Category: Fiber Transceivers

Optical transceivers, Active Optic Cables, data center switches, and cable management in data centers.

Cisco Gigabit Ethernet SFP Transceiver Modules introductions

by http://www.fiber-mart.com

SFP (Small Form-Factor Pluggable) transceiver is a hot swappable I/O device that plugs into a Gigabit Ethernet port or slot, linking the port with the network. Gigabit Ethernet SFP transceivers are designed to use with Gigabit networks. They are compliant to IEEE 802.3 standards and with different types to fit for 1000Base SX, LX/LH, EX and 1000Base-T applications.
It is important to choose a proper Gigabit Ethernet transceiver to act as an interface between the device and the cable. The interconnecting cable can be made of copper or fiber optic. This choice of transceiver is usually made based on the type of fiber optic cable.
Cisco Gigabit Ethernet SFP Transceiver Modules mainly include 1000BASE-T (or GLC-T) SFP, 1000BASE-SX SFP, 1000BASE-LX/LH SFP, 1000BASE-EX SFP, 1000BASE-ZX SFP, 1000BASE-BX10-D and 1000BASE-BX10-U SFP.
1000BASE-T (or GLC-T) SFP: Cisco GLC-T is a copper wire SFP for 1000Base-T applications. The 1000BASE-T SFP operates on standard Category 5 unshielded twisted pair copper cabling of up to 100m length. Cisco 1000BASE-T SFP module supports 10/100/1000 auto-negotiation and Auto MDI/MDIX.
1000BASE-SX SFP: 1000Base-SX SFP supports Multimode fiber only. It operates on 50μm multimode fiber links up to 550m, up to 1km over laser-optimized 50μm multimode fiber cable and on 62.5μm Fiber Distributed Data Interface (FDDI)-grade multimode fibers up to 220m.
1000BASE-LX/LH SFP: 1000BASE-LX supports both Multimode and Single-mode fibers. 1000BASE-LX/LH SFP operates on standard single-mode fiber optic link spans of up to 10km and up to 550m on any multimode fibers.
1000BASE-EX SFP: The 1000BASE-EX SFP supports for Single-Mode Fibers. 1000BASE-EX SFP operates on standard single-mode fiber-optic link length up to 40 km. A 5-dB inline optical attenuator should be inserted between the fiber-optic cable and the receiving port on the SFP at each end of the link for back-to-back connectivity.
1000BASE-ZX SFP: The 1000BASE-ZX SFP operates on standard single-mode fiber-optic link spans of up to approximately 70 km in length. The SFP provides an optical link budget of 23 dB, but the precise link span length depends on multiple factors such as fiber quality, number of splices, and connectors.
1000BASE-BX10-D and 1000BASE-BX10-U SFP: The 1000BASE-BX-D & 1000BASE-BX-U SFPs used for Single-Fiber Bidirectional applications. A 1000BASE-BX10-D device is always connected to a 1000BASE-BX10-U device with a single strand of standard SMF with an operating transmission range up to 10km.
The communication over a single strand of fiber is achieved by separating the transmission wavelength of the two devices. 1000BASE-BX10-D transmits a 1490-nm channel and receives a 1310-nm signal, whereas 1000BASE-BX10-U transmits at a 1310nm wavelength and receives a 1490nm signal. Wavelength-division multiplexing (WDM) splitter integrated into the SFP to split the 1310nm and 1490nm light paths, up to 80km or 1000base BX-D BIDI tx1550 rx1490nm 80km in simple.

Will Wideband Multimode Fiber be a new option for future data centers?

Multimode fiber has evolved from OM1, OM2, OM3 to OM4 since entering the market in the 1980s. Among them, OM3 is for vertical cavity surface emitting laser (VCSEL) light source optimized multimode fiber, the effective mode bandwidth (EMB) reaches 2000MHZ. Km, which can support transmit distance up to 300 meters or 100 meters for 10GBASE-SR transceiver 100GBASE-SR10 transceiver respectively. OM4’s effective mode bandwidth (EMB) compared to OM3 is increased by more than 1 times to 4700MHZ.Km, but support 100GBase-SR10 transceiver is only 150 meters effective transmission distance, that is relative to the OM3 fiber, 100G Ethernet Network transmission distance increased by only 50%, obviously can not meet the needs of the future high-speed network.
As one of the options for future data centers, Wideband Multimode Fiber (WBMMF) is far superior to traditional OM4 multimode fiber in efficient mode bandwidth (EMB), so it can provide enough margin space for high-speed Ethernet in future. Before understanding it, we need to carefully understand the development of single mode and multimode fiber.
Single mode Fiber and Multimode Fiber, the Difference and Selection
Single mode fiber is mainly used in multi-frequency data transmission applications, the transmission mode is usually used wavelength division multiplexing (WDM) transmission mode, after the multiplexing of optical signals only need to use a single mode fiber to achieve data transmission. The 100GBASE-LR4 released in 2010, using 2-core Single mode Fiber (1 transmit +1 receive) 1, can be in one core fiber simultaneously multiplexing 4 wavelengths, each wavelength transmits at 25Gbps.
The transmission rate of single mode fiber is higher than that of multimode, and the transmission distance is 50 times higher than that of multimode, so its price is higher than MMF. Compared with MMF, the core diameter of single mode fiber is much smaller, and the characteristics of small core diameter and single mode transmission make the optical signal transmitted in SMF not distorted by the overlapping of light pulses. In all fiber types, the signal attenuation rate of single mode fiber is the lowest and the transmission speed is the most.
However, single mode fiber needs to use a high-cost laser (LD) Light transmitter transceiver, single mode laser transceiver Price is at least more than 3 times than the multimode transceiver, the power consumption of at least twice times.
Conventional multimode fiber generally uses serial transmission mode, in this mode, increasing the transmission rate of Ethernet must increase the transmission rate of each core fiber/channel. At present, the Ethernet maximum serial transmission rate is 10Gbps/channel, and IEEE is developing 25Gbps/Channel, 50Gbps/Channel Network standard. Take 400G Ethernet as an example, there will be 3 different versions like 25Gbps, 50Gbps, and 100Gbps, the fiber core number needs 32 cores/16 cores/8 cores respectively. 400G Ethernet uses NRZ, PAM4, DMT, and higher-level coding means more complex circuits and power consumption, and thus higher costs.
In the past practical applications, the most common determinant of selecting multimode or single mode is distance. If only child miles are preferred multimode because the LED transmitter/receiver is much cheaper than a single mode laser. If the distance is greater than 5 miles, the single mode fiber is best. Another issue to consider is bandwidth; if future applications may include the transmission of large-bandwidth data signals, the single mode will be the best choice.
In the 2010 years, with 100g-NG, 200G/400G Ethernet, and even 1T Ethernet, the traditional multimode fiber has become the bottleneck of the future Ethernet network, and the emergence of the WBMMF breaks the bottleneck of traditional multimode fiber. It draws on the wavelength division multiplexing (WDM) technology of single mode fiber, extends the range of available wavelengths in network transmission, and can support 4 wavelengths on one-core MMF, and reduces the number of fiber cores needed to 1/4.
Shortwave Wavelength Division Multiplexing technology utilizes a high cost-effective vertical cavity surface emitting laser (VCSEL) light source, the optimized wideband multimode fiber (WBMMF) can support 4 wavelengths on one-core multimode fiber, reduce the number of fiber cores required to 1/4, and increase the effective mode bandwidth (EMB), so the 40 / 100G transmission distance extended to 300 meters.
Conclusion
At present, 96% of the world’s data centers, the backbone of the network core (Spine) switch to the Server cabinet Branch (Leaf) switch within 300 meters, therefore, shortwave wavelength division multiplexing (SWDM) and wideband multimode fiber (WBMMF) will continue to be the tradition of multimode fiber as the mainstream transmission medium of data center 40G/100G/400G Ethernet. In the future, the combination of shortwave wavelength division multiplexing (SWDM) and parallel transmission technology requires only 8-core wideband multimode fiber (WBMMF) to support more high-speed applications such as 200G/400G Ethernet.

10GBASE-T SFP or 10G Fiber SFP, which one will you choose?

Now big data become IT industry hot words, it gives 10 Gigabit transmission popularity to bring a strong driving force. For a data center switch port, both of 10GBASE-SR Fiber SFP and 10GBASE-T SFP+ can meet your need, but which one will you choose?
What is 10GBASE-SR?
“SR” in 10GBase-SR is an abbreviation for “short range”, which is used only for short distance connections. The specification supports 64B / 66B shortwave (wavelength 850nm) multimode fiber (MMF), the effective transmission distance of 2 meters to 300 meters. However, it is necessary to use a laser-optimized 50 μm line OM3 fiber to reach the maximum distance.
What is 10GBASE-T?
10GBASE-T is an Ethernet specification using a copper connection (Category 6 shielded or unshielded twisted pair) with an effective bandwidth of 10 Gbit / s and a maximum transmission distance of up to 100 meters. The IEEE standard corresponding to 10GBASE-T is 802.3an-2006. This can be regarded as a revolutionary advance in 10 Gigabit Ethernet because until then, it has been thought that it is impossible to achieve such a high transmission rate on a twisted pair, because the loss is too large to run at such a high operating frequency (at least 500MHz). But standard builders rely on the following 4 technical artifacts to make 10GBASE-T a reality: Loss elimination, analog to digital conversions, cable enhancements, and coding improvements.
In coding, 10GBase-T is not the original 1000Base-T PAM-5, but the use of PAM-8 encoding, support 833Mbps, and 400MHz bandwidth, the cabling system bandwidth requirements are also modified to 500MHz, and if still The use of PAM-5’s 10GBase-T requires a routing bandwidth of 625MHz. On the connector side, the 10GBase-T uses a 650 MHz version of the RJ-45 connector that has been widely used in Ethernet. The longest effective transmission distance on Category 6 is 55 meters and can reach 100 meters on Category 6a.
10GBASE-SR Fiber SFP+ VS 10GBASE-T SFP+, which one should you choose?
The 10GBase-T system is the first choice for the 10G network system. 10GBASE-SR fiber with the similar cabling system construction process complexity (cost) approach like 10GBASE-T, but 10GBase-SR fiber system (including fiber switch, optical transceiver, fiber-optic network card) will cost about 3 times or more than 10GBASE-T system. The comprehensive point of view, 10 Gigabit copper with cheap, good compatibility, easy construction and so on. At the same time, Base-T usually runs up to 100 meters twisted pair, such as the use of Category 6A and 7 cables. 10GBase-T technology supports backward compatibility and is automatically negotiated between higher and lower rates, eliminating the need for a one-time upgrade to the network.
For the transmission distance, optical fiber transmission due to anti-interference ability is undoubted to achieve the park million network backbone of the ideal solution. In other words, more 10G optical fiber network is used for vertical systems, such as a unit of various branches, or the school has each dormitory is connected to the network center, the distance is likely to be more than 100 meters or even 1-2 kilometers. But in a data center or unit of the horizontal range, the radius of 100 meters range, 10G Base-T has more advantages. Therefore, we can’t simply say the fiber 10G and 10G Base-T who is faster who slower or who replace the another, the difference is that the data transmission distance and anti-jamming performance, different application scenarios determine their value.
Glossary
1000BASE-T: 1000BASE-T uses a unshielded twisted pair as the transmission medium to transmit the longest distance is 100 meters. 1000BASE-T does not support 8B / 10B encoding, but the use of more complex coding. 1000BASE-T has the advantage of the user can be 100BASE-T on the basis of a smooth upgrade to 1000BASE-T.
25GBase-T: 25G Ethernet is also based on copper cable transmission, and also with Cat. 8 class copper cable twisted pair is medium.
5GBase-T: 5 Gbit/s up to at least 100 m of Cat 6.
2.5GBase-T: 2.5 Gbit/s up to at least 100 m of Cat 5e.
40GBase-T: use bandwidth spectrum from 1 MHz to approximately 1,600 MHz., maximum length of cabling will be limited to around 30 meters over Category 8.

What is the market demand for the Data Center Optical transceiver?

High-speed optical transceiver market began to grow at the end of 2016, its core driving force is the international large/ultra large data center upgrades and new demands on 40G and 100G transceiver.
As the core device to achieve photoelectric conversion, the Optical transceiver module is widely used in the data center. In the scene of 10G above rate and 10 meters above transmission distance, the use of optical transceiver (or active optical cable-AOC that is packaged with multimode fiber and VCSEL optical transceiver module) is the only mature cabling scheme. High-definition video, live, VR and other new applications to promote global network traffic growth, cloud computing, IaaS services, large data and other emerging application requirements for data center internal data transmission higher requirements. Therefore, the data center needs not only continuously expands, new deployment, but also needs to evolve continuously in the network architecture to achieve the Non-blocking network performance.
In the process of data Center architecture evolution, the demand for the Optical transceiver has undergone new changes. Through the analysis of the traditional three-layer architecture, the improved three-layer architecture and the new two-layer architecture, it can be seen that the number of optical transceivers that the single cabinet needs to configure has increased significantly with the continuous evolution of the network architecture.
The traditional three-layer architecture, the number of Optical transceiver module is about 8.8 times the number of cabinets (8 unit 40G Optical transceiver module, 0.8 unit 100G Optical transceiver module);
The improved three-layer architecture, the number of Optical transceiver module is about 9.2 times the number of cabinets (8 unit  40G Optical transceiver module, 1.2 unit 100G Optical transceiver module);
The number of optical transceivers in the emerging two-tier architecture is about 44 or 48 times the number of cabinets (80-90% of which are 10G optical transceivers, 8 unit 40G transceivers or 4 unit 100G transceivers); let us highlight the high-speed optical transceiver demand, the international market due to cloud computing and other downstream applications of high-speed development, third-party data centers and cloud computing manufacturers self-built data center rapid development, network architecture to the continuous evolution of flat, high-speed optical transceiver module produce strong demand. In contrast, the Chinese market due to the downstream application is still in the early stages of development, network traffic pressure and data center data transmission within the lower demand, China’s high-speed optical transceiver market data still need to wait for the downstream application of manure.
The flattening of data center mainly solves the data flow in the data center, and the data is mainly driven by cloud computing, cloud Migration, IaaS virtual configuration, distributed computing and large data (data volume, storage, and different servers). The international market AWS, Google, such as cloud platforms, Salesforce, and other professional platforms, as well as distributed computing and large-data applications of high-speed development to promote the flattening of the data center, high-speed optical transceiver module generated strong demand. In contrast, the downstream application of China is still in the stage of market cultivation and business model exploration and the market of China High-Speed Digital transceiver is expected to enter the growth period after the downstream application enters the mature stage.

Industrial – High and Low Temperature XFP Transceiver

Industrial XFP transceiver, which is used in the field of industrial control Ethernet switch equipment, due to the use of network standards open and good, widely used; can adapt to low temperature, anti-electromagnetic interference, anti-salt spray, strong earthquake resistance. Using a transparent and unified TCP / IP protocol, Ethernet has become the main communication standard in the field of industrial control. Mainly used in industrial control automation, road traffic control automation, mine automatic control system, oil field control automation, hydropower station control automation, power system control automation. XFP transceiver is more suitable for “civilian”, more common, common. In the harsh environment, all aspects of resistance than industrial grade.
The fiber-mart.com series is a green industrial XFP transceiver that supports two Ethernet ports and one optical port. The RS Series offers 12/24/48 VDC redundant power input with a rugged IP40 aluminum housing and supports rail mounting and high EMI / EMC protection, working over a wide temperature range of -40 to 85 ° C and via 100 % Of the aging test to ensure quality. The RS series offers these features to provide a cost-effective industrial solution.
● Features
12/24/48 VDC dual redundant power supply
Jumbo frame transfer rate up to 10KB
Power or port failure can be triggered by relay output
Broadcast storm protection
Operating temperature – 40 ~
PoE function is optional

What are the XFP transceiver connection methods?

Before purchasing XFP transceiver, understand the XFP transceiver connection method, more convenient for users according to their own needs to set up the network to meet the needs of user transmission. XFP transceiver as a signal conversion transmission equipment, the front of the Ethernet signal, through the XFP transceiver transmitter to the Ethernet electrical signal converter for long-distance transmission of optical signals, XFP transceiver will receive the optical signal Electrical signal. The product is generally used in Ethernet networks that can not be covered and must be used to extend the transmission distance of the actual network environment, and are usually located in the broadband layer network access layer applications. Fiberland based on the customer in the process, summed up the 3-point XFP transceiver link method:
First, the ring backbone network
The backbone backbone network is constructed by using SPANNING TREE to construct the backbone of the metro area. This structure can be deformed into a mesh structure, which is suitable for the high density central area on the metropolitan area network, forming a fault-tolerant core backbone network. Ring backbone network support for IEEE.1Q and ISL network features can guarantee compatibility with most mainstream backbone networks, such as cross-switch VLAN, TRUNK and other functions. Ring backbone network for the financial, government, education and other industries set up broadband virtual private network.
Second, the chain backbone network
The chain-type backbone network can save a large number of backbone light quantity by chain connection, which is suitable for constructing high-bandwidth and low-cost backbone network at the edge of the city and the suburbs. The model can also be used for highway, oil and transmission Lines and other environments. Chain backbone network IEEE802.1Q and ISL network characteristics of the support, can guarantee compatible with the vast majority of the backbone network, for the financial, government, education and other industries set up broadband virtual private network. Chain backbone network is able to provide image, voice, data and real-time monitoring of integrated transmission of multimedia networks.
Third, the user access system
User access system using 10Mbps / 100Mbps adaptive and 10Mbps / 100Mbps automatic conversion function, you can connect any client device, no need to prepare a variety of XFP transceiver, the network can provide a smooth upgrade program. At the same time the use of half-duplex / full-duplex adaptive and half-duplex / full-duplex automatic conversion function, you can configure the user inexpensive half-duplex HUB, several times to reduce the cost of the client network, improve network operators The competitiveness of the. At the same time, the built-in switching core of the equipment improves the transmission efficiency of the access equipment, reduces the network broadcasting, controls the flow rate, and detects the transmission failure.
Users know how to buy XFP transceiver before the purchase method, according to the use of the scene to buy equipment, play a multiplier effect, but also save a lot of construction and installation time.