)
Supported (configurable IP and MAC address per port)
System is fully field upgradeable to product releases (FPGA images and Software)
Histogram statistics (2)
Two real-time histograms per port. Each histogram can measure one of RX/TX packet length, IFG, jitter, or latency distribution for all traffic, a specific stream, or a filter.
Enable/disable of optical laser or copper link
IGMPv2 multicast join/leave
IGMPv2 continuous multicast join, with configurable repeat interval
• L1RX2TX – RX-to-TX, transmit byte-by-byte copy of the incoming packet
• L2RX2TX – RX-to-TX, swap source and destination MAC addresses (1)
• L3RX2TX – RX-to-TX, swap source and destination MAC addresses and IP addresses (1)
• TXON2RX – TX-to-RX, packet is also transmitted from the port
• TXOFF2RX – TX-to-RX, port’s transmitter is idle
• Port-to-port – Inline loop mode where all traffic is looped 100% transparent at L1
• Initial Accuracy is 3 ppm
• Frequency drift over 1st year: +/- 3 ppm (over 15 years: +/- 15 ppm)
• Temperature Stability: +/- 20 ppm (Total Stability is +/- 35 ppm)
Number of transmit streams per port
256 (wire-speed) continuous. Each stream can generate millions of traffic flows through the use of field modifiers.
Test payload insertion per stream
Wire-speed packet generation with timestamps, sequence numbers, and data integrity signature optionally inserted into each packet.
Stream statistics (1)
TX Mbit/s, packets/s, packets, bytes, FCS error
Burst size and density can be specified. Uniform and bursty bandwidth profile streams can be interleaved
16-bit or 32-bit header field modifiers with inc, dec, or random mode. Each modifier has configurable bit-mask, repetition, min, max, and step parameters. 8 16-bit modifiers per stream or 4 32-bit modifiers per stream.
Packet length controls
Fixed, random, butterfly, and incrementing packet length distributions. Packet length from 56 to 12288 bytes
Repeated user specified 1 to 18B pattern, an 8-bit incrementing pattern
Fixed full custom payloads can be generated for each stream with payload sizes up to 12288 bytes
Undersize length (56 bytes min.) and oversize length (12288 bytes max.) packet lengths, injection of sequence, misorder, payload integrity, and FCS errors
TX packet header support and RX autodecodes
Ethernet, Ethernet II, VLAN, ARP, IPv4, IPv6, UDP, TCP, LLC, SNAP, GTP, ICMP, RTP, RTCP, STP, MPLS, PBB, or fully specified by user
Packet scheduling modes
• Normal (stream interleaved mode). Standard scheduling mode, precise rates, minor variation in packet inter-frame gap.
• Strict Uniform. New scheduling mode, with 100% uniform packet inter-frame gap, minor deviation from configured rates.
• Sequential packet scheduling (sequential stream scheduling). Streams are scheduled continuously in sequential order, with configurable number of packets per stream.
• Burst. Up to 10000 packets per stream are organized in bursts. Bursts from active streams form a burst group. The user specifies time from start of one burst group till start of next burst group
Number of traceable Rx streams per port
Automatic detection of test payload for received packets
Real-time reporting of statistics and latency, loss, payload integrity, sequence error, and misorder error checking
Jitter (Packet Delay Variation) measurements compliant to MEF10 standard with 8 ns accuracy. Jitter can be measured on up to 32 streams.
Stream statistics (2)
• RX Mbit/s, packets/s, packets, bytes.
• Loss, payload integrity errors, sequence errors, misorder errors
• Min latency, max latency, average latency
• Min jitter, max jitter, average jitter
Latency measurements accuracy
±16 ns (opto/elec).
Latency measurement resolution
8 ns (Latency measurements can calibrate and remove latency from transceiver modules)
Number of filters:
• 6 x 64-bit user-definable match-term patterns with mask, and offset
• 6 x frame length comparator terms (longer, shorter)
• 6 x user-defined filters expressed from AND/OR’ing of the match and length terms.
Filter statistics (2)
Per filter: RX Mbit/s, packets/s, packets, bytes.