Hubbry Logo
Serial Attached SCSISerial Attached SCSIMain
Open search
Serial Attached SCSI
Community hub
Serial Attached SCSI
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Serial Attached SCSI
Serial Attached SCSI
from Wikipedia
SAS
Serial Attached SCSI
Four red cables lead into a wide black electrical connector
SAS connector
Width in bits1
No. of devices65,535
Speed
  • SAS-1: Full-duplex[1] 3 Gbit/s (2004)
  • SAS-2: Full-duplex 6 Gbit/s (2009)
  • SAS-3: Full-duplex 12 Gbit/s (2013)
  • SAS-4: Full-duplex 22.5 Gbit/s (2017)[2]
StyleSerial
Hotplugging interfaceYes

In computing, Serial Attached SCSI (SAS) is a point-to-point serial protocol that moves data to and from computer-storage devices such as hard disk drives, solid-state drives and tape drives. SAS replaces the older Parallel SCSI (Parallel Small Computer System Interface, usually pronounced "scuzzy"[3][4]) bus technology that first appeared in the mid-1980s. SAS, like its predecessor, uses the standard SCSI command set. SAS offers optional compatibility with Serial ATA (SATA), versions 2 and later. This allows the connection of SATA drives to most SAS backplanes or controllers. The reverse, connecting SAS drives to SATA backplanes, is not possible.[5]

The T10 technical committee of the International Committee for Information Technology Standards (INCITS) develops and maintains the SAS protocol; the SCSI Trade Association (SCSITA) promotes the technology.

Introduction

[edit]
Storage servers housing 24 SAS hard disk drives per server

A typical Serial Attached SCSI system consists of the following basic components:

  1. An initiator: a device that originates device-service and task-management requests for processing by a target device and receives responses for the same requests from other target devices. Initiators may be provided as an on-board component on the motherboard (as is the case with many server-oriented motherboards) or as an add-on host bus adapter.
  2. A target: a device containing logical units and target ports that receives device service and task management requests for processing and sends responses for the same requests to initiator devices. A target device could be a hard disk drive or a disk array system.
  3. A service delivery subsystem: the part of an I/O system that transmits information between an initiator and a target. Typically cables connecting an initiator and target with or without expanders and backplanes constitute a service delivery subsystem.
  4. Expanders: devices that form part of a service delivery subsystem and facilitate communication between SAS devices. Expanders facilitate the connection of multiple SAS End devices to a single initiator port.[6]

History

[edit]
  • SAS-1: 3.0 Gbit/s, introduced in 2004[7]
  • SAS-2: 6.0 Gbit/s, available since February 2009
  • SAS-3: 12.0 Gbit/s, available since March 2013
  • SAS-4: 22.5 Gbit/s called "24G",[8] standard completed in 2017[7][2]
  • SAS-5: 45 Gbit/s development started 2018[9]

Identification and addressing

[edit]

A SAS Domain is the SAS version of a SCSI domain—it consists of a set of SAS devices that communicate with one another by means of a service delivery subsystem. Each SAS port in a SAS domain has a SCSI port identifier that identifies the port uniquely within the SAS domain, the World Wide Name. It is assigned by the device manufacturer, like an Ethernet device's MAC address, and is typically worldwide unique as well. SAS devices use these port identifiers to address communications to each other.

In addition, every SAS device has a SCSI device name, which identifies the SAS device uniquely in the world. One does not often see these device names because the port identifiers tend to identify the device sufficiently.

For comparison, in parallel SCSI, the SCSI ID is the port identifier and device name. In Fibre Channel, the port identifier is a WWPN and the device name is a WWNN.

In SAS, both SCSI port identifiers and SCSI device names take the form of a SAS address, which is a 64 bit value, normally in the NAA IEEE Registered format. People sometimes refer to a SCSI port identifier as the SAS address of a device, out of confusion. People sometimes call a SAS address a World Wide Name or WWN, because it is essentially the same thing as a WWN in Fibre Channel. For a SAS expander device, the SCSI port identifier and SCSI device name are the same SAS address.

Comparison with parallel SCSI

[edit]
  • The SAS "bus" operates point-to-point while the SCSI bus is multidrop. Each SAS device is connected by a dedicated link to the initiator, unless an expander is used. If one initiator is connected to one target, there is no opportunity for contention; with parallel SCSI, even this situation could cause contention.
  • SAS has no termination issues and does not require terminator packs like parallel SCSI.
  • SAS eliminates clock skew.
  • SAS allows up to 65,535 devices through the use of expanders, while Parallel SCSI has a limit of 8 or 16 devices on a single channel.
  • SAS allows a higher transfer speed (SAS-1, SAS-2, SAS-3, and SAS-4 supports data bandwidth of 3, 6, 12, and 24 Gbits/sec, respectively)[10] than most parallel SCSI standards. SAS achieves these speeds on each initiator-target connection, hence getting higher throughput, whereas parallel SCSI shares the speed across the entire multidrop bus.
  • SAS devices feature dual ports, allowing for redundant backplanes or multipath I/O; this feature is usually referred to as the dual-domain SAS.[11]
  • SAS controllers may connect to SATA devices, either directly connected using native SATA protocol or through SAS expanders using Serial ATA Tunneling Protocol (STP).
  • Both SAS and parallel SCSI use the SCSI command set.

Comparison with SATA

[edit]

There is little physical difference between SAS and SATA.[12]

  • SAS protocol provides for multiple initiators in a SAS domain, while SATA has no analogous provision.[12]
  • Most SAS drives provide tagged command queuing, while most newer SATA drives provide native command queuing.[12]
  • SATA uses a command set that is based on the parallel ATA command set and then extended beyond that set to include features like native command queuing, hot-plugging, and TRIM. SAS uses the SCSI command set, which includes a wider range of features like error recovery, reservations and block reclamation. Basic ATA has commands only for direct-access storage. However SCSI commands may be tunneled through ATAPI[12] for devices such as CD/DVD drives.
  • SAS hardware allows multipath I/O to devices while SATA (prior to SATA 2.0) does not.[12] Per specification, SATA 2.0 makes use of port multipliers to achieve port expansion, and some port multiplier manufacturers have implemented multipath I/O using port multiplier hardware.
  • SATA is marketed as a general-purpose successor to parallel ATA and has become common in the consumer market, whereas the more-expensive[when?] SAS targets critical server applications.
  • SAS error-recovery and error-reporting uses SCSI commands, which have more functionality than the ATA SMART commands used by SATA drives.[12]
  • SAS uses higher signaling voltages (800–1,600 mV for transmit, and 275–1,600 mV for receive[clarification needed]) than SATA (400–600 mV for transmit, and 325–600 mV for receive[clarification needed]). The higher voltage offers (among other features) the ability to use SAS in server backplanes.[12]
  • Because of its higher signaling voltages, SAS can use cables up to 10 m (33 ft) long, whereas SATA has a cable-length limit of 1 m (3.3 ft) or 2 m (6.6 ft) for eSATA.[12]
  • SAS is full duplex, whereas SATA is half duplex. The SAS transport layer can transmit data at the full speed of the link in both directions at once, so a SCSI command executing over the link can transfer data to and from the device simultaneously. However, because SCSI commands that can do that are rare, and a SAS link must be dedicated to an individual command at a time, this is generally not an advantage with a single device.[13]

Characteristics

[edit]

Technical details

[edit]

The Serial Attached SCSI standard defines several layers (in order from highest to lowest): application, transport, port, link, PHY and physical. Serial Attached SCSI comprises three transport protocols:

  • Serial SCSI Protocol (SSP) – for command-level communication with SCSI devices.
  • Serial ATA Tunneling Protocol (STP) – for command-level communication with SATA devices.
  • Serial Management Protocol (SMP) – for managing the SAS fabric.

For the Link and PHY layers, SAS defines its own unique protocol.

At the physical layer, the SAS standard defines connectors and voltage levels. The physical characteristics of the SAS wiring and signaling are compatible with and have loosely tracked that of SATA up to the 6 Gbit/s rate, although SAS defines more rigorous physical signaling specifications as well as a wider allowable differential voltage swing intended to allow longer cabling. While SAS-1.0 and SAS-1.1 adopted the physical signaling characteristics of SATA at the 3 Gbit/s rate with 8b/10b encoding, SAS-2.0 development of a 6 Gbit/s physical rate led the development of an equivalent SATA speed. In 2013, 12 Gbit/s followed in the SAS-3 specification.[14] SAS-4 is slated to introduce 22.5 Gbit/s signaling with a more efficient 128b/150b encoding scheme to realize a usable data rate of 2,400 MB/s while retaining compatibility with 6 and 12 Gbit/s.[15]

Additionally, SCSI Express takes advantage of the PCI Express infrastructure to directly connect SCSI devices over a more universal interface.[16]

Architecture

[edit]
The architecture of SAS layers

SAS architecture consists of six layers:

  • Physical layer:
    • defines electrical and physical characteristics
    • differential signaling transmission
    • Multiple connector types:
      • SFF-8482 – SATA compatible
      • Internal four-lane connectors: SFF-8484, SFF-8087, SFF-8643
      • External four-lane connectors: SFF-8470, SFF-8088, SFF-8644
  • PHY Layer:
  • Link layer:
    • Insertion and deletion of primitives for clock-speed disparity matching
    • Primitive encoding
    • Data scrambling for reduced EMI
    • Establish and tear down native connections between SAS targets and initiators
    • Establish and tear down tunneled connections between SAS initiators and SATA targets connected to SAS expanders
    • Power management (proposed for SAS-2.1)
  • Port layer:
    • Combining multiple PHYs with the same addresses into wide ports
  • Transport layer:
    • Contains three transport protocols:
      • Serial SCSI Protocol (SSP): for command-level communication with SCSI devices
      • Serial ATA Tunneled Protocol (STP): for command-level communication with SATA devices
      • Serial Management Protocol (SMP): for managing the SAS fabric
  • Application layer

Topology

[edit]

An initiator may connect directly to a target via one or more PHYs (such a connection is called a port whether it uses one or more PHYs, although the term wide port is sometimes used for a multi-PHY connection).

SAS expanders

[edit]

The components known as Serial Attached SCSI Expanders (SAS Expanders) facilitate communication between large numbers of SAS devices. Expanders contain two or more external expander-ports. Each expander device contains at least one SAS Management Protocol target port for management and may contain SAS devices itself. For example, an expander may include a Serial SCSI Protocol target port for access to a peripheral device. An expander is not necessary to interface a SAS initiator and target but allows a single initiator to communicate with more SAS/SATA targets. A useful analogy: one can regard an expander as akin to a network switch in a network, which connects multiple systems using a single switch port.

SAS 1 defined two types of expander; however, the SAS-2.0 standard has dropped the distinction between the two, as it created unnecessary topological limitations with no realized benefit:

  • An edge expander allows for communication with up to 255 SAS addresses, allowing the SAS initiator to communicate with these additional devices. Edge expanders can do direct table routing and subtractive routing. (For a brief discussion of these routing mechanisms, see below). Without a fanout expander, you can use at most two edge expanders in a delivery subsystem (because you connect the subtractive routing port of those edge expanders together, and you can not connect any more expanders). Fanout expanders solve this bottleneck.
  • A fanout expander can connect up to 255 sets of edge expanders, known as an edge expander device set, letting even more SAS devices be addressed. The subtractive routing port of each edge expanders connects to the phys of fanout expander. A fanout expander cannot do subtractive routing, it can only forward subtractive routing requests to the connected edge expanders.

Direct routing allows a device to identify devices directly connected to it. Table routing identifies devices connected to the expanders connected to a device's own PHY. Subtractive routing is used when you are not able to find the devices in the sub-branch you belong to. This passes the request to a different branch altogether.

Expanders exist to allow more complex interconnect topologies. Expanders assist in link-switching (as opposed to packet-switching) end-devices (initiators or targets). They may locate an end-device either directly (when the end-device is connected to it), via a routing table (a mapping of end-device IDs and the expander the link should be switched to downstream to route towards that ID), or when those methods fail, via subtractive routing: the link is routed to a single expander connected to a subtractive routing port. If there is no expander connected to a subtractive port, the end-device cannot be reached.

Expanders with no PHYs configured as subtractive act as fanout expanders and can connect to any number of other expanders. Expanders with subtractive PHYs may only connect to two other expanders at a maximum, and in that case they must connect to one expander via a subtractive port and the other via a non-subtractive port.

SAS-1.1 topologies built with expanders generally contain one root node in a SAS domain with the one exception case being topologies that contain two expanders connected via a subtractive-to-subtractive port. If it exists, the root node is the expander, which is not connected to another expander via a subtractive port. Therefore, if a fanout expander exists in the configuration, it must be the domain's root node. The root node contains routes for all end devices connected to the domain. Note that with the advent in SAS-2.0 of table-to-table routing and new rules for end-to-end zoning, more complex topologies built upon SAS-2.0 rules do not contain a single root node.

Connectors

[edit]

SAS connectors are much smaller than traditional parallel SCSI connectors. Commonly, SAS-3 provides for point data transfer speeds up to 12 Gbit/s.[18] Currently, SAS-4 is available with up to 24 Gbps; with SAS-5 under development, according to T10.

The physical SAS connector comes in several different variants:[19]

Code-
name[20]
other names external/
internal
Pins No of devices
/ lanes
Comment Image
SFF-8086 Internal mini-SAS,
internal mSAS
internal 26 4 devices
4 lanes
This is a less common implementation of internal mSAS than SFF-8087's 36-circuit version.
The fewer positions is enabled by it not supporting sidebands.
SFF-8087[21][22] Internal mini-SAS,
internal mSAS,
internal iSAS,
internal iPass
internal 36 4 devices
4 lanes
Unshielded 36-circuit implementation of SFF-8086.
Molex iPass reduced width internal 4× connector; 12 Gbit/s capability.
SFF-8088[23][24] External mini-SAS,
external mSAS,
external iSAS,
external iPass
external 26 4 devices
4 lanes
Shielded 26-circuit implementation of SFF-8086.
Molex iPass reduced width external 4× connector; 12 Gbit/s capability.
SFF-8431[25][26] SFP+ external 20 1 device
1 lane
SFF-8436[27][28] QSFP+,
Quad SFP+
external 38 1 device
4 lanes
Commonly used with many NetApp storage systems.
Often seen with SFF-8088 or SFF-8644 on the other end; 6 Gbit/s capability.
SFF-8470[29][30] InfiniBand CX4
connector,
Molex LaneLink
external 34 4 devices
4 lanes
High-density external connector (also used as an internal connector).
SFF-8482[31][32] internal 29 1 device
1 lane
This form factor is designed for compatibility with SATA but can drive a SAS device.
A SAS controller can control SATA drives, but a SATA controller cannot control SAS drives. Lower pins (S1-S7, P1-P11) defined as in SATA. Upper pins S8-S14 provide additional lane of data.

The most common connection[33] for SAS drives connecting to backplanes in servers, i.e. PowerEdge[34] and ProLiant[35]

SFF-8484[36][37] internal 32 or
19
4 devices
4 lanes
High-density internal connector, 2 and 4 lane versions are defined by the SFF standard.
SFF-8485[38] Defines SGPIO (extension of SFF 8484),
a serial link protocol used usually for LED indicators.
SFF-8613[39]
(SFF-8643[40][41])
Mini-SAS HD,
U.2 internal
internal 36 1 device
4 lanes
Mini-SAS HD (introduced with SAS 12 Gbit/s)

Also known as a U.2 port[42] along with SFF-8639.

SFF-8614[43]
(SFF-8644[44][45])
external Mini-SAS HD external 1 device
4 lanes
Mini-SAS HD (introduced with SAS 12 Gbit/s)
Sideband
connector
internal Often seen with 1× SFF-8643 or 1× SFF-8087 on the other end –
internal fan-out for 4× SATA drives.
Connects the controller to drives without backplane or
to the (SATA) backplane and optionally, to the status LEDs.
SFF-8680[46][47] internal
  • 1
  • (2 ports)
  • SAS 12 Gbit/s backplane connector;
  • same pinout as SFF-8482, but with electrical requirements for 12 Gbit/s;
  • SFF-8678 is analogously defined for 6 Gbit/s.
SFF-8639[48][49] U.2[50]Also used for SATA Express and U.3 (SFF-TA-1001). internal 68 1 device
4 lanes
  • SAS 12 Gbit/s backplane connector;
  • downward-compatible with SFF-8680.

SFF-8638[51]
  • Four 1x ports at up to 24 Gb/s each;
  • two 2x ports at up to 48 Gb/s each;
  • one 4x port at up to 96 Gb/s.
SFF-8640[52]
  • Four 1x ports at up to 24 Gb/s each;
  • two 2x ports at up to 48 Gb/s each;
  • one 4x port at up to 96 Gb/s.[53]
SFF-8681[54]
  • Two 1x ports at up to 24 Gb/s each;
  • one 2x ports at up to 48 Gb/s each.
SFF-8654[55] SlimSAS[56] internal 4X: 38

8X: 74

1 device
4 lanes
4X and 8X SAS-4 plug and receptacle

Nearline SAS

[edit]

Nearline SAS (abbreviated to NL-SAS, and sometimes called midline SAS) drives have a SAS interface, but head, media, and rotational speed of traditional enterprise-class SATA drives, so they cost less than other SAS drives. When compared to SATA, NL-SAS drives have the following benefits:[57]: 20 

  • Dual ports allowing redundant paths
  • Ability to connect a device to multiple computers
  • Full SCSI command set
  • No need for using Serial ATA Tunneling Protocol (STP), which is necessary for SATA HDDs to be connected to a SAS HBA.[57]: 16 
  • No need for SATA interposer cards, which are needed for pseudo–dual-port high availability of SATA HDDs.[57]: 17 
  • Larger depth of command queues

See also

[edit]

References

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia
from Grokipedia
Serial Attached SCSI (SAS) is a point-to-point serial protocol designed for high-speed data transfer between computers and storage devices, such as hard disk drives and solid-state drives, serving as an evolutionary successor to with enhanced performance, scalability, and reliability for enterprise applications. Developed by the T10 technical committee under ANSI/INCITS, SAS was standardized in 2003 (INCITS 376-2003), with products introduced in 2004, building on commands while adopting serial transmission to overcome limitations like and device count in parallel interfaces. It utilizes a layered architecture, including physical, link, transport, and application layers, with data rates starting at 3 Gbit/s and advancing to 22.5 Gbit/s in recent generations. At its core, SAS employs three key transport protocols: the Serial SCSI Protocol (SSP) for direct SCSI command execution, the Serial ATA Tunneling Protocol (STP) to enable compatibility with drives on the same infrastructure, and the Serial Management Protocol (SMP) for configuring and monitoring expanders that extend connectivity. This dual support for and devices allows seamless integration in mixed environments, reducing costs and simplifying deployment in data centers. SAS domains can scale to support up to 65,536 devices through cascaded expanders, each handling up to 128 connections, making it ideal for mission-critical storage arrays requiring and . The protocol's latest iteration, SAS Protocol Layer 5 (SPL-5) under INCITS 554-2023, further refines serial interconnect rules for improved efficiency in modern peripherals like tape drives and SSDs, with ongoing work on 24G+ enhancements and SAS-5.

Overview

Definition and Purpose

Serial Attached (SAS) is a point-to-point serial protocol that leverages the SCSI command set to enable data transfer between computer systems and storage devices, including hard disk drives (HDDs), solid-state drives (SSDs), and tape drives. This interface uses thin serial cables to establish direct connections, facilitating reliable and efficient communication in storage-intensive applications. The purpose of SAS is to deliver high-performance, scalable connectivity for storage in enterprise environments, such as data centers, servers, and storage area networks (SANs). It supports up to 16,384 devices per domain, enabling expansive configurations that meet the demands of large-scale and archiving. Key components of SAS include initiators, which are host bus adapters or controllers that send commands; , the peripheral storage devices that receive and process those commands; expanders, which act as intelligent switches to route connections and extend the network; and service delivery subsystems, encompassing the physical infrastructure like cables and backplanes that link these elements. SAS emerged as the successor to , overcoming its predecessor's constraints on cable length and the number of attachable devices. It maintains compatibility with the command set, ensuring broad interoperability with established storage protocols.

Key Features and Advantages

Serial Attached SCSI (SAS) supports dual-port , allowing each device to have two independent ports for enhanced and capabilities in high-availability environments. This dual-port design ensures that if one path fails, the other can seamlessly take over, minimizing in enterprise storage systems. SAS incorporates and port through its expander devices, enabling logical isolation of storage resources and improved scalability. , standardized in the SAS-2 specification, allows administrators to create secure zones that control access between hosts and targets, similar to , supporting up to 256 devices per expander for denial-of-service and device isolation. Port in expanders permits a single physical to function as multiple logical ports, expanding connectivity without additional controller ports and facilitating large-scale deployments. The protocol operates in full-duplex mode, enabling simultaneous bidirectional data transfer, and supports wide ports comprising up to four lanes per port to aggregate bandwidth for higher throughput. Additionally, SAS is compatible with SATA devices through the SATA Tunneling Protocol (STP), which encapsulates SATA commands within SAS frames, allowing seamless integration of SATA drives into SAS domains for mixed-environment flexibility. SAS offers superior reliability via robust error detection and correction mechanisms, including cyclic redundancy checks (CRC) on all frames to ensure data integrity during transmission. It supports cable lengths up to 10 meters for external copper connections, providing greater deployment flexibility compared to shorter alternatives. Furthermore, the protocol accommodates multiple initiators accessing a single target, enhancing resource sharing in multi-host configurations.

History and Standards

Development and Evolution

The development of (SAS) originated in the early as an evolutionary successor to , addressing key limitations such as restricted device counts per bus (typically 15 devices) and short cable lengths due to signal skew and noise in parallel transmission. In May 2002, the INCITS Technical Committee T10, responsible for SCSI standards, accepted a project proposal from industry leaders including , HP, and to create a serial interface that retained 's command set while enabling point-to-point connections for improved scalability. The Trade Association (STA), formed to promote technologies, collaborated closely with T10 to drive this initiative, culminating in the first ANSI-approved SAS standard, INCITS 376-2003, published in 2003. Key milestones marked SAS's progression from specification to market reality. The first SAS-compliant products, including host bus adapters and drives from vendors like Seagate and LSI Logic, became available in 2004, with demonstrations of functional silicon occurring as early as January of that year. By 2009, SAS achieved widespread adoption in enterprise environments, driven by the release of enhanced standards that supported broader deployment in servers and storage arrays. Post-2010, SAS integrated seamlessly with solid-state drives (SSDs), with major announcements such as Seagate's SAS SSDs in 2011 and subsequent entries from and Micron, enabling high-performance flash storage in data centers. The primary drivers for SAS's evolution were the demand for higher data transfer rates beyond parallel SCSI's Ultra320 limits, enhanced to support thousands of devices through switched topologies, and the industry shift toward serial interfaces for reduced and better reliability in expanding storage ecosystems. ANSI-accredited INCITS, through its T10 committee, handled the core technical specifications and protocol definitions, ensuring compatibility and . Meanwhile, the STA—now operating under the Storage Networking Industry Association (SNIA)—played a pivotal role in promotion, , and roadmap development to accelerate industry adoption.

Generations and Speeds

Serial Attached SCSI (SAS) has progressed through multiple generations, each advancing transfer rates while incorporating encoding improvements to enhance efficiency and reduce overhead. The initial generations relied on 8b/10b encoding, which maps 8 bits of to 10-bit symbols for DC balance and , resulting in 20% overhead. Later generations shifted to more efficient schemes like 128b/130b encoding, which prepends a 2-bit sync header to 128 bits, yielding approximately 98.5% efficiency. SAS-1, standardized as INCITS 376-2003 and introduced in products by 2004, operates at 3 Gbit/s per lane using 8b/10b encoding as its basic serial implementation. This provides an effective throughput of about 300 MB/s after encoding overhead, marking the transition from parallel SCSI to serial point-to-point connections. SAS-2, ratified as INCITS 457-2010 and available since 2009, doubles the speed to 6 Gbit/s per lane while retaining 8b/10b encoding. It introduces the SATA Tunneling Protocol (STP) to encapsulate SATA commands over SAS links, enabling compatibility with SATA devices, and includes improved power management for better energy efficiency in enterprise environments. SAS-3, approved as INCITS 519-2014 and released in 2013, achieves 12 Gbit/s per by adopting 128b/130b encoding for greater efficiency. This scheme reduces overhead compared to 8b/10b, with the effective data rate calculated as 128130×\frac{128}{130} \times raw rate, approaching 11.8 Gbit/s while maintaining compatibility with prior topologies.
GenerationYearSpeed (Gbit/s per lane)EncodingKey Improvements
SAS-120043 (effective ~2.4)8b/10bBasic serial point-to-point links
SAS-220096 (effective ~4.8)8b/10bSTP for compatibility,
SAS-3201312 (effective ~11.8)128b/130bHigher efficiency encoding
SAS-4201722.5 (effective ~19.2)128b/130b + FEC (eff. 128b/150b)FEC for reliability, tri-mode support with PCIe/SATA
SAS-5202345 (target, signaling)TBDProtocol enhancements for hyperscale, low latency
SAS-4, standardized as INCITS 534-2019 and announced in 2017, supports a 22.5 Gbit/s signaling rate per lane (effective ~19.2 Gbit/s or 2400 MB/s after 128b/130b encoding and 20-bit FEC, equivalent to 128b/150b), with (FEC) to ensure reliability over longer channels. This generation aligns with PCIe Gen4 for shared connector designs in enterprise storage, enabling tri-mode support for SAS, SATA, and PCIe in unified backplanes. The effective encoding efficiency is 128150\frac{128}{150} due to FEC overhead. SAS-5 (including SPL-5, ratified as INCITS 554-2023 for the protocol layer), targets 45 Gbit/s per lane in its ongoing development (project INCITS 561, in development as of 2025), with enhancements to the protocol layer optimized for hyperscale centers and HDD workloads, including reduced latency and better for large-scale patterns. As of 2025, SAS-5 is still in development with no commercial implementations, while SAS-4 has seen adoption in high-performance SSDs and HDDs. This evolution builds on prior encoding efficiencies while preparing for even higher raw rates in future physical layer implementations.

Technical Architecture

Identification and Addressing

In Serial Attached SCSI (SAS) networks, devices are uniquely identified using 64-bit SAS addresses, which function as World Wide Names (WWNs) and ensure global uniqueness across the domain, similar to those used in protocols. These addresses follow the Network Address Authority (NAA) IEEE Registered format, beginning with NAA 5h, followed by a 24-bit IEEE Company ID and a 32-bit vendor-specific identifier, and are assigned to each SAS port on initiators, targets, and expanders. The SAS address serves as the primary SCSI port identifier, replacing the traditional parallel SCSI bus ID scheme and enabling unambiguous device recognition without conflicts. The addressing scheme in SAS combines the 64-bit SAS address of a port with an 8-bit SCSI ID for targets, allowing initiators to specify destinations in command frames and connection requests. This structure supports routing to individual targets within the domain, where the SCSI ID differentiates targets sharing the same port address, while logical unit numbers (LUNs) up to 64 bits address subunits on a target. Through expander devices, which maintain routing tables with 16-bit indices supporting up to 65,535 entries, SAS domains can scale to support up to 16,384 devices in practical configurations, far exceeding direct-attach limitations. Device discovery occurs domain-wide via the Serial Management Protocol (SMP), where management application clients issue SMP DISCOVER requests to expanders to build routing tables and map . This process involves level-order traversal starting from the root expander, reporting attached device types, SAS addresses, negotiated link rates, and PHY details after link resets or broadcast changes, ensuring all devices are discoverable and routable. SAS ports are classified as narrow (single PHY) or wide (multiple PHYs sharing the same SAS address), with each PHY assigned a unique 8-bit identifier for point-to-point link management within the port. Port identifiers, derived from the SAS address, are exchanged in IDENTIFY address frames during the identification sequence, facilitating connection establishment and at the . Expanders play a key role in scaling by using these identifiers to route connections across up to 128 PHYs per device.

Protocol Layers and Protocols

The (SAS) is organized into a that facilitates reliable data exchange between initiators and targets. This structure includes the physical (PHY) layer, , layer, , and , each handling specific aspects of communication from to command execution. The PHY and manage the serial interface and basic connection setup, while the layer coordinates multiple PHYs within a device. The oversees frame-based data transfer across the three primary protocols, and the processes SCSI commands. This layered approach ensures compatibility with SCSI semantics while leveraging serial point-to-point connectivity. The (PHY) defines the electrical and signaling characteristics for SAS links, using 8b/10b encoding to transmit data at rates compatible with SATA signaling. It handles (OOB) signaling for link initialization, such as COMRESET and COMWAKE primitives, and supports dual-port configurations for . The builds on the PHY by managing primitive sequences for connection establishment and , including OPEN and CLOSE primitives to initiate and terminate connections, as well as error detection through invalid dword identification. The port layer aggregates multiple PHYs into logical , enabling wide port configurations for increased bandwidth without altering upper-layer protocols. At the transport layer, SAS employs three distinct protocols to handle different types of traffic: the Serial Protocol (SSP) for native I/O, the Serial ATA Tunneled Protocol (STP) for device compatibility, and the Serial Management Protocol (SMP) for discovery and configuration. SSP frames are structured with start-of-frame (SOF) and end-of-frame (EOF) delimiters, a 32-byte frame header including protocol type and connection rate, a variable up to 1,024 bytes for data, and a 32-bit CRC for integrity. STP encapsulates frames with additional headers and uses -specific primitives like SATA_X_RDY for flow control. SMP frames support request-response exchanges for tasks like PHY logging, marked by dedicated SOF/EOF pairs. Flow control is credit-based, with SSP utilizing up to 255 receive ready (RRDY) credits per connection to regulate data transmission and prevent ; STP relies on hold/dma ready primitives. recovery mechanisms include ACK/NAK responses for frame retransmission, with NAKs signaling issues like CRC errors or invalid primitives, and retry limits enforced per connection to maintain link reliability. The application layer in SAS primarily implements the full SCSI command set via SSP, enabling initiators to issue commands to targets for data access and management. Core commands include READ and WRITE operations, which transfer logical blocks of data; for instance, the 16-byte variants (READ(16) and WRITE(16)) support transfer lengths up to 2^{32} blocks as defined by their command descriptor blocks (CDBs), though device-specific limits reported in the Block Limits VPD page (e.g., via INQUIRY) may cap this at lower values such as 16 million blocks to align with buffer capacities. These commands encapsulate SCSI Primary Commands (SPC) and Block Commands (SBC) elements, with responses including status bytes and sense data for error reporting. SMP extends application-layer functionality for non-SCSI tasks, such as reporting SAS addresses or configuring expander routing, ensuring domain-wide management without disrupting I/O flows.

Topology and Expanders

Serial Attached SCSI (SAS) supports flexible to connect initiators, targets, and expanders in a domain, enabling scalable storage configurations without loops or multiple paths between devices. The simplest topology is direct attach, a point-to-point connection between an initiator and a target , limited to a single link per connection. Daisy-chaining allows sequential connections of multiple devices, where each device's output links to the next input, though this is constrained by the number of available ports and addressing limits. For larger setups, switched topologies via expanders form or networks, where a central or interconnected expanders route traffic to numerous end devices, supporting communication across the domain. SAS expanders function as intelligent switches that extend connectivity by managing multiple point-to-point links, using internal tables to direct connection requests based on destination SAS addresses. Each expander possesses a unique 64-bit SAS address and up to 128 physical interfaces (phys), allowing it to connect initiators, , and other expanders while retiming signals and handling . Expanders support a domain-wide addressing scheme accommodating up to 65,535 unique SAS addresses, enabling vast configurations through cascaded connections. There are two primary types: edge expanders, which primarily connect end devices like drives and use simpler for local sets; and expanders, which link multiple edge expander sets or direct ports using advanced tables, with a maximum of one per domain to avoid complexity. Routing in expanders relies on per-phy attributes to forward connection requests efficiently. Direct routing handles immediate local connections without table lookup, while table routing consults an indexed table of up to 128 entries per phy to match destination addresses and select output phys. Subtractive routing serves larger domains by forwarding unmatched requests from all non-direct and non-table phys to a single designated subtractive port, typically upstream to another expander, allowing hierarchical discovery and connectivity. In expansive setups without a fanout expander, up to two edge expander sets—each limited to 128 SAS addresses—can connect via subtractive routing. Zoning enhances security in SAS domains by configuring expander routing tables to restrict unauthorized connections, using commands like REPORT ZONE ROUTE TABLE to map and enforce access policies during discovery. Zoning expanders traverse the , gathering information via SMP to populate zone-aware tables that limit phy-to-phy interactions, preventing lateral access between isolated groups of devices. This feature, introduced in SAS-2, ensures compartmentalized communication in shared environments. SAS topologies impose physical and structural limits to maintain and domain manageability. Each link supports a maximum of 10 meters per hop using passive cables, beyond which active cables or optical extensions may be required for longer distances. A domain includes 1 fanout expander connected to up to 128 edge expander device sets, with practical configurations supporting up to 16,384 devices.

Hardware Components

Connectors and Cables

Serial Attached SCSI (SAS) employs standardized connectors defined by the Small Form Factor (SFF) Committee to facilitate high-speed serial connections between hosts and storage devices. These connectors support multiple for aggregating bandwidth and are designed for both internal and external applications, ensuring compatibility across SAS generations. Internal connectors include the SFF-8087, a 36-circuit Mini SAS plug and receptacle that supports up to four for SAS-1 and SAS-2 implementations, commonly used in server backplanes and direct drive attachments. For higher-speed SAS-3 and SAS-4, the SFF-8643 provides an unshielded Mini Multilane interface with 36 pins, accommodating up to four (or eight via dual connectors) at up to 22.5 Gb/s per lane in SAS-4 while reducing connector footprint. For higher-density applications in SAS-4, SlimSAS (SFF-8654) supports up to eight internal . External connectors, such as the SFF-8470, feature a high-density 34-pin design with jack screws for , supporting four in shielded configurations suitable for cable-to-cable or interconnections. SAS cables utilize differential twisted-pair wiring to transmit serial data, with impedance matched at 100 ohms to maintain across . Internal cables are typically limited to short runs of 0.5 to 1 meter due to enclosure constraints, while external cables extend up to 10 meters, leveraging SAS's higher signaling tolerance compared to . For extended distances, some SAS implementations incorporate optic extensions with active transceivers, enabling reaches beyond 10 meters in specialized enterprise environments. Backward compatibility is inherent in SAS connector designs; for instance, SAS-4 interfaces like those based on SFF-8643 are compatible with prior-generation interfaces such as SFF-8087 via appropriate cabling or adapters, allowing upgrades without full hardware replacement. Lane aggregation enables flexible port widths, with configurations supporting x1 (single lane), x2, or x4 (four lanes) wide s to scale bandwidth from 3 Gb/s (x1 SAS-1) to 90 Gb/s (x4 SAS-4) per , depending on the generation. Pinouts for SAS connectors feature dedicated differential pairs for transmit (Tx) and receive (Rx) signals per lane, alongside ground and sideband pins for vendor-specific control or . Signaling employs low-voltage differential pairs with a recommended peak-to-peak voltage of up to 1,200 mV, operating at approximately 0.6 V differential to ensure robust transmission over copper media while minimizing .

Host Adapters and Devices

Host Bus Adapters (HBAs) serve as the primary interface between a host system and SAS storage devices, typically implemented as PCIe expansion cards that provide multiple SAS ports for connectivity. These adapters, such as those from (formerly LSI), support configurations like the 9300 series with up to 16 internal ports and optional integrated functionality for and optimization, or the 9500 series with support for 24 Gb/s SAS. For instance, the LSI SAS 9311-8i model utilizes eight lanes of PCIe 3.0 to deliver 12 Gb/s SAS connectivity, enabling high-throughput access to enterprise storage arrays. Target devices in SAS environments include hard disk drives (HDDs), solid-state drives (SSDs), and tape drives, which act as endpoints responding to commands from the host. SAS HDDs and SSDs, such as Seagate's Exos series or Western Digital's Ultrastar DC models, feature dual-porting to allow simultaneous access from two initiators, enhancing in clustered systems. These drives are backward-compatible with devices through SAS controllers or adapters, permitting mixed environments where HDDs or SSDs can connect via SAS-to-SATA bridging. SAS tape drives, like Quantum's LTO series, provide high-capacity archival storage with SAS interfaces for reliable backup in data centers. Initiator software, embedded in operating systems, manages command issuance from the host to SAS targets via HBA drivers. In Windows and environments, drivers from manufacturers like handle SAS protocol operations, including the MegaRAID SAS driver package that supports installation and configuration for PCIe-based HBAs. These drivers ensure seamless integration, allowing the OS to discover and utilize SAS devices without custom coding. SAS devices adhere to standardized form factors for enterprise deployment, with 2.5-inch small form factor (SFF) and 3.5-inch large form factor (LFF) drives being predominant. The SFF-8200 series specifications, developed by the SNIA, define dimensions and mounting for 2.5-inch SAS drives, including connector placements for hot-plug compatibility in server chassis. Power requirements for these devices typically range from 5-10 watts for SFF SSDs to higher draws for LFF HDDs, optimized for rack-mounted systems.

Performance Characteristics

Core Specifications

Serial Attached SCSI (SAS) operates on a full-duplex basis, enabling simultaneous data transmission and reception over each lane, which distinguishes it from half-duplex interfaces and supports efficient bidirectional communication in enterprise environments. The protocol's bandwidth scales across generations, with SAS-1 providing 3 Gbit/s per lane, SAS-2 at 6 Gbit/s, SAS-3 at 12 Gbit/s, and SAS-4 reaching a signaling rate of 24 Gbit/s with an effective data rate of 22.5 Gbit/s per lane; aggregated wide ports utilizing four lanes can achieve up to 90 Gbit/s in SAS-4 configurations, while SAS-5 development targets exceed 24 Gbit/s per lane for future enhancements. When paired with solid-state drives (SSDs), SAS delivers operations per second () capabilities up to several hundred thousand in random read/write scenarios for enterprise models, such as 440,000 random read reported for high-capacity SAS SSDs, enabling high-performance handling of transactional workloads. Latency in SAS environments remains low for command-response cycles, as low as 0.1 ms for SSD-based systems due to the protocol's efficient command queuing and direct attachment model, though HDD implementations may approach 3-4 ms under load from mechanical seek times. The interface supports extensive queue depths, accommodating up to tagged commands per initiator port as defined in the architecture, which allows for deep queuing and improved throughput in multi-device topologies without overwhelming individual device limits, often capped at 254-512 per target in practice. This allows for deep queuing and improved throughput in multi-device topologies without overwhelming individual device limits, often capped at 254-512 per target in practice. Reliability features in SAS include cyclic redundancy check (CRC) for end-to-end data integrity verification on every frame, reducing undetected errors to rates below 10^{-12}, and link reset primitives that enable rapid recovery from transient faults by reinitializing the without full disruption. Enterprise SAS drives further enhance dependability with (MTBF) ratings exceeding 2 million hours, reflecting robust design for continuous operation in data centers, including vibration tolerance and error recovery mechanisms. Power management in SAS incorporates features like power disable primitives, allowing initiators to remotely control target device power states for energy efficiency, with typical consumption ranging from 5-10 W per 3.5-inch enterprise drive during active operations—typically 7-14 W for 2.5-inch SAS SSDs active—and idle modes dropping to 4-6 W, supporting scalable power budgeting in dense storage arrays.

Comparison with Parallel SCSI

Serial Attached SCSI (SAS) represents a fundamental architectural evolution from Parallel SCSI, shifting from a parallel multidrop bus to a serial point-to-point interface. In Parallel SCSI, data is transmitted simultaneously across multiple parallel lines, which introduces challenges such as signal skew—where signals on different lines arrive at slightly different times due to varying delays—and requires termination resistors at both ends of the bus to prevent signal reflections. In contrast, SAS employs a serial transmission over dedicated point-to-point links, eliminating skew and the need for termination resistors, as each connection is independent and does not share a common bus. This design allows for more reliable signaling at higher speeds and simplifies cabling. Scalability is significantly enhanced in SAS compared to . is limited to up to 8 devices for narrow or 16 for wide buses, and maximum cable lengths are constrained—up to 25 meters for high-voltage differential (HVD) but often shorter (e.g., 12 meters for low-voltage differential) due to skew and issues at higher speeds. SAS, however, supports up to 16,384 devices through the use of expanders that route connections in a topology, and it maintains effective distances of up to 10 meters on cables without the skew-related degradation that plagues parallel implementations. This enables larger, more flexible storage configurations in enterprise environments. In terms of speed progression, SAS begins at 3 Gbit/s (approximately 300 MB/s per link), roughly equivalent to the peak of Ultra-320 at 320 MB/s, but SAS avoids the bus contention inherent in parallel designs where all devices share bandwidth, leading to performance degradation as more devices are added. SAS scales efficiently by aggregating multiple links (e.g., wide ports with 4 lanes reaching 1.2 GB/s at 3 Gbit/s generation) and has progressed to higher generations like 12 Gbit/s and beyond without the timing complexities that capped 's evolution. Backward compatibility in SAS is maintained at the protocol level but not physically with . SAS retains the SCSI command set and upper-layer protocols, allowing existing software and applications designed for to operate unchanged on SAS hardware. However, the physical layer differs entirely—SAS uses serial connectors (e.g., SFF-8482), incompatible with 's parallel 68-pin or 50-pin connectors—necessitating new cabling and adapters for migration.

Comparison with SATA

Serial Attached SCSI (SAS) and Serial ATA (SATA) are both serial interfaces for connecting storage devices, but SAS is designed for enterprise environments requiring high reliability and scalability, while SATA targets consumer and lower-end applications focused on cost efficiency. SAS leverages the full command set, enabling advanced features such as reservations for controlling shared access to devices in multi-initiator setups, error recovery mechanisms, and block reclamation, which are essential for mission-critical operations. In contrast, SATA employs the ATA command set, which lacks these enterprise-oriented capabilities and is optimized for simpler, single-host interactions. In terms of connectivity, SAS supports dual-port configurations on drives, allowing redundant paths and simultaneous access from multiple initiators within a domain, which enhances in clustered systems. SATA drives, however, feature single-port designs limited to one host connection, without native support for multi-initiator environments. SAS further incorporates expanders that function as intelligent switches, enabling point-to-point topologies with up to thousands of devices in a single domain, whereas relies on direct point-to-multipoint connections without switching capabilities. Performance differences stem from SAS's emphasis on reliability over distance and . SAS employs higher signaling voltages—typically around 850 mV differential compared to SATA's 500 mV—which reduces susceptibility to noise and supports longer cable runs of up to 10 meters, ideal for rack-mounted enterprise setups. SATA, with its lower voltage, is constrained to 1-meter internal cables to maintain signal quality, limiting its use in expansive configurations. Both interfaces share a common serial derived from serial ATA signaling, but SAS's full-duplex operation and robust error handling provide superior sustained throughput in demanding workloads. SAS offers backward compatibility with SATA through the Serial ATA Tunneling Protocol (STP), allowing SAS hosts and expanders to directly connect and communicate with SATA drives by encapsulating ATA commands within SAS frames. The reverse is not possible, as SATA controllers cannot handle SAS drives due to incompatible signaling voltages and protocol requirements, preventing damage but enforcing one-way interoperability. This design makes SAS more versatile for mixed environments, though it comes at a higher cost—enterprise SAS components and drives are typically 2-10 times more expensive than equivalent SATA options, justified by enhanced durability features like higher MTBF ratings and 24/7 operation support.

Variants and Applications

Nearline SAS

Nearline SAS (NL-SAS) is a variant of Serial Attached SCSI (SAS) hard disk drives designed for cost-optimized, high-capacity nearline storage, utilizing the SAS interface on slower-spinning platters typically operating at 7,200 RPM to balance capacity and efficiency in secondary storage tiers. These drives support capacities up to 24 TB per unit, such as Seagate's Exos X24 model in helium-sealed 3.5-inch form factors, enabling massive scale-out environments. Unlike high-performance SAS drives, which prioritize speed with 10,000 RPM or 15,000 RPM spindles for random access workloads, NL-SAS emphasizes sequential data handling with lower latency variance and reduced power draw, providing improved watts per terabyte efficiency. Key features of NL-SAS include dual-porting for , full compatibility with the command set for enterprise management, and support for SAS topologies like expanders, allowing seamless integration into existing SAS infrastructures without performance bottlenecks from interposers or adapters. Reliability metrics, such as a (MTBF) of 2.5 million hours and (AFR) of 0.35%, ensure durability in demanding environments, while options like self-encrypting drives () enhance . These attributes make NL-SAS a bridge between enterprise SAS and consumer , offering superior and at a modest cost premium over equivalents. In data centers, NL-SAS drives are primarily deployed for archival storage, disk-to-disk backups, and bulk data retention, where high-capacity needs outweigh the demand for ultra-low latency. Applications include virtualization platforms, cloud storage tiers, and media repositories, supporting sequential workloads like large file streaming or long-term retention with throughputs optimized for 12 Gb/s SAS interfaces as of 2025. Compared to standard SAS, NL-SAS reduces operational costs through lower RPM-induced power savings—typically around 5.4 W idle—with topology flexibility for up to thousands of drives in expanded arrays.

Integration with SSDs and Enterprise Storage

Serial Attached SCSI (SAS) solid-state drives (SSDs) are engineered for enterprise environments, offering high endurance ratings typically ranging from 1 to 3 drive writes per day (DWPD) over a five-year period, which supports intensive write workloads in mission-critical systems. These drives provide capacities from 1.92 TB up to 30.72 TB, enabling dense storage configurations in data centers. Performance metrics include random read exceeding 700,000 and sequential throughput up to 4,200 MB/s on 24G SAS interfaces, making them suitable for latency-sensitive operations. In enterprise applications, SAS SSDs power all-flash arrays, where their dual-port architecture facilitates and by allowing simultaneous connections to multiple host bus adapters, minimizing downtime during path failures. They integrate seamlessly into (HCI) platforms, such as those supporting environments, by providing consistent low-latency access for virtualized workloads. For database systems, including (OLTP), SAS SSDs deliver the endurance and required to handle high-volume read-write operations without degradation. The SAS SSD market, as a segment of the broader enterprise SSD sector, is projected to contribute to the overall enterprise SSD market reaching $32 billion in 2025, with a (CAGR) of 15.5% through 2030, fueled by surging demand from AI and workloads that require rapid data access and high reliability. Hybrid storage systems leverage SAS expanders to combine SSDs for tiers with traditional SAS hard disk drives (HDDs) for capacity tiers, enabling automated data placement in tiered architectures that optimize cost and speed for enterprise storage pools. This approach supports scalable enclosures where SSDs accelerate hot data while HDDs store archival content, maintaining SAS protocol compatibility across mixed media.

Modern Context and Future

Comparison with NVMe

Serial Attached SCSI (SAS) and Non-Volatile Memory Express (NVMe) represent two distinct approaches to enterprise storage interfaces, with SAS relying on dedicated serial links that employ the protocol for point-to-point or expander-based connections to multiple devices. This architecture supports dual-port configurations for enhanced and is optimized for environments requiring robust, scalable cabling over distances up to 10 meters using or longer with fiber optics. In contrast, NVMe operates as a protocol over the PCIe bus, providing a direct path to the CPU but sharing bandwidth with other high-performance components such as GPUs, which can introduce contention in densely populated systems. Performance differences stem from their architectural designs, particularly in latency and command queuing. SAS SSDs typically deliver end-to-end latencies of 150–200 μs, influenced by the command overhead and single-queue model that limits concurrency to a depth of 256 commands per device. NVMe, however, achieves lower latencies of around 80–100 μs through reduced protocol overhead and parallel processing, scaling to 64,000 queues with up to 64,000 commands each, enabling significantly higher —often exceeding 500,000 for random 4K operations compared to SAS's 200,000–300,000. These attributes make NVMe particularly effective for workloads demanding massive parallelism, while SAS suffices for sequential or moderately concurrent access patterns. In terms of use cases, SAS persists in legacy storage area networks (SANs) and expander topologies for enterprise arrays, where its compatibility with both SSDs and HDDs, along with features like hot-swapping and , supports tiered storage in traditional data centers. NVMe, by comparison, dominates in hyperscale environments, powering high-IOPS applications such as databases, , and AI training in facilities operated by providers like cloud hyperscalers. Coexistence is facilitated by SAS-to-NVMe bridges and adapters that allow NVMe drives to integrate into SAS backplanes, while SAS remains the choice for HDD-based systems where NVMe's capabilities would be overprovisioned and cost-ineffective.

Recent Developments and Outlook

In 2024, the Trade Association (STA) and INCITS/ unveiled enhancements to the 24G SAS standard, including command duration limits to optimize long-running operations, logical depopulation for efficient drive management, rebuild assist features tailored for SSDs to accelerate recovery processes, and persistent connections to maintain stable links in dynamic environments. These updates build on SAS-4's 22.5 Gbit/s baseline, enabling better support for high-density storage in AI-driven workloads by reducing latency and improving resource utilization in enterprise arrays. Approximately 39% of SAS storage module manufacturers introduced 24G SAS-compatible products in 2024, focusing on hyperscale centers and HDD-optimized configurations for cost-effective scaling. The SNIA's 2024 SAS roadmap extends through 2027, emphasizing sustained evolution for hybrid storage ecosystems, with features like enhanced error handling and improvements to address growing data volumes from AI inference and . While direct integration with (CXL) remains limited, SAS's dual-port architecture complements CXL-extended memory pools in disaggregated systems, facilitating memory-intensive AI tasks without full protocol overhauls. Despite NVMe's dominance in all-flash arrays, SAS maintains a strong foothold in edge and hybrid cloud deployments, particularly for HDD-centric archival and mixed workloads, with connected storage capacity projected to grow steadily. The SAS SSD market, a key indicator, is expected to expand from $4.71 billion in 2025 to $10 billion by 2035, driven by demand in reliable, multi-initiator environments. Key challenges include intensifying competition from NVMe over Fabrics (NVMe-oF), which offers lower latency for networked storage and is increasingly adopted in AI clusters, potentially eroding SAS's share in shared infrastructures.

References

Add your contribution
Related Hubs
User Avatar
No comments yet.