Lacie FASTKEY User Manual

WHITE PAPER: LACIE FASTKEY USB 3.0 SSD
Technical Brief
This white paper discusses the advantages of solid-state drives over traditional hard disk drives, focusing on LaCie’s first implementation of this emerging technology, the FastKey USB 3.0.
To illustrate the benefits of SSD, benchmark tests comparing the FastKey with other LaCie products were conducted.
LaCie White Paper: FastKey USB 3.0 SSD
WHAT IS AN SSD?
SSD means Solid State Drive. An SSD has no mov­ing parts and is essentially an HDD emulator based on flash memory. It is comprised of a printed circuit board, a set of NAND flash memory chips, SDRAM cache, a memory controller, an interface controller, and an interface connector such as IDE, SATA, SAS, USB, or even fiber channel.
In addition to the performance and energy advan­tages of SSD, the lack of moving parts means that the drives can withstand high vibration and shock. In fact, some SSDs can withstand shock up to 1500G, the equivalent of a drop from 26 feet.
Basic specifications of Solid State Drives are:
MTBF 1,000,000 hours
Data Integrity 10 years
Shock (operating) 1500G, 3 axes
Vibration (operating) 16G, each axis
Operating Temperature 0°C to 70°C
The LaCie FastKey is an SSD USB 3.0-connected device. The FastKey uses MLC NAND flash memory chips and two main chipsets:
Indylinx barefoot: SSD controller Symwave 6316: USB 3.0 controller
MLC OR SLC, WHAT IS THE DIFFERENCE?
Single-level cell (SLC) and multi-level cell (MLC) Flash memory are similar in their design. MLC Flash devices usually cost less and allow for higher storage density. SLC Flash devices provide faster write per­formance and greater reliability, even at high indus­trial temperatures above the operating range of MLC Flash devices. Speed performances between SLC and MLC are comparable.
The endurance of SLC Flash is around five times that
of MLC Flash. The endurance of MLC Flash decreas­es during the product’s life. This is a main reason why SLC Flash is considered industrial-grade Flash and MLC Flash is considered consumer-grade Flash. MLC lifetime is limited to 1,000,000 “Programmed / Erased” cycles (10,000 cycles per cell).
HARD DRIVE DISK OR SOLID STATE DISK?
Advantages of SSD over HDD
Faster start-up because no spin-up is required.
Fast random access because there is no “seek-
ing” motion as is required with rotating disk plat­ters and the read/write head and head-actuator mechanism
Low read latency times for RAM drives. In ap-
plications where hard disk seeks are the limiting factor, this results in faster boot and application launch times.
Consistent read performance because physical
location of data is irrelevant for SSDs.
File fragmentation has negligible effect because
data access degradation due to fragmentation is primarily due to much greater disk head seek activ­ity, as data reads or writes are spread across many different locations on the disk; SSDs have no heads and thus no delays due to head motion (seeking).
Silent operation due to the lack of moving parts.
SSDs typically have lower power consumption
than HDDs.
High mechanical reliability, as the lack of mov-
ing parts almost eliminates the risk of mechani­cal failure.
Ability to endure extreme shock, high altitude,
vibration, and extremes of temperature.
Immune to magnets.
For low-capacity SSDs (like the LaCie FastKey),
lower weight and size: although size and weight
per unit storage are still better for traditional hard drives.
Failures occur less frequently while writing/eras-
ing data, which means there is a lower chance of irrecoverable data damage.
SSDs are random access by nature and can per-
form parallel reads on multiple sections of the drive (as opposed to a HDD, which requires seek time for each fragment, assuming a single head assembly).
Can also be configured to smaller form factors
and reduced weight.
DISADVANTAGES OF SSD COMPARED TO HDD:
Flash-memory drives have limited lifetimes and
will often wear out after 1,000,000 P/E cycles (10,000 per cell) for MLC, and up to 5,000,000 P/E cycles (100,000 per cell) for SLC.
SSDs using wear leveling cannot be defragment-
ed in order to provide maximum sequential read speed. Optimizations do not work efficiently if files are fragmented (access time of flash-based SSDs is about 0.1 ms).
Wear leveling used by most SSDs intrinsically in-
duces fragmentation. Moreover, defragmenting a SSD by a defragmenter is harmful since it adds wear to the SSD for no benefit.
As of 4Q 2010, SSDs are still much more ex-
pensive per gigabyte than hard drives. Whereas hard drives are around US$0.10 per gigabyte for 3.5”, or US$0.20 for 2.5”, a typical flash drive is closer to US$3 per gigabyte in 2010
The maximum capacity of SSDs is currently lower
than that of hard drives.
SSD write performance is significantly impacted
by the availability of free, programmable blocks. Previously written data blocks that are no longer in use can be reclaimed by TRIM; however, even with TRIM, fewer free, programmable blocks translate into reduced performance.
As a result of wear leveling and write combin-
ing, the performance of SSDs degrades with use. However, most modern SSDs now support the TRIM command and thus return the SSD back to its factory performance when using OSes that support it like Windows 7, Windows Server, 2008 and Linux.
WHAT IS USB 3.0?
The Universal Serial Bus (USB) 3.0 specification is a new industry-standard peripheral connection tech­nology, developed by USB Implementors forum, for connecting peripherals to PCs and laptops.
The USB 3.0 specification draws from the same ar­chitecture of the wired USB specification and there­fore is a backward-compatible standard with the same ease-of-use and plug-and-play capabilities of previous USB technologies, but with a 10 times per­formance increase and lower power consumption. The USB 3.0 specification uses two additional high­speed differential pairs for SuperSpeed mode, which boosts its bandwidth to 5 GB/s.
For end-users of the USB 3.0 specification, the goals of connecting peripherals with PCs or laptops are still the same as the Hi-Speed (USB 2.0) specification, but with significantly increased speed and reduced power consumption.
The SuperSpeed USB specification, therefore, is not simply an upgrade to earlier versions of the USB 2.0 specification. Due to the broad deployment of USB
2.0 devices in the market, SuperSpeed USB devices need to be backward-compatible, but the backward­compatibility portion of the SuperSpeed USB specifi­cation targets only the device drivers and connector architecture. The higher speed and reduced power consumption for the USB 3.0 specification uses ad­vanced mechanisms and techniques similar to ones that were used for other high bandwidth interfaces, such as the PCI Express (PCIe) specification. As a result, the SuperSpeed USB specification has many differences compared to earlier generations of USB specifications (1.1/2.0/OTG).
Page 2 Page 3
LaCie White Paper: FastKey USB 3.0 SSD
DIFFERENCES BETWEEN SUPERSPEED USB 3.0 AND HI-SPEED USB 2.0
The SuperSpeed USB specification is similar to earlier USB versions in terms of the connector and device drivers. The end-user and device driver engineer may find SuperSpeed USB similar to earlier versions, but it is significantly different to implementors of Super­Speed USB host and devices.
At a mechanical level, the SuperSpeed USB specifi­cation supports dual-bus architecture for backward compatibility to a USB 2.0 device.
This means that the SuperSpeed USB cable needs to support eight primary wires, two wires for USB 2.0 con­nectors, two shared between the USB 2.0 and Super­Speed USB specifications (PWR and GND), and four for SuperSpeed USB dual-simplex differential signals.
SuperSpeed USB 3.0 Hi-Speed USB 2.0
Dual-simplex, unicast protocol Half-duplex, broadcast protocol
Uses asynchronous notification (NRDY, ERDY) Uses polling mechanism
Supports streaming for bulk transfers Does not support streaming
Many changes were required in the existing USB 2.0 data flow to maximize the advantages of the Super­Speed USB bi-directional dual-simplex data inter­face. Though the SuperSpeed USB specification is still a host-directed protocol and preserves the con­cepts of endpoints, pipes transfer types, etc., the traf­fic flow has changed to asynchronous as opposed to polling traffic flow in previous USB specifications. In addition, there are many fundamental differences at the Protocol level as shown in the inset table.
SuperSpeed USB 3.0 also manages power consump­tion more efficiently, which results in some differences at the Protocol level:
1. SuperSpeed supports link-level power manage­ment, which means either a host or a device can initiate link power management. In USB 2.0, it is always initiated by the host.
2. SuperSpeed USB allows isochronous devices to enter in the low power link states between ser-
THE WEAR LEVELING FUNCTION FOR SSD
MLC lifetime is limited to 1,000,000 “Programmed
/ Erased” cycles. LaCie Fastkey uses special static
wear levelling in order to mitigate this problem by
spreading writes over the entire device (and not al­ways on the same memory blocks).
Static wear levelling uses a map to link the Logical Block Addresses (LBAs) from the Operating System to physical memory addresses. Each time the OS writes replacement data, the map is updated so the original physical block is marked as invalid data, and a new block is linked to that map entry. Each time a block of data is re-written to the Flash memory it is written to a new location. This “rotational” effect enables the SSD to operate until most of the blocks are near their end of life.
TRIM FUNCTION
1
FOR SSD (TO
Pages to delete
RESIST WRITE
via TRIM
PERFORMANCE DECLINE)
2
512KB blocks. This is important because this is the smallest structure that can be erased. You can read and write at a page level, but you can only erase an entire 512KB block. This means that you can read 4KB at a time and write 4KB at a time to an empty space, but you can’t overwrite a page. You must first erase the content.
Windows 7 and Windows Server 2008 R2 support the TRIM function, which they use when they detect that a file is being deleted from an SSD. When the OS deletes a file on an SSD, it updates the file system but also tells the SSD via the TRIM command which pages should be deleted. At the time of the delete, the SSD can read the block into memory, erase the block, and write back only pages with data in them, as illustrated below. The delete is slower, but you get better performance for future writes because the pages are already empty, and write performance is generally considered the most important factor.
NAND Block
Supports continuous bursting Does not support bursting
For OUT, token is integrated into data OUT is three separate parts (Token, Data, and
Handshake)
For IN, token is replaced by Handshake IN is three separate parts (Token, Data, and
Handshake)
Split error protection, recovery, and flow control functionality between protocol layer and link layer
The SuperSpeed USB specification supports a dual­simplex data interface with four differential wires for simultaneous data flow in both directions. It should be noted that adding a bi-directional data interface was necessary to support the SuperSpeed USB speci­fication’s target speed because the halfduplex, two wire differential signals of USB 2.0 and unidirection­al data flow were not enough to support the Super­Speed USB specification’s high bandwidth.
Protocol layer manages error detection, recovery, and flow control functionality
vice intervals. This mechanism is not supported in USB 2.0.
3. SuperSpeed USB allows devices to inform the host of their latency tolerance using the Latency Tolerance Messaging mechanism.
This allows the host to enter in low power states for better power performance.
TRIM is a useful command for Linux 2.6.33, Windows 7, and Windows Server 2008. The La­Cie FastKey supports TRIM which improves performance when you delete files to prepare the space for future writing. If you overwrite an existing file, TRIM doesn’t help and you’ll get the same write per­formance as without TRIM.
SSDs behave very differently from traditional mechanical, platter­based hard disks. SSDs are made up of cells. These cells are or­ganized into pages, the smallest readable/writable unit in most SSDs, and are normally 4KB.
These pages are then organized into blocks, tra­ditionally 128 pages per block, for a block size of
3
4
Copy
Memory Cache
ERASE BLOCK
Copy
Page 4 Page 5
Loading...
+ 6 hidden pages