What differences are there between "home" switches and "professional" switches?
Looking at the official specifications for the D-Link device and comparing them with the NETGEAR device, I would go with the D-LINK.
D-LINK presents the same relevant specifications, even more than NETGEAR. Also it's cheaper :)
Concerning the latency issue, the D-LINK provides you with a QoS method for controlling traffic priority, allowing you to prioritize the most important traffic to reduce latency for time-sensitive transmissions, such as streaming media.
The reason behind my choice:
╔═══════════════════════════════════════╦══════════════════╦══════════════════╗
║ RELEVANT SPECIFICATION ║ D-LINK ║ NETGEAR ║
╠═══════════════════════════════════════╬══════════════════╬══════════════════╣
║ Packet Filtering/Forwarding Rates ║ ║ ║
║ Ethernet (pps per port): ║ 14,880 ║ 14,800 ║
║ Fast Ethernet (pps per port): ║ 148,800 ║ 148,000 ║
║ Gigabit Ethernet (pps per port): ║ 1,488,000 ║ 1,480,000 ║
╠═══════════════════════════════════════╬══════════════════╬══════════════════╣
║ Standards ║ ║ ║
║ IEEE 802.3 10BASE-T ║ Yes ║ --- ║
║ IEEE 802.3i 10BASE-T ║ --- ║ Yes ║
║ IEEE 802.3u 100BASE-TX ║ Yes ║ Yes ║
║ IEEE 802.3ab 1000BASE-T ║ Yes ║ Yes ║
║ ANSI/IEEE 802.3 Nway autonegotiation ║ Yes ║ --- ║
║ IEEE 802.3x Flow control ║ Yes ║ Yes ║
║ IEEE 802.1p QoS ║ Yes ║ --- ║
╠═══════════════════════════════════════╬══════════════════╬══════════════════╣
║ buffer ║ ║ ║
║ Size: ║ 128 KBytes/device║ 2 Mb ║
║ Method: ║ store-and-forward║ store-and-forward║
╚═══════════════════════════════════════╩══════════════════╩══════════════════╝
Product pages with Official Specifications:
- D-Link: DGS-1005G 5-Port Gigabit Desktop Switch
- NETGEAR: ProSafe® 16-port Gigabit Ethernet Switch 10/100/1000 Mbps JGS516
The main differences between "home" and "professional" equipment are
- manageability: there is an interface to configure priorities, virtual circuits and load balancing, and to query port status and statistics
- scalability: the switch needs to maintain a table which devices are connected to which port. If this table is full because more devices are on the network than entries exist in the table, then performance degrades (when a device is unknown, packets to the device are broadcast over the entire network, in the hope that the device answers and so becomes known). Consumer devices typically have table sizes of 20-50 entries; professional equipment should support 1000 entries at least
- extensibility: many professional switches are "stackable", meaning that you can combine multiple switches into a single one, configure it centrally and have shared address tables over all of them.
- management protocol support: to support redundant links, switches have to be aware of each other, and discuss the network topology via protocols such as Spanning Tree Protocol (STP). Bad things happen if there are cyclic links and no inter-switch routing protocol is enabled.
Both of the devices you mention I'd classify as "home" equipment, and I don't think the Netgear switch will improve things. You want something that is "managed", which allows you to set up traffic priorities.
Is there a significant difference that we would notice in terms of latency between these two switches?
I didn't see any mention of it in those descriptions, but the latency of a switch will mostly be determined by its switch buffering method: either store-and-forward or cut-through. Store-and-forward means that the complete Ethernet frame has to be received (intact) before that frame leaves the switch to the next destination. Cut-through switching will try to retransmit the Ethernet frame as soon as the IP header with the destination address has been received and processed.
Almost all SOHO/home switches are store-and-forward. Unless stated otherwise, assume that a switch uses store-and-forward. Note that cut-through switches will propagate bad frames, whereas a store-and-forward switch will filter out bad frames.