Demartek Storage Networking Interface Comparison
Updated 7 October 2016
By Dennis Martin, Demartek President
Interactive PDF Version
Recently, Demartek updated the Storage Networking Interface Comparison and included a new Interactive PDF version. With this format you are capable of viewing and interacting with the information in ways not available with standard PDF readers. To get the most out of this guide, we encourage you to upgrade to the latest version of the Adobe Acrobat Reader. If you experience any problems viewing or interacting with this document, please contact us regarding the problem.
Download the Demartek Storage Networking Interface Comparison Interactive PDF
(PDF, 8.6 MB)
(If using Internet Explorer, Firefox, Safari or Opera,
right-click, save the file and open in your PDF reader of choice.)
Because of the number of storage interface types and related technologies that are used for storage devices, we have compiled this summary document providing some basic information for each of the interfaces. This document will be updated periodically. This document may become larger over time. Contact us if you’d like to see additional information in this document.
The interface types listed here are known as “block” interfaces, meaning that they provide an interface for “block” reads and writes. They simply provide a conduit for blocks of data to be read and written, without regard to file systems, file names or any other knowledge of the data in the blocks. The host requesting the block access provides a starting address and number of blocks to read or write.
We are producing deployment guides for some of the technologies described in this document.
- Demartek Fibre Channel Deployment Guide
- Demartek iSCSI Deployment Guide
- Demartek SSD Deployment Guide
To be notified when we post new deployment guides, evaluation reports and commentaries, sign-up for our free monthly newsletter, Demartek Lab Notes. We do not give out, rent or sell our email list.
- Storage Networking Interface Comparison Table
- Transfer Rate, Bits vs. Bytes, and Encoding Schemes
- Cables: Fiber Optics and Copper
- Connector Types
- PCI Express® (PCIe®)
- FC — Fibre Channel (also see Demartek FC Zone)
- FCoE — Fibre Channel over Ethernet (also see Demartek FCoE Zone)
- IB — InfiniBand
- iSCSI — Internet Small Computer System Interface (also see Demartek iSCSI Zone)
- NVMe — NVM Express
- PCIe — PCI Express
- SAS — Serial Attached SCSI
- SATA — Serial ATA
- USB — Universal Serial Bus
- 10GbE — 10 Gigabit Ethernet
- CNA — Converged Network Adapter (used with FCoE)
- HBA — Host Bus Adapter (used with FC, iSCSI, SAS, SATA)
- HCA — Host Channel Adapter (used with IB)
- NIC — Network Interface Controller or Network Interface Card (used with FCoE, iSCSI)
- ISL — Inter-Switch Link
- SAN — Storage Area Network
- Gb — Gigabit
- GB — Gigabyte
- Mb — Megabit
- MB — Megabyte
- Tb — Terabit
- TB — Terabyte
- Gb/s — Gigabits per second
- Gbit/s — Gigabits per second
- Gbps — Gigabits per second
- GB/s — Gigabytes per second
- GBps — Gigabytes per second
- Mb/s — Megabits per second
- Mbit/s — Megabits per second
- Mbps — Megabits per second
- MB/s — Megabytes per second
- MBps — Megabytes per second
- Tb/s — Terabits per second
- Tbit/s — Terabits per second
- Tbps — Terabits per second
- TB/s — Terabytes per second
- TBps — Terabytes per second
- HDD — Hard Disk Drive
- SSD — Solid State Drive (also see Demartek SSD Zone)
- SSHD — Solid State Hybrid Drive
- SDR — Single Data Rate (InfiniBand)
- DDR — Double Data Rate (InfiniBand)
- QDR — Quad Data Rate (InfiniBand)
- FDR — Fourteen Data Rate (InfiniBand)
- EDR — Enhanced Data Rate (InfiniBand)
PCIe data rates are provided in the PCI Express section below.
Transfer rate, sometimes known as transfer speed, is the maximum rate at which data can be transferred across the interface. This is not to be confused with the transfer rate of individual devices that may be connected to this interface. Some interfaces may not be able to transfer data at the maximum possible transfer rate due to processing overhead inherent with that interface or the protocol used on that interface. Some interface adapters provide hardware offload functions to improve performance, manageability and/or reliability of the data transmission across the respective interface. The transfer rates listed are across a single port at half duplex (from point A to point B, one direction at a time).
Bits vs. Bytes and Encoding Schemes
Transfer rates for storage interfaces and devices are generally listed as MB/sec or MBps (MegaBytes per second), which is generally calculated as Megabits per second (Mbps) divided by 10. Many of these interfaces use “8b/10b” encoding which maps 8 bit bytes into 10 bit symbols for transmission on the wire, with the extra bits used for command and control purposes. When converting from bits to bytes on the interface, dividing by ten (10) is exactly correct. 8b/10b encoding results in a 20 percent overhead (10-8)/10 on the raw bit rate.
Beginning in 2010 newer encoding schemes emerged that improve data transfer efficiency. The first of these newer encoding schemes is known as “64b/66b” and is used for 10GbE, 10GbFC (for ISL’s) and 16Gb FC and some of the higher data rates for IB. 64b/66b encoding is not directly compatible with 8b/10b, but the technologies that implement it are built so that they can work with the older encoding scheme. 16Gb Fibre Channel uses a line rate of 14.025 Gbps, but with the 64b/66b encoding scheme results in a doubling of the throughput of 8Gb Fibre Channel, which uses a line rate of 8.5 Gbps with the 8b/10b encoding scheme. 64b/66b encoding results in a 3 percent overhead (66-64)/66 on the raw bit rate.
PCIe versions 1.x and 2.x use 8b/10b encoding. PCIe version 3 uses 128b/130b encoding, resulting in a 1.5 percent overhead on the raw bit rate. Additional PCIe information is provided in the PCI Express section below.
USB 3.1 Gen 2 (10 Gbps) uses use 128b/132b encoding. See Roadmaps section below.
|8b/10b||20%||1GbE, FC (up to 8Gb), IB (SDR, DDR & QDR), PCIe (1.0 & 2.0) SAS, SATA, USB (up to 3.0)|
|64b/66b||3%||10GbE, 100GbE, FC (10Gb, 16Gb & 32Gb), FCoE, IB (FDR & EDR), Thunderbolt 2|
|128b/130b||1.5%||PCIe 3.0, 24Gb SAS (likely)|
|128b/132b||3%||USB 3.1 Gen 2 (10 Gbps, see Roadmaps section below)|
Fibre Channel speeds are often abbreviated into the form nnGFC, indicating nn Gbps Fibre Channel.
InfiniBand connections can be aggregated into 4x (4 lanes) and 12x (12 lanes), depending on the application and connector. QSFP and QSFP+ connectors are used for 4x connections, and CXP connectors are typically used for 12x connections. See the Connector Types section below for more details on the connector types. Newer IB speeds have been announced that are described in the Roadmaps section below.
Products became available with the interface speeds listed during these years. Newer interface speeds are often
available in switches and adapters long before they are available in storage devices and storage systems.
All the interface speeds listed are single direction (from point A to point B). Some of these interfaces can also operate in bi-directional or full-duplex mode (both directions simultaneously). For bi-directional rates, double the data rates.
- FC — 1 Gb/s in 1997, 2 Gb/s in 2001, 4 Gb/s in 2005, 8 Gb/s in 2008, 10 Gb/s (ISL only) 2004, 16 Gb/s in 2011, 32 Gb/s in 2016 (FC is generally backward compatible with the previous two generations)
- FCoE — FC: 4 Gb/s and Ethernet: 10 Gb/s in 2008, 10 Gb/s in 2009.
(FC-BB-5 was approved in June 2009, INCITS 462-2010 was approved in Spring 2010)
- IB — 10 Gb/s in 2002, 20 Gb/s in 2005, 40 Gb/s in 2008, 56 Gb/s in 2011, 100 Gb/s in 2015.
- iSCSI — 1 Gb/s in 2003, 10 Gb/s in 2007 (basic 10GbE first appeared in 2002)
- NVMe — Version 1.0 specification published in March 2011. Version 1.1 specification finalized in October 2012. Version 1.2 specification released in November 2014.
- SAS — 3 Gb/s in 2005, 6 Gb/s in 2009, 12 Gb/s in 2H 2013
- SATA — 1.5 Gb/s in 2003, 3 Gb/s in 2005, 6 Gb/s in 2010 (traditional SATA is not expected to extend beyond 6 Gb/s, see Roadmaps section below)
- SATA μSSD was introduced in August 2011 as a new implementation of SATA for embedded SSDs. These devices do not have the traditional SATA interface connector but use a single ball grid array (BGA) package that can be surface mounted directly on a system motherboard. These SATA μSSD devices are intended for mobile platforms such as tablets and ultrabooks, and consume less electric power than traditional SATA interface devices.
- SATA Revision 3.2 has been ratified and the announcement was made in August 2013. This revision includes new SATA form factors, details for SATA Express, power management enhancements and optimizations for solid state hybrid drives (SSHDs). One of the new SATA form factors is M.2 that enables small form-factor SATA SSDs suitable for thin devices such as tablets and notebooks. M.2 was formerly known as NGFF, is defined by the PCI-SIG and supports a variety of applications including WiFi, WWAN, USB, PCIe and SATA. The v3.2 specification standardizes the SATA M.2 connector pin layout. SATA v3.2 also introduces USM Slim, which reduces the thickness of the module from 14.5mm to 9mm.
- Thunderbolt — 10 Gb/s in 2011, 20 Gb/s in late 2013, 40 Gb/s in 2015
- USB — 1.5 Mb/s in 1997, 12 Mb/s in 1999, 480 Mb/s in 2001, 5 Gb/s in 2009, 10 Gb/s specification completed in 2014
PCIe history is provided in the PCI Express section below.
These roadmaps include the estimated calendar years that higher speeds may become available and are based on our industry
research, which are subject to change. Past history indicates that several of these interfaces are on a three to five
year development cycle for the next improvement in speed. It is reasonable to expect that pace to continue.
It should be noted that it typically takes several months after the specification is complete before products are generally available in the marketplace. Widespread adoption of those new products takes additional time, sometimes years.
Some of the standards groups are now working on “Energy Efficient” versions of these interfaces to indicate additions to their respective standards to reduce power consumption.
All the interface speeds listed are single direction (from point A to point B). Some of these interfaces can also operate in bi-directional or full-duplex mode (both directions simultaneously). For bi-directional rates, double the data rates.
See the Connector Types section below for additional roadmap information.
Roadmap slides are courtesy of the respective industry organizations.
- In 2015, and again in 2016, the Ethernet Alliance published its roadmap for new speeds of Ethernet. In addition to 25 Gbps, development of 2.5 Gbps and 5 Gbps speeds are underway that will support faster speeds of WiFi. Faster speeds, in single-lane and multi-lane configurations, are also in development. All of these new speeds have obvious implications for storage applications such as FCoE and iSCSI block protocols as well as file and object storage protocols. The Connector Types section below has additional information on single-lane and multi-lane connection technology.
- Fibre Channel (FC)
Fibre Channel Roadmap
- 64GFC — Development work has begun on 64GFC and 256GFC in FC-PI-7, with technical stability expected in 2017. Both the single-lane (64GFC) and quad-lane (256GFC) specifications will be included in one standard. Each FC revision is expected to be backwards compatible with at least the two previous generations.
- SAN interface — FC has a future as a SAN interface for the foreseeable future. There has been a huge investment (US$ Billions) in FC infrastructure over the years, primarily in enterprise datacenters, which is likely to remain deployed for many years. According to various industry analysts, Fibre Channel continues to be the most popular host interface for external storage systems. Some newer storage vendors that did not originally include FC as a host interface have plans to add it.
- Disk drive interface — FC has reached end-of-life as a disk drive interface, as the disk drive and SSD manufacturers have largely moved to SAS and NVMe as the primary device interfaces for enterprise drives.
- FC-NVMe — Development of drivers to support FC-NVMe in existing FC host adapters is well underway. When completed, Fibre Channel host adapters will support traditional Fibre Channel and FC-NVMe simultaneously. Fibre Channel switches require no changes to support FC-NVMe.
- Roadmap diagram — FCoE appears on the Fibre Channel roadmap diagram above
- 100 Gb/s — 100 Gb/s FCoE support is in development. Also see the Ethernet roadmap above. See the connector types section below for discussion on connectors for various speeds.
- InfiniBand (IB)
- InfiniBand Routing — IB routing has been in the specification for some time, but the first IB routers are expected to come to market in 2016. These routers are expected to support various subnet topologies.
- InfiniBand Virtualization — IB virtualization support is in development to support virtualized workloads.
- HDR (High Data Rate) 200 Gb/s (4 x 50 Gb/s) — HDR IB is expected to become available in 2H 2017. 200 Gb/s adapters will require either PCIe 3.0 x32 or PCIe 4.0 x16 slots in the servers to support adapters running at the full data rate.
- HDR Copper Cables — Copper cables are expected to be supported at HDR (200 Gb/s), for distances up to at least 3 meters (50 Gb/s per lane).
- Beyond HDR — NDR and XDR have been announced, but as of September 2016, the release dates and specific data rates have not been determined.
- NDR — Next Data Rate is expected after HDR.
- XDR — eXtreme Data Rate is expected after NDR.
- iSCSI — follows Ethernet roadmap.
- NVMe 1.3 — NVMe 1.3 is expected to be completed by year-end 2016. The features expected are sanitize, virtualization, directives, thin provisioning, host controlled thermal management and others. Version 1.3 is also expected to support mobile devices.
- NVMe-MI 1.1 — Development is underway on the next version of the NVMe Management Interface specification, version 1.1, which will focus on enclosure management.
- NVMe over Fabrics — Version 1.0 of the NVMe over Fabrics specification was published in June 2016. Prototype products were shown at various industry events during the summer of 2016. NVMe over Fabrics products may appear on the market by the end of 2016.
- 24 Gb/s SAS specifications — Development is currently underway on the 24 Gb/s SAS specification, known as SAS-4 by the INCITS T10 committee. A SAS-4 working draft revision 0.7 was created in July 2016. The SAS-4 specification is expected to enter Letter Ballot Review at the September 2016 T10 meetings and is expected to exit Letter Ballot Review in January 2017. Letter Ballot complete implies that the specification is functionally complete and silicon will be taped out to that revision of the specification. SPL-4 is the protocol layer specification and is expected to exit Letter Ballot Review at the September 2016 T10 meetings. 24 Gb/s SAS is expected to be backward compatible with 12 Gb/s and 6 Gb/s SAS but may use a different encoding scheme than previous versions.
- 24 Gb/s SAS products — The first 24 Gb/s SAS plugfest is expected to occur in 2H 2017. The development and prototypes for 24 Gb/s SAS will use PCIe 3.x technology, but it is likely that the 24 Gb/s SAS final products will be aligned with the server platforms that use PCIe 4.0 technology, with 24 Gb/s SAS products available in approximately 2019. See the PCI Express section below for information on PCIe 4.0. After the official release of 24 Gb/s SAS products, it is expected that work will be done to reduce latency for SAS.
- SCSI Express — SCSI Express was an effort to place the SCSI protocol over the PCI Express interface and includes SCSI over PCIe (SOP) and PCIe queueing interface (PQI). SCSI Express was first submitted as a proposal in 2011 and accepted as a formal project in 2012, but was de-emphasized during 2015. Although the SCSI Express specification was completed in 2013, as of mid-2016 there has been little or no industry adoption, work on SCSI Express development has stopped and no SCSI Express products are expected to become available.
- SAS Advanced Connectivity — New SAS cabling options are offering longer distances by using active copper (powered signal) and optical cables. The Mini SAS HD connector can be used for 6 Gb/s SAS and is used for 12 Gb/s SAS connections. See the connector types section below for discussion of Mini SAS and Mini SAS HD connectors. For both 12 Gb/s SAS and 24 Gb/s SAS, external passive copper cables are expected to support up to 6 meters, active copper up to 20 meters and optical cables at least 100 meters. Internal cables are expected to support 1 meter for speeds up to 24 Gb/s SAS. SAS Device to Mid-plane interconnects are expected to support 1-port, 2-port and 4-port connections up to 96 Gb/s (4 x 24 Gb/s) at half-duplex.
- SAS Advanced Features — New features emerging in 12 Gb/s SAS include zoned block commands (ZBC) to support shingled magnetic recording (SMR) which will drive HDD capacities higher and storage intelligence (hints) for optimized SSD performance.
- SATA Express — SATA Express is included in SATA Revision 3.2 (see history section above). SATA Express enables client SATA and PCI Express (PCIe) solutions to coexist. SATA Express will support increased interface speeds up to 2 lanes of PCIe (2 Gb/s for PCIe 3.0, 1 Gb/s for PCIe 2.0), compared with current 0.6 Gb/s (6 Gb/s) SATA technology. While SATA Express is supported on a few motherboards as of mid-2015, few, if any, devices are expected to support SATA Express anytime soon. It is unclear if SATA Express will gain acceptance in the market.
- SATA Enterprise Features — New features expected in the future include enterprise features such as remote power disable, improved rebuilds for arrays and optimizations for devices containing NAND flash. Support for shingled magnetic recording for higher capacity disk drives is also planned.
- Thunderbolt 3 (40 Gb/s) — Thunderbolt 3 was announced in June 2015. Thunderbolt 3 supports speeds up to 40 Gb/s with active copper or optical cables, or 20 Gb/s with passive copper cables. Thunderbolt 3 supports up to four lanes of PCI Express 3.0.
- USB data rates — The USB 3.0 Promoter Group announced at the end of July 2013 that the USB 3.1 specification had been completed. USB 3.1 enables USB to operate at 10 Gbps, and is backward compatible with existing USB 3.0 and 2.0 hubs and devices. USB 3.1 uses a 128b/132b encoding scheme, with four bits used for command and control for the protocol and the cable management. The first public prototype demonstration of USB running at 10 Gbps was made in September 2013. USB 3.1 is expected to support USB Power Delivery (described below). In the summer of 2015, the USB-IF renamed the 5 Gbps and 10 Gbps USB speeds to SuperSpeed USB (USB 3.1 Gen 1, 5 Gbps) and SuperSpeed USB 10 Gbps (USB 3.1 Gen 2).
- USB Power Delivery — USB is becoming a power delivery interface, with an increasing number of devices charging or receiving power via USB ports in computers or wall sockets and power strips. The USB Power Delivery (PD) Specification, version 1.0, was introduced in July 2012 to allow an increased amount of power to be carried via USB. This specification proposes to raise the limit from 7.5 watts up to 100 watts of power, depending on cable and connector types. Devices negotiate with each other to determine voltage and current levels for the power transmission, and power can flow in either direction. Devices can adjust their power charging rates while transmitting data. Prototypes began to appear in late 2013.
- USB Type-C Cable — The specification for the new Type-C USB cable and connector was completed in August 2014. This USB cable has an entirely new design with a smaller connector size that can be easily used with a variety of types of devices. With this new specification, the connector and cable are reversible with respect to the plug orientation and the direction of the cable. Type-C USB cables will have the same type of connector on each end of the cable, so it will not matter which end of the cable is plugged into a computer, hub, charger or other device. The Type-C USB cables are electronically marked so that cable information can be passed to the device. Initial cable lengths will be up to 1m in length and use passive copper technology. Active copper and optical cables could be produced in the future. The USB Type-C cable is also used in Thunderbolt 3 applications (see Thunderbolt above) that supports USB, Thunderbolt and DisplayPort protocols.
SAS-SATA Connector Compatibility
Source: SCSI Trade Association
Express Bay Connector Backplane
Source: SCSI Trade Association
SATA Express Connector Mating Matrix
Cables: Fiber Optics and Copper
As interface speeds increase, expect increased usage of fiber optic cables and connectors for most interfaces. At higher Gigabit speeds (10Gb+), copper cables and interconnects generally have too much amplitude loss except for short distances, such as within a rack or to a nearby rack. This amplitude loss is sometimes called a poor signal-to-noise ratio or simply “too noisy”.
Fiber-optic cables: Single-mode fiber vs. Multi-mode fiber
There are two general types of fiber optic cables available: single-mode fiber and multi-mode fiber.
- Single-mode fiber (SMF), typically with an optical core of approximately 9 µm (microns), has lower modal dispersion than multi-mode fiber and can support distances up to 80-100 Km (Kilometers) or more, depending on transmission speed, transceivers and the buffer credits allocated in the switches.
- Multi-mode fiber (MMF), with optical core of either 50 µm or 62.5 µm, supports distances up to 600 meters, depending on transmission speeds and transceivers.
When planning datacenter cabling requirements, be sure to consider that a service life of 15 to 20 years can be expected for fiber optic cabling, so the choices made today need to support legacy, current and emerging data rates. Also note that deploying large amounts of new cable in a datacenter can be labor-intensive, especially in existing environments.
There are different designations for fiber optic cables depending on the bandwidth supported.
- Multi-mode: OM1, OM2, OM3, OM4
- Single-mode: OS1 is for indoor use, OS2 is for outdoor use.
OM3 and OM4 are newer multi-mode cables that are “laser optimized” (LOMMF) and support 10 Gigabit Ethernet and faster applications. OM3 and OM4 cables are also the only multi-mode fibers included in the IEEE 802.3ba 40G/100G Ethernet standard that was ratified in June 2010. The 40G and 100G speeds are currently achieved by bundling multiple channels together in parallel with special multi-channel (or multi-lane) connector types. This standard defines an expected operating range of up to 100m for OM3 and up to 150m for OM4 for 40 Gigabit Ethernet and 100 Gigabit Ethernet, using 10 Gbps per lane technology. See the Connector Types section below for additional detail.
Newer multi-mode OM2, OM3 and OM4 (50 µm) and single-mode OS1 (9 µm) fiber optic cables have been introduced that can handle tight corners and turns. These are known as “bend optimized,” “bend insensitive,” or have “enhanced bend performance.” These fiber optic cables can have a very small turn or bend radius with minimal signal loss or “bending loss.” The term “bend optimized” multi-mode fiber (BOMMF) is sometimes used.
OS1 and OS2 single-mode fiber optics are used for long distances, up to 10,000m (6.2 miles) with the standard transceivers and have been known to work at much longer distances with special transceivers and switching infrastructure.
Each of the multi-mode and single-mode fiber optic cable types includes two wavelengths. The higher wavelengths are used for longer-distance connections.
Update: 24 April 2012 — The Telecommunications Industry Association (TIA) Engineering Committee TR-42 Telecommunications Cabling Systems has approved the publication of TIA-942-A, the revised Telecommunications Infrastructure Standard for Data Centers. A number of changes were made to update the specification with respect to higher transmission speeds, energy efficiency and harmonizing with international standards. For backbone and horizontal cabling and connectors, the following are some of the important updates:
- Copper cabling — Cat 6 is the minimum requirement, Cat 6a recommended
- Fiber optic cabling — OM3 is the minimum requirement, OM4 is recommended
- Fiber optic connectors — LC is the standard for one or two fiber connectors
The TIA revises its standards approximately every five years and work is underway on TIA-942-B. Possible updates include Category 8 cabling that is designed to support 25GBASE-T and 40GBASE-T and wideband multimode fiber (WBMMF) that can use multiple wavelengths such as 850nm, 880nm, 910nm and 940nm.
10Gb Ethernet Fiber-Optic Cables
- 10GBASE-SR — Currently, the most common type of fiber-optic 10GbE cable is the 10GBASE-SR cable that supports an SFP+ connector with an optical transceiver rated for 10Gb transmission speed. These are also known as “short reach” fiber-optic cables.
- 10GBASE-LR — These are the “long reach” fiber optic cables that support single-mode fiber optic cables and connectors.
Indoor vs. Outdoor cabling
Indoor fiber-optic cables are suitable for indoor building applications. Outdoor cables, also known as outside plant or OSP, are suitable for outdoor applications and are water (liquid and frozen) and ultra-violet resistant. Indoor/outdoor cables provide the protections of outdoor cables with a fire-retardant jacket that allows deployment of these cables inside the building entrance beyond the OSP maximum distance, which can reduce the number of transition splices and connections needed.
¹ OM1 cable is not recommended for 16GFC, but is expected to operate up to 15m.
Distances supported in actual configurations are generally less than the distance supported by the raw fiber optic cable. The distances shown above are for 850 nm wavelength multi-mode cables. The 1300 nm wavelength multi-mode cables can support longer distances.
Copper cables: Active Copper vs. Passive Copper
Copper cables are available in passive and active designs. Passive copper cables consume no power and have shorter reach. Active copper cables include components that boost the signal, reduce the noise and work with smaller-gauge cables, improving signal distance, cable flexibility and airflow. These active copper cables consume some power and are more expensive than passive copper cables. From a distance perspective, active copper cables reach longer than passive copper but shorter than fiber-optic cables. The copper cables, both passive and active, typically include a connector or transceiver mounted directly on the cable.
Passive copper cables are very common in datacenters and are frequently used in configurations with top-of-rack (“TOR”) switches because even at the higher speeds, passive copper cables can usually reach the full height of a rack. Many newer datacenters are designed in such a way to keep cable lengths short to take advantage of lower-cost passive copper cables rather than use more expensive active copper cables.
The newer speeds of some interfaces are expected to increase the use of active copper cables over time.
Copper cables: Distances
Passive copper cables, depending on the rated speed, typically are in the single-digit meters in terms of distance. Active copper cables can typically reach into the low double-digits of meters. As the data rates increase, the distance decreases. With single-lane data rates of 25 Gbps and 32 Gbps, copper cables have a relatively short reach.
Research is being conducted to determine the feasibility of copper cables at 100 Gbps per lane. One example of this was shown at DesignCon 2016. An incubation startup company demonstrated a 100 Gbps serial data communication over a 1.5 meter twinaxial copper cable. Additional information is available at www.bifast.io.
Copper: 10GBASE-T and 1000BASE-T
1000BASE-T cabling is commonly used for 1Gb Ethernet traffic in general, and 1Gb iSCSI for storage connections. This is the familiar four pair copper cable with the RJ45 connectors. Cables used for 1000BASE-T are known as Cat5e (Category 5 enhanced) or Cat6 (Category 6) cables.
10GBASE-T cabling supports 10Gb Ethernet traffic, including 10Gb iSCSI storage traffic. The cables and connectors are similar to, but not the same as the cables used for 1000BASE-T. 10GBASE-T cables are Cat6a (Category 6 augmented), also known as Class EA cables. These support the higher frequencies required for 10Gb transmission up to 100 meters (330 feet). Cables must be certified to at least 500MHz to ensure 10GBASE-T compliance. Cat7 (Category 7, Class F) cable is also certified for 10GBASE-T compliance, and is typically deployed in Europe. Cat6 cables may work in 10GBASE-T deployments up to 55m, but should be tested first. 10GBASE-T cabling is not expected to be deployed for FCoE applications in the near future. Some newer 10GbE switches support 10GBASE-T (RJ45) connectors.
10GBASE-CR — Currently, the most common type of copper 10GbE cable is the 10GBASE-CR cable that uses an attached SFP+ connector, also known as a Direct Attach Copper (DAC). This fits into the same form factor connector and housing as the fiber optic cables with SFP+ connectors. Many 10GbE switches accept cables with SFP+ connectors, which support both copper and fiber optic cables. These cables are available in 1m, 3m, 5m, 7m, 8.5m and longer distances. The most commonly deployed distances are 3m and 5m.
10GBASE-CX4 — These cables are older and not very common. This type of cable and connector is similar to cables used for InfiniBand technology.
USB Type-C Cables
Information on the new USB Type-C cables is located in the Roadmaps section above.
Several types of connectors are available with cables used for storage interfaces. This is not an exhaustive list but is intended to show the more common types. Each of the connector types includes the number of lanes (or channels) and the rated speed. The speeds listed are across a single port at half duplex (from point A to point B, one direction at a time).
As of early 2011, the fastest generally available connector speeds supported were 10 Gbps per lane. Significantly higher speeds are currently achieved by bundling multiple lanes in parallel, such as 4x10 (40 Gbps), 10x10 (100 Gbps), 12x10 (120 Gbps), etc. Most of the current implementations of 40GbE and 100GbE use multiple lanes of 10GbE and are considered “channel bonded” solutions.
14 Gbps per lane connectors appeared in the last half of 2011. These connectors support 16Gb Fibre Channel (single-lane) and 56Gb (FDR) InfiniBand (multi-lane).
25 Gbps per lane connectors began to appear in volume in 2016. The underlying technology (“25G/28G”) used for 25 Gbps connectors for Ethernet is essentially the same as the technology used for 32Gb Fibre Channel. The Ethernet implementation provides 25 Gbps per lane and the Fibre Channel implementation provides 32 Gbps per lane. This same technology is bundled into four-lane (“quad”) configurations to achieve 100GbE and 128GFC. Other variations of bundling multiple lanes of 25 Gbps may be possible, such as 10x25 (250 Gbps), 12x25 (300 Gbps) or 16x25 (400 Gbps).
With the announcement of 25 Gbps Ethernet, some in the industry believe that single-lane 25 Gbps Ethernet infrastructure (cables, connectors and adapters) will gain faster market acceptance than multi-lane 40 Gbps Ethernet infrastructure.
Two of the popular fiber-optic cable connectors are the SFP-style and the QSFP-style (see diagrams below). SFP stands for “small form-factor pluggable” and QSFP is “quad small form-factor pluggable.” As the data rates for various interfaces have increased, the internal technology in these connectors has changed, and the names have changed slightly. The table below indicates the name of the technology and the interfaces that use it. See the Roadmaps section above for additional details on particular interfaces.
Connectors: development of smaller form factor QSFP connectors is underway.
Development of on-board optical connections is underway via the Consortium for On-Board Optics (COBO).
|Fibre Channel||1GFC, 2GFC, 4GFC||8GFC, 16GFC||32GFC||—||128GFC|
* QSFP28 is also known as QSFP100 for some 100Gb (4 x 25) Ethernet applications.
Note the encoding schemes described above for additional detail on speeds available for various connector and cable combinations.
PCIe data rates and connector types are provided in the PCI Express section.
|Mini SAS HD||SAS HD|
|Small Form-factor Pluggable||SFP, SFP+, SFP28|
|Quad Small Form-factor Pluggable||QSFP, QSFP+, QSFP28|
PCIe connector types are provided in the PCI Express section.
In the second half of 2010, a new variant of the SFP/SFP+ connector was introduced to accommodate the Fibre Channel backbone with 64-port blades and the planned increased density Ethernet core switches. This new connector, known as mSFP, mini-SFP or mini-LC SFP, narrows the optical centerline of a conventional SFP/SFP+ connector from 6.25 mm to 5.25 mm. Although this connector looks very much like a standard SFP style connector, it is narrower and is required for the higher-density devices. The photo at the right shows the difference between mini-SFP and the standard size.
CXP and CFP
The CXP (copper) and CFP (optical) connectors are expected to be used initially for switch-to-switch connections. These are expected for Ethernet and may also be used for InfiniBand. CFP connectors currently support 10 lanes of 10 Gbps connections (10x10) that consume approximately 35-40 watts. CFP2 is a single board, smaller version of CFP that also supports 10x10 but uses less power than CFP. During 2013, quite a bit of development activity is focused on CFP2. A future CFP4 connector is in the planning stages that is expected to use the 25/28G connectors and support 4x25. CFP4 is expected to handle long range fiber optic distances.
Mini SAS and Mini SAS HD
The Mini SAS connector is the familiar 4-lane connector available on most SAS cables today. The Mini SAS HD connector provides twice the density as the Mini SAS connector, and is available in 4-lane and 8-lane configurations. The Mini SAS HD connector is the same connector for passive copper, active copper and optical SAS cables. The diagrams below compare these two types of SAS connectors.
Source: SCSI Trade Association
PCI Express (PCIe)
PCI Express®, also known as PCIe®, stands for Peripheral Component Interconnect Express and is the computer industry standard for the I/O bus for computers introduced in the last few years. The first version of the PCIe specification, 1.0a, was introduced in 2003. Version 2.0 was introduced in 2007 and version 3.0 was introduced in 2010. These versions are often identified by their generation (“gen 1”, “gen 2”, etc.). It can take a year or two between the time the specification is introduced and general availability of computer systems and devices that use the new version of the PCIe specification.
The PCIe specifications are developed and maintained by the PCI-SIG® (PCI Special Interest Group). PCI Express
and PCIe are registered trademarks of the PCI-SIG.
PCIe 2.0 — Servers that have PCIe 2.0 x8 slots can support two ports of 10GbE or two ports of 16GFC on one adapter.
PCIe 3.0 — On 6 March 2012, the major server vendors announced their next generation servers that support PCIe 3.0, which, among other things, doubles the I/O throughput rate from the previous generation. These servers also provide up to 40 PCIe 3.0 lanes per processor socket, which is also at least double from the previous server generation. Workstation and desktop computer motherboards that support PCIe 3.0 first appeared in late 2011. PCIe 3.0 graphics cards appeared in late 2011. Other types of adapters supporting PCIe 3.0 were announced in 2012 and 2013. The PCIe 3.0 specification was completed in November 2010.
PCIe 3.1 — The PCIe 3.1 specification was released in October 2014. It incorporates M-PCIe and consolidates numerous protocol extensions and functionality for ease of access.
PCIe 4.0 — The PCIe 4.0 specification will maintain backward compatibility with previous generations of the PCIe architecture such as PCIe 1.x, 2.x and 3.x. As of mid-2016, the PCIe 4.0, Revision 0.7 specification is available and Revision 0.9 is expected to be available in 2H 2016. The final version of the PCIe 4.0 specification is expected to be complete in 1Q 2017. After the final PCIe 4.0 specification is complete, it may take up to a year or more for products that support the PCIe 4.0 architecture to become generally available, though early prototypes have been shown in 2016.
PCIe 5.0 — As the PCIe 4.0 specification nears completion, PCI-SIG is beginning to discuss the next development of PCIe technology with PCIe 5.0. PCI-SIG is considering supporting 25 GT/s or 32 GT/s per lane, or a combination of each. It is still very early in the process and much investigation remains into what is feasible.
Data rates for different versions of PCIe are shown in the table below. PCIe data rates are expressed in Gigatransfers per second (GT/s) and are a function of the number of lanes in the connection. The number of lanes is expressed with an “x” before the number of lanes, and is often spoken as “by 1”, “by 4”, etc. PCIe supports full-duplex (traffic in both directions). The data rates shown below are in each direction. Note the explanation of encoding schemes described above.
M.2 — M.2 is the next generation PCIe connector that can be used for internally mounted devices such as boot drives in a variety of devices, from mobile to server. Its multiple socket definitions support WWAN, SSD and other applications. M.2 can support PCIe protocol or SATA protocol, but not both at the same time on the same device. M.2 supports a variety of board width and length options. M.2 is available in single-sided modules that can be soldered down, or single-sided and dual-sided modules used with a connector. M.2 PCIe is also available in a Ball Grid Array (BGA) form factor.
Mini-PCIe — PCI Express cards are also available in a mini PCIe form factor. This is a special form factor for PCIe that is approximately 30mm x 51mm or 30mm x 26.5mm, designed for laptop and notebook computers, and equivalent to a single-lane (x1) PCIe slot. A variety of devices including WiFi modules, WAN modules, video/audio decoders, SSDs and other devices are available in this form factor.
U.2 — U.2 (formerly SFF-8639) is the I/O backplane connector designed for high-density SSD storage devices and is backward compatible with existing storage interfaces. SFF-8639 supports PCIe/NVMe, SAS and SATA devices and enables hot plug and hot swap of devices while the system is running.
M-PCIe™ — M-PCIe is the specification that maps PCIe over the MIPI® Alliance M-PHY® technology used in low-power mobile and handheld devices. M-PCIe is optimized for RFI/EMI requirements and supports M-PHY gears 1, 2 and 3 and will be extended to support gear 4.
I/O Virtualization (SR-IOV & MR-IOV) — In 2008, the PCI-SIG announced the completion of its I/O Virtualization (IOV) suite of specifications including single-root IOV (SR-IOV) and multi-root IOV (MR-IOV).
- SR-IOV enables a PCIe device to support multiple Virtual Functions and is currently used with several 10GbE and faster NICs and with various hypervisors.
- MR-IOV enables multiple non-coherent Root Complexes (RCs) to enhance resource utilization of PCI hardware and is primarily the overlaying of multiple Virtual Hierarchies (VHs) over a shared physical set of multi-root aware (MRA) and non-MRA components.
PCIe External Cables —The concept of sharing PCIe devices or providing access to PCIe devices that may be physically larger than some smaller form-factor systems can accommodate has led to the development of external connections to some PCIe devices. Cables have been developed for extending the PCIe bus outside of the chassis holding the PCIe slots. These cables are specified by indicating the number of PCIe lanes (x4, x8, etc.) supported. Cables are typically available for x4, x8 and x16 lane configurations. Common cable lengths are 1m and 3m. The photo below shows some PCIe cables and connectors. PCIe can also be carried over fiber-optic cables for longer distances. In the future we will begin to see the shift from PCIe cables and connectors to OCuLink cables and connectors, also shown in the image below.
OCuLink — OCuLink is intended to be a low-cost, small cable form factor for PCIe internal and external devices such as storage devices or graphics adapters, offering bit rates starting at 8 Gbps, with headroom to scale, and new independent cable clock integration. OCuLink supports x1, x2 and x4 lanes of PCIe 3.0 connectivity. OCuLink supports passive cables capable of reaching up to 2—3 meters and active copper and optical cables. Active copper cables can reach up to 3—10 meters while active optical cables can reach up to 300 meters in length. The OCuLink specification Revision 1.0 was published in October 2015 and is available to PCI-SIG members via the online specification library.
The original version of this page is available at www.demartek.com/Demartek_Interface_Comparison.html on the Demartek website.