Business , NETGEAR
share
145 views

There is no question that edge switches have become more sophisticated in recent year. In fact, we now see some edge switches with throughput and failover capabilities that, in the past, one would only find in a pricey, chassis-based LAN switch.

NETGEAR commissioned Tolly to benchmark performance, power consumption and non-stop forwarding (NSF)stack failover of its M4300 Intelligent Switch family along with similar products from Cisco Systems and Aruba. Testing was done on single switches and stacks of three switches from each vendor.

In this blog post, I will give you an overview of the test results. I recommend you also look at our full Tolly report #220143 for all of the details. You can find that report at both the Tolly and NETGEAR sites.
NETGEAR site at: https://blog.netgear.com/blog/tolly-test-of-netgear-lan-switches-vs-cisco-and-aruba/

The NETGEAR M4300-52G LAN switch was compared to the Aruba 2930F-48G-4SFP+ and the Cisco Systems SG550X-48P switches. Testing was performed both on single switches and a stack of three switches from each vendor. Each switch provided 48 Gigabit Ethernet ports and four 10 Gigabit Ethernet uplink/stacking ports. Herewith, the results in a nutshell:

Performance

The NETGEAR M4300 delivered line-rate throughput at all frame sizes across all ports, Gigabit Ethernet and 10GbE. This performance matched that of the Cisco switch and was better than Aruba that only reached line-rate at 256-byte frames and above. See Figure 1.Fig 1 – Throughput


Latency

The latency of all three vendor switches is extremely low. In fact, it is measured now in microseconds or millionths of a second. All the switches had latency in single digits of microseconds. Low latency? Check. See Figure 2.Fig 2 – Latency

Stack Failover

This test included several scenarios using three switches from each vendor. A key proof point here was to illustrate the instant failover of NETGEAR Non-Stop Forwarding in the event a “commander” (i.e. master) switch went down. These tests showed zero second failover for NETGEAR compared to 28 seconds for Cisco in one scenario and 8.2 seconds for Cisco in another. See Table 1 for results and Figure 3 for a logical diagram.Table 1 – Failover results
  Fig 3 – Stack Testing Diagram

Power Consumption

LAN switches are “always on” devices and thus power efficiency is important. Tolly benchmarked the power consumption of each switch according to industry standard ATIS and TEER recommendations. The NETGEAR M4300 provided the same throughput as Cisco with significantly lower power consumption. Thus, the TEER calculation of Gbps/Watt is better for NETGEAR at 1.88 than for Cisco at 1.37. See Table 2.Table 2 – Power Consumption Results

Systems Under Test

For details of the systems under test, see Table 3.Table 3 – Systems Under Test

Conclusion

Remember, all LAN switches are not created equal. Whenever possible, get benchmark data and compare so that you know that you will be getting the best switch for the money and for your business needs.

Founder, The Tolly Group

Kevin Tolly

Founder, The Tolly Group

Tolly
  • View Comments

    Leave a Reply

    Your email address will not be published. Required fields are marked *