GLOBAL ENTERPRISE ROUTERS MARKET: DEMAND, UPCOMING TRENDS AND BUSINESS OPPORTUNITY 2020 TO 2024

GLOBAL ENTERPRISE ROUTERS MARKET: DEMAND, UPCOMING TRENDS AND BUSINESS OPPORTUNITY 2020 TO 2024

Fancy Wang 1603 2020

The following sections are from John F. Kim and Internet


The Global Enterprise Routers Market is poised to grow at a CAGR of 6.2%, during the forecast period (2019 – 2024). Tri-Band Routers are coming in trend which is capable of multiple Gigabit speeds, with 4,000 Gbps on the 5 GHz band. It also reduces signal interference even more than dual-band devices.

No alt text provided for this image


 Increasing network virtualization is driving the market. The growing demands placed on IT infrastructure have made network virtualization an increasingly popular solution for businesses ranging from small to enterprise-level, due to reduced hardware & power consumption, which will drive the demand of enterprise router efficiently.

No alt text provided for this image


– Rapid growth in IP traffic enterprise is driving the market as the demand for bandwidth is climbing. Offices are leaning more on applications and customers are expecting more cloud-based services. WAN traffic grows at a rate of around 200 percent in real estate, energy, and travel sectors, which increases the penetration of the enterprise router.

– Security and privacy concerns are restraining the market. According to the study, five out of every six routers in the United States are running out-of-date firmware with exploitable security vulnerabilities, which is also hindering the market growth globally.

The enterprise routers market is growing as IP traffic have increased five-fold during the past five years and is still increasing exponentially. So, to support this huge growth in IP traffic, routers need to be upgraded, thus growing the enterprise routers market in the various end-user segment such as BFSI, IT & Telecom, Transport, etc.

Wireless Connectivity Through Access Point is Driving the Market

No alt text provided for this image


– An ordinary wireless router can only support 10-20 users access, while wireless access router point is able to allow over 50 or even hundreds of users access, and it has stronger ability to send and receive signals.

– Generally, the range of signal transmission that a wireless router can cover is just dozens of meters. A wireless access router can cover further distances, up to 100-300 meters, which enables users to roam freely in the network. Especially for enterprises, their office space is usually larger, and even some need to communicate across the buildings, and the number of users accessing the network is so large that they need a larger wireless router network coverage.

– The networking mode of the wireless router is relatively single with low flexibility. On the contrary, a wireless access point has a variety of modes to choose, which meant to be very flexible, such as Simplex AP, Wireless Client, Wireless Bridge, Multi-point Bridge, etc. and it can be managed in a centralized way with the cooperation.

– Netgear Nighthawk X10 AD7200 is one of industry’s fastest router for media streaming with Plex media server. Nighthawk X10 with 802.11ac and 802.11ad WiFi technology delivers ultra-smooth 4K streaming, VR gaming and instant downloads. With a powerful 1.7GHz Quad-core processor, 60GHz 11ad, and Quad-stream architecture, one can enjoy combined wireless speeds of up to 7.2Gbps, which drives the market.

No alt text provided for this image



North America Holds the Largest Share in Enterprise Routers Market

– North America is expected to hold the largest market size with the United States holding the major chunk of the market due to high dependence on internet connectivity.

– Most of the enterprise router providing companies are headquartered in North America. Also, most of the cloud service providers and network virtualization companies are set up in this region, which will also boost the growth rate of the market.

No alt text provided for this image


– Growing demand for hyperscale data centers is now steering growth for network hardware industry in the United States. Surging demand from hyperscale cloud providers led to a record 303 megawatts (MW) of absorption in the seven primary United States data center markets in 2018, which will lead to the demand of high IP traffic, by which the demand of enterprise router will increase.

– Digi International announced that its TransPort family of M2M routers now support Gobi 4G LTE. They are the first M2M routers that give customers the ability to connect to either Verizon, AT&T, or Sprint’s 2G, 3G and 4G networks across North America using a single device. This fully programmable and upgradeable enterprise level 3G and 4G LTE remote communications solutions are ideal for transportation, retail, POS, utilities, energy, medical, financial, digital signage, cellular backup, and remote device connectivity applications.

Competitive Landscape

The enterprise routers market is fragmented as the global players are investing in R&D for the latest innovations in the product for several small to medium-sized companies. Key players are Cisco Systems, Inc., Juniper Networks, Inc., Hewlett Packard Enterprise, etc. The companies operating in the market continuously strive to launch new solutions to help and assist service providers to achieve a holistic approach to infrastructure transformation and gain new orders in that course for strategic aim.

– Feb 2019 – Juniper Networks, announced a major refresh to its metro, edge and core solutions to accelerate service providers’ 5G transformation. Comprised of IP optimized silicon enabling industry to lead 400GbE density on a new 14.4Tb line card, new ACX access and aggregation platforms and an expanded MX 5G router portfolio.

These new speeds are so hot that, like the ageless celebrities you just saw on the Oscar Night red carpet, we say “25 is the new 10 and 50 is the new 40.” But whoa! Sure everyone wants to look younger for the camera, but no 25-year old actor wants to look 10. More importantly, why would anyone want 25GbE or 50GbE when we already have 40GbE and 100GbE?


The answer is about performance and cost at cloud scale.

 

SpeedLanesFibers/ Wires*Ratified byFirst Product Avail.1GbE1 x 12IEEE in 1999200010GbE1 x 102IEEE in 2002200325GbE1 x 25225GE Consortium in 2014,

IEEE in processlate 201540GbE4 x 108*IEEE in 2010201150GbE2 x25425GE Consortium in 2014,

IEEE in processlate 2015100GbE10×1020*IEEE in 20102011 for switches only100GbE4 x 258*IEEE in 2010early 2015

*Optical networking can combine multiple lanes into one fiber with wavelength division multiplexing (WDM), so a 40GbE or 100GbE (4×25) fiber cable could use either 8 fibers without WDM or 2 fibers with WDM.

Figure 2: Chart showing Ethernet speed standards including the number of lanes and wires/fibers typically needed.

 

Performance

First, servers are getting faster. A mid-range-end x86 server can easily drive 20Gb/s of network throughput, while high-end servers can push 40, 50, or even 80Gb/s.

 

Second, storage is getting faster– NVMe SSDs today support 3GB/s (24Gb/s) of sequential read throughput, each! Vendors are sampling NVMe SSDs that support over 4GB/s (32Gb/s). A recent Mellanox demo with Microsoft Windows Storage Spaces showed one server with three NVMe SSDs could sustain 95Gb/s of throughput over a Mellanox 100GbE link using the Windows SMB Direct protocol. (See the Windows 100GbE demo on the Mellanox Microsoft solutions page.)

 

Third—here’s the other Oscars tie-in—more movies and TV shows are being shot and edited using 4K (3840×2160 or 4096×2160) resolution. Uncompressed 4K video needs 12 to 26 Gb/s of bandwidth per stream (depending on color depth) to capture, edit, and render, so both 10GbE and 8Gb Fibre Channel are too slow. Some studios are already testing 6K and 8K video and expect 8K resolution video will be commonplace by 2020.

 

At this point someone is obliged to point out that not all servers are high-end, not all applications read from 3 NVMe SSDs at maximum speed, and not every film is shot and viewed in 4K resolution. All true, but even today’s low-end and midrange servers with 3 SAS SSDs or a dozen fast HDDs (or 20 slow HDDs) can easily overwhelm a single 10GbE connection. In many cases, 25GbE is ideal for many new servers, SSDs, and ultra-high definition video workflows.

 

Cost

Obviously, if higher bandwidth didn’t cost more, everyone would use it, but cost matters too. Many mid-range application servers and clients today need more than 10Gb/s but less than 25Gb/s.

But why not just use 2x10GbE or 4x10GbE? Because it’s less expensive to use 25GbE, which provides 2.5x more bandwidth over the same number of wires and switch ports but costs about 1.5x more per NIC port than 10GbE. This gives 25GbE a 40% lower cost/bandwidth than 10GbE. Analysts forecast 25GbE will reach the same price per port as 10GbE in just 3 years, by 2019, giving 25GbE an even better cost/bandwidth advantage.

No alt text provided for this image


 What if you need more than 25Gb/s?

You can always use 40GbE (on 8 wires/fibers), which is growing rapidly in cloud and enterprise data centers. But there is also a 50GbE option which carries 25% more data. The adapter ports cost about the same as 40GbE, making the cost/bandwidth 20% lower than for 40GbE. Part of the cost savings is simply from using one NIC and one switch port instead of two, but another important reason is due to the way 25, 50 and 100 do signaling. 10 and 25GbE both use a single lane (two wires or fibers), 50GbE uses two lanes, and 40 and 100GbE use 4 lanes (eight wires or fibers in most cases). The more lanes used, the greater the power consumption for the adapter and switch, the fewer ports you can squeeze into one switch, and the more expensive the cable. For example a 100GbE switch port can support one connection at 40 or 100GbE but two connections of 25 or 50GbE. Because 25, 50 and 100GbE send and receive more data per lane than 10 or 40GbE, they make networking more efficient and lower cost.

No alt text provided for this image


Networking at Cloud Scale

Large enterprises may deploy thousands of servers per year while large cloud and Web 2.0 customers deploy thousands of servers per month and may roll out 2 or 3 new data centers each year. At these quantities, small cost savings in each network connection add up to large numbers and additional special factors come into play.

 

Besides the lower cost of the adapter ports themselves, one 25GbE port consumes less power, half the cabling, and half the switch ports of two 10GbE ports; reducing the number of switch ports saves more power plus rack space. Also, it’s easier to manage one 25 or 50GbE connection than bonding two or four 10GbE connections per server, and you reduce the number of inter-switch links (ISLs) by 60% when using 100GbE instead of 40GbE connections between switches.

 

As a bonus, 25GbE can run over existing fiber optic cable plant designed for 10GbE, and 50 or 100GbE can run over existing fiber optic cable plant designed for 40GbE, just by changing the transceivers. This is a tremendous cost savings for large data centers that already use fiber cabling for 10 and 40GbE. And for those hyperscale customers who use the Open Compute Platform (OCP), these special server, rack and NIC designs result in additional capital cost, power and space savings, as well as the ability to have four servers share one NIC using a special Mellanox feature called Multi-Host.

No alt text provided for this image


Saving $100 per server on adapters and another $300 per server in cabling and switch ports is only mildly interesting for customers deploying 20 servers per quarter ($8000 savings) but immensely appealing for those deploying 20,000 servers at a time ($8 million savings), plus the ongoing savings in power, cooling, and rack space.

 

The Bottom Line About the Bottom Line

The key message is that most new servers and the flash storage inside require connections faster than 10GbE. Likewise, storage systems with 20+ spinning disks or 3+ SSDs (even the slower SAS/SATA SSDs) often need 20Gb/s or more of bandwidth. This makes 25GbE ideal for new server deployments, 40 or 50GbE ideal for storage, and 100GbE the most efficient link to aggregate switch traffic or bridge datacenter racks and rows. 25, 50 and 100 offer not only faster performance but lower cost and easier scaling for large datacenters.

 

Mellanox is happy to support all these solutions as we are the clear market leader in 40GbE adapters and first vendor to ship 25, 50, and 100GbE adapters. In fact, Mellanox is the only vendor to offer an end-to-end solution for 10, 25, 40, 50, and 100GbE, including adapters, switches, cables and transceivers, and a leading supplier of OCP-compatible network cards.

 

You can see the Mellanox 25, 50, and 100GbE solutions, including our OCP adapters, next week at the 2016 US OCP Summit, March 9-10 in San Jose.

RESOURCES

Chloe Jian Ma’s Blog in EE Times: Schumpeter’s Gale and 25 Is The New 10

Q&A with Kevin Deierling in Electronic Design: Q&A: Ethernet Evolution—25 is the New 10 and 100 is the New 40

Visit the 25 Gig Ethernet Consortium

Mellanox Multihost adapter solutions

Mellanox Ethernet switch systems, first to support 10/25/40/50/56/100 GbE

Mellanox Ethernet Adapters supporting 10/25/40/50/56/100 GbE

Mellanox cables and transceivers for 10/25/40/50/56/100 GbE

 About Us

About Shenzhen Speed Technology Co.,ltd, we are a research and development switch manufacturer located in Shenzhen, Guangdong. We started in 1997 with a variety of product lines, supporting OEM, ODM customization.

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics