apogeeweb title
Home  Memory

Sep 3 2020

Types of Hard Disk Drive Interface

Introduction

The hard disk interface is the connecting part between the hard disk and the host computer system, and its function is to transmit data between the hard disk cache and the host memory. Different hard disk interfaces determine the data transmission speed between the hard disk and the computer. In the entire system, the quality of the hard disk interface directly affects the speed of the program and the performance of the system. From this article, you can understand the connector concepts of IDE, SATA, SCSI, Fibre Channel (FC) and SAS and their development process, and finally two important interface protocols: AHCI and NVMe.

SAS, SATA, SCSI, FC, and IDE Explained

Catalog

Introduction

Ⅰ Bus Interface Types

Ⅱ What is IDE?

2.1 Integrated Drive Electronics Definition

2.2 IDE Mode

2.3 IDE Advantages and Disadvantages

Ⅲ What is SCSI?

3.1 SCSI Basics

3.2 SCSI Version

Ⅳ What is Fiber Channel (FC)?

4.1 Overview of Fibre Channel

4.2 Fibre Channel Protocol

Ⅴ What is SATA?

5.1 Serial ATA Definition

5.2 SATA Interface

5.3 IDE vs SATA Interface

Ⅵ What is M.2?

Ⅶ What is SAS?

Ⅷ Tech Guide: AHCI and NVMe Protocol

8.1 AHCI Protocol

8.2 NVMe Protocol

8.3 Tech Note


Ⅰ Bus Interface Types

From an overall point of view, hard disk interfaces are divided into five types: parallel ATA (PATA, also called IDE or EIDE), SATA, SCSI, Fibre Channel, and SAS. IDE is mostly used in household products. And some of it are used in website servers. SCSI is mainly used in the server market. While Fibre Channel is only used in high-end servers and is expensive. SATA is now the mainstream hard disk. Most notebook computers and desktop computers use it, and solid state hard drives also use it. SAS is generally used in servers. It has fast transmission speed and strong reliability. Under the broad categories of IDE and SCSI, there has a variety of specific interface types that can be divided according to different technical specification and transmission speed.

 

Ⅱ What is IDE?

2.1 Integrated Drive Electronics Definition

The full name of IDE is "Integrated Drive Electronics", and its original meaning refers to the hard disk drive that integrates the "hard disk controller" and the "disk body". This approach reduces the number and length of cables for the hard disk interface, enhances the reliability of data transmission, and makes hard disk manufacturing easier. Many hard disks used to have IDE interfaces, but now almost all hard disk interfaces are standard with SATA.

IDE represents a type of hard disk interfaces, but in actual applications, people are also used to call it the first IDE-type hard disk ATA-1. This type of interface has been eliminated with the development of technology. With the time passes, more types of hard disk interfaces are developed, such as ATA, Ultra ATA, DMA, Ultra DMA and other interfaces are all IDE hard disk interfaces.

IDE Definition

2.2 IDE Mode

There are three transmission modes for IDE: PIO (Programmed I/O), DMA (Driect Memory Access), and Ultra DMA (UDMA).

The biggest drawback of PIO mode is that it consumes a lot of CPU resources. The IDE interface running in PIO mode has a data transfer rate ranging from 3.3MB/s (PIO mode 0) to 16.6MB/s (PIO mode 4). There are two types of DMA modes: Single-Word DMA and Multi-Word DMA. The highest transfer rate of Single-Word DMA mode is 8.33MB/s, and Multi-Word DMA (Double Word) can reach 16.66MB/s. The biggest difference between the DMA and the PIO is that the DMA mode does not rely too much on CPU instructions to run, which can save the processor's operating code.

Due to the emergence and rapid popularity of the UDMA mode, PIO and DMA are immediately replaced by UDMA. UDMA is a standard protocol under the Ultra ATA system, which is based on the 16-bit Multi-Word DMA mode. One of the advantages of UDMA is that in addition to the advantages of DMA mode, it also applies CRC (Cyclic Redundancy Check) technology to enhance the performance of error detection and debugging during data transmission. Since the introduction of the Ultra ATA standard, its interface has applied DDR (Double Data Rate) technology to double the transmission speed, with a transmission speed of up to 100MB/s.

 

2.3 IDE Advantages and Disadvantages

  • Advantages: Compatible and cost-effective.
  • Disadvantages: slow data transmission speed, short cable length, fewer connected devices, no support for hot swap, poor upgrading ability of interface speed.

 

Ⅲ What is SCSI?

3.1 SCSI Basics

The full name of SCSI is "Small Computer System Interface", which is a completely different interface from IDE. SCSI is not specifically designed for hard disks, but a high-speed data transmission technology widely used in minicomputers. It has the advantages of wide application range, multitask, large bandwidth, low CPU occupancy rate, and hot swap. So SCSI is mainly used in medium and high-end servers and high-end workstations. But the higher price makes it difficult to popularize like IDE. SCSI also has some potential problems. It has limited system BIOS support, and it has to be set for each computer. There's also no common SCSI software interface.

SCSI Interface

Figure 1. SCSI Interface

3.2 SCSI Version

SCSI Version

Descriptions

SCSI-1

developed in 1986

(obsolete)

Introduced in 1979, supported synchronous and asynchronous SCSI peripherals.

SCSI-2

adopted in 1994

Introduced in 1992, also known as fast SCSI, supported any SCSI device.

SCSI-3

debuted in 1995

It is the standard currently in use.


 

Ⅳ What is Fiber Channel (FC)?

4.1 Overview of Fibre Channel

Fibre Channel is the same as SCSI. It is not originally an interface technology developed for hard disk design and development, but specifically designed for network systems. However, as storage systems develop, they are gradually applied to hard disk systems. Fibre Channel is developed to improve the speed and flexibility of multi-disk storage systems. And it greatly improves the communication speed of multi-disk systems. The main characteristics of Fibre Channel are: hot swap, high-speed bandwidth, remote connection, large number of connected devices, etc. It can meet the high data transmission rate requirements of high-end workstations, servers, mass storage sub-networks, and peripherals for bidirectional and serial data communication through hubs, switches and point-to-point connection.

Fiber Channel FC Controller

4.2 Fibre Channel Protocol

Fiber Channel Protocol

Descriptions

FC-0

Physical layer, customizes different media, set transmission distance and signal mechanism standards, defines optical fiber, copper interfaces, and cable indicators

FC-1

Encode/Decode

FC-2

Framing protocol /flow control

FC-3

common services such as data encryption and compression

FC-4

Protocol mapping layer, which defines the interface between fibre channel and upper-layer protocol. Upper-layer applications such as SCSI protocol, HBA FC-4 interface functions. FC-4 supports multiple protocols, such as FCP-SCSI, FC-IP, and FC-VI.


 

Ⅴ What is SATA?

5.1 Serial ATA Definition

SATA stands for "Serial Advanced Technology Attachment" or "Serial ATA". It is an interface used to connect ATA hard drives to a computer's motherboard. SATA adopts serial connection mode. Serial ATA bus uses embedded clock signal, which has stronger error correction ability. Compared with the past, its biggest difference is that it can check transmission instructions (not just data). Errors are automatically corrected, which greatly improves the reliability of data transmission.

Now the general interface is SATA interfaces. The reason why it can replace IDE is because the performance of the SATA is much better than that of the IDE. SATA speed is also much higher than IDE, and it also supports hot swap/hot-plugging.

Serial ATA Cable

5.2 SATA Interface

Most of the computers we use are also SATA interfaces. The current SATA interface has three versions 1.0, 2.0, and 3.0. The larger the version number, the later it appears, the better the performance, which mainly due to the faster data transfer rate. SATA 3.0 is the most common interface used today, though there have been four revisions since its introduction, namely 3.1 through 3.4. The SATA interface version is backward compatible, and the higher version is compatible with the lower. Some SATA hard disks provide jumpers. Due to the jumper settings are different, the version number of the SATA interface of the same hard disk is different. In addition, the actual transfer rate of the interface requires the support of the SATA motherboard.

SSD has better performance, smaller size, and higher interface requirements. High-performance SSDs have basically switched to M.2, U.2 and PCIe, but SATA interfaces will not be eliminated in a short time. It is still in mainstream market, especially the HDD market. The SATA 3.3 specification upgraded by SATA-IO also brings some new features, optimizing the support of SMR, and can be powered off remotely. SMR shingled magnetic recording technology can increase the storage density of HDD by 25%.

The SATA interface entered the 6Gbps era from the 3.0 standard in 2009. In 2011, SATA 3.1 was updated, SATA 3.2 was updated in 2013, and then SATA 3.3 was updated in 2016. These subversion upgraded have not brought many new functions. After all, the bottleneck of HDDs is not the speed, and it is difficult to make big improvements from the interface.

 

5.3 IDE vs SATA Interface

SATA hard disk has a new design structure, fast data transmission, save space, and many other advantages over IDE hard disk:

1) SATA hard disk has a higher transmission speed than IDE hard disk. SATA can provide a peak transfer rate of 150MB/s. It will reach 300 MB/s and 600 MB/s with the development. At that time, we will get a transfer rate nearly 10 times faster than IDE hard drives.

2) Compared with the PATA40-pin data cable of IDE hard disks, the SATA cable is small and thin. And the transmission distance is long, which can be extended to 1 meter, making it easier to install equipment and wiring in the machine. Because the size of the connector is small, this kind of cable effectively improves the air flow inside the computer and also speedsthe heat dissipation in the case.

3) Thepower consumption has been reduced. SATA hard drives can work with 500 mA of current.

4) SATA can be backward compatible with PATA devices by using multi-purpose chipsets or serial-parallel converters. Since SATA and PATA can use the same drive, there is no need to upgrade or change the operating system.

5) SATA does not need to set the master and slave disk jumpers. The BIOS will number it in the order of 1, 2, 3. Whilethe IDE hard disk needs to set the master and slave disks through jumpers.

6) SATA also supports hot plugging and can be used like a U disk. IDE hard disks do not support hot swap.

 

Ⅵ What is M.2?

The M.2 interface is a new interface specification. It is a new standard tailored for Ultrabooks to replace the original mSATA interface. Whether it is a smaller size or higher transmission performance, M.2 is far better than mSATA.

M.2 interfaces are generally divided into two types. When buying M.2 SSDs, you need to pay attention to internal agreements. One is to use the traditional SATA AHCI protocol, which has no difference in performance with ordinary SATA solid hard drives; another is to use the brand-new NVMe protocol, which can provide SSD performance up to 3000MB/s or more.

 

Ⅶ What is SAS?

SAS (Serial Attached SCSI) is a new generation of SCS technology, which is the same as the current popular SATA technology. It uses serial technology to obtain higher transmission speed and improves internal space by shortening the cable. SAS is a new interface developed after the parallel SCSI. This interface is designed to improve the performance, availability, and expandability of the storage system, and to provide compatibility with the SATA.

SAS technology can be backward compatible with SATA. Specifically, the compatibility of the two is mainly reflected in the compatibility of the physical part and the protocol. At the physical layer, the SAS interface and the SATA interface are fully compatible, and the SATA hard disk can be directly used in the SAS environment. In terms of interface standards, SATA is a sub-standard of SAS, so the SAS controller can directly control SATA hard drives, but SAS cannot be directly used in the SATA environment. Because the SATA controller cannot control the SAS hard disk. As for the protocol, SAS is composed of three types of protocols, which use corresponding protocols for data transmission according to different devices. Among them, the serial SCSI protocol (SSP) is used to transmit SCSI commands, the SCSI management protocol (SMP) is used to maintain and manage connected devices, and the SATA channel protocol (STP) is used to transfer data between SAS and SATA. Therefore, under the cooperation of three protocols, SAS can seamlessly work with SATA and some SCSI devices.

SAS controller

The backplane of the SAS system can be connected to dual-port, high-performance SAS drives and high-capacity, low-cost SATA drives. So SAS drives and SATA drives can exist in a storage system at the same time. But it should be noted that the SATA system is not compatible with SAS, so SAS drives cannot be connected to the SATA backplane. Due to the compatibility of the SAS system, users can use hard drives with different interfaces to meet the capacity or performance requirements of various applications. So they have more flexibility when expanding the storage system, allowing storage devices to maximize application benefits.

In the system, each SAS port can connect up to 16256 external devices, and SAS adopts a point-to-point serial transmission directly with a transmission rate of up to 3Gbps. It is estimated that there will be 6Gbps or even 12Gbps high-speed interfaces in the future. The SAS interface performance has also been greatly improved. It also provides 3.5-inch and 2.5-inch interfaces, so it can meet the requirements of different server environments. SAS relies on SAS expanders to connect more devices. Most expanders have 12 ports.

Compare with the traditional parallel SCSI, SAS has a significant increase in interface speed. With the use of serial cables, it not only can achieve a longer connection distance, but also improve the anti-interference ability. In addition, this cable can also significantly improve the heat dissipation inside the chassis.

 

Ⅷ Tech Guide: AHCI and NVMe Protocol

Here we will focus on the AHCI protocol and NVMe protocol of the solid state drive (SSD).

There are two mainstream transmission protocols for SSD (Solid State Drive): One is the AHCI protocol, and the other is the NVMe protocol.

8.1 AHCI Protocol

Advanced Host Controller Interface (AHCI), sets the operation of Serial ATA (SATA) host controllers in a non-implementation-specific manner in its motherboard chipsets. That is, AHCI allows storage drivers to connect advanced SATA functions. When we use SATA SSD, we must enable AHCI mode in the motherboard settings. This is because when the AHCI mode is turned on, the number of useless seeks of the SSD can be greatly shortened and the data search time can be reduced. So that the SSD under multi-tasking can exert all the performance and effects. According to related performance tests, after the AHCI mode is turned on, the SSD read and write performance is increased by about 30%. However, with the gradual enhancement of SSD performance, these standards have also become a major bottleneck restricting solid state drives. Because the AHCI standard designed for hard disk drives is not suitable for low-latency solid state drives.

8.2 NVMe Protocol

Another transmission protocol is the NVMe protocol that represents the future performance trend. The so-called NVMe protocol is to make full use of the low latency and parallelism of PCI-E channels, greatly improve the read and write performance of SSDs under controllable storage costs. It reduces the high latency caused by the AHCI, and completely liberates ultimate performance of SATA SSD.

NVMe Specification

1.0 (March 2011)

1.1 (October 2012)

1.2 (November 2014)

Fabric's NVMe (2014)

NVM-MI (November 2015)

1.3 (April 2017)

1.4 (July 2019)


Due to the flash memory particles and the main control, the SSD(solid state drives) price with M.2 NVMe protocol is very high, which is about twice the price of SATA SSD. So buy the corresponding level of solid state hard drive based on the configuration and requirements of the computer. Otherwise it will cause performance waste.

In terms of software layer, the delay of NVMe standard is less than half of AHCI. NVMe streamlines the calling method and does not need to read registers when executing commands. Each command of AHCI needs to read registers 4 times, which consumes 8000 CPU times in total loop, causing a delay of about 2.5 ms. NVMe can support receiving commands and prioritizing requests from multi-core processors at the same time.

NVMe has automatic power state switching and dynamic power management functions. The device can switch to Power State 1 after being idle for 50ms from Power State 0. If it continues to be idle, it will enter Power State 2 with lower power consumption after 500ms. There will be a short delay when switching. The SSD can be controlled at a very low level when it is idle. In terms of power management, the NVMe SSD will have a greater advantage than the AHCI SSD. This is important for mobile devices, which can significantly increase the power endurance of notebooks. Moreover, NVMe SSD can be easily matched to different platforms and systems, and can work normally without the corresponding driver provided by the manufacturer. At present, Windows, Linux, Solaris, Unix, VMware, UEFI, etc, support the NVMe SSD.

PCIe SSDs based on the NVMe protocol far exceed the traditional AHCI-based SATA SSDs in terms of performance and practicability. It can be said to be the future of the development of the SSD industry. But the traditional SATA interface will become the first choice for ordinary machine installations under the background of reduced manufacturing costs.

8.3 Tech Note

8.3.1 PCI-E Basic

PCI-E (peripheral component interconnect express) is a high-speed serial computer expansion bus standard. Its original name is "3GIO". It was proposed by Intel in 2001 to replace the old PCI, PCI-X, and AGP bus standard. It belongs to high-speed serial point-to-point high-bandwidth dual-channel transmission. The connected devices allocate exclusive channel bandwidth and do not share bus bandwidth. It mainly supports active power management, error reporting, end-to-end reliable transmission, hot plugging, quality of service ( QOS), and other functions. PCI-E also has a variety of specifications, from PCI-E x1 to PCI-E x32, which can meet the needs of low-speed devices and high-speed devices in a certain period of time in the future.

The PCI-E bus protocol can be directly connected to the CPU with almost no delay, making it an excellent companion to the NVMe standard. In the era of the AHCI standard, the actual performance of PCIe can hardly be exerted due to the agreement limit.

  • Table: PCle Version

Version

Year

Description

PCIe 1.0a

2003

The data rate per channel is 250 MB/s, and the transmission rate is 2.5 GT/s.

PCIe 1.1

2005

The data rate has not changed and it is fully compatible with PCIe 1.0a.

PCIe 2.0

2007

It doubles the transfer rate from PCIe 1.0 to 5 GT/s, and the throughput per channel rises from 250 MB/s to 500 MB/s.

PCIe 2.1

2009

Its speed is the same as PCIe 2.0, supporting troubleshooting system.

PCIe 3.0

2010

Transmitter and receiver equalization, PLL improvements, clock data recovery, and channels are all improved.

PCIe 3.1

2014

Various improvements based on the PCIe 3.0 specifications.

PCIe 4.0

2016

Double the bandwidth provided by PCIe 3.0, maintain software support, and have backward compatibility for the used mechanical interfaces.

PCI-E SD 7.0

2018

A new generation of SD 7.0 standard specifications


 

8.3.2 Interface Size Introduction

The size and application of the hard disk can be divided into:

  • 0.85 inches, mostly used in portable devices such as mobile phones.
  • 1 inch, mostly used in digital cameras (CF type II interface).
  • 1.8 inches, used in some notebook computers and external hard disk enclosures.
  • 2.5 inches, commonly used in notebook computers and external hard disk enclosures.
  • 3.5 inches, mostly used in desktop computers. External hard drive enclosures with 3.5-inch requires an external power supply.

Related Articles

0 comment

Leave a Reply

Your email address will not be published.

 
 
   
Rating: