OCZ Storage Solutions Saber 1000 SB1CSK31MT5D0-0960 Benutzerhandbuch
Produktcode
SB1CSK31MT5D0-0960
2
Product Brief | Saber 1000 Series | V1.1 | © OCZ Storage Solutions – A Toshiba Group Company
OCZ may make changes to specifications and product descriptions at any time, without notice. For the latest version, visit ocz.com.
OCZ may make changes to specifications and product descriptions at any time, without notice. For the latest version, visit ocz.com.
Superior Sustained Performance
Utilizing cutting-edge A19nm NAND flash from Toshiba coupled with OCZ’s own Barefoot 3 controller, Saber
1000 delivers consistent, sustained I/O across all capacities to meet the demands of enterprise applications
and hyperscale workloads. The performance metrics for a 960GB capacity Saber 1000 include read bandwidth
of up to 550 MB/s, write bandwidth of up to 470 MB/s, 4KB random read speed of up to 98,000 input/output
operations per second (IOPS), and 4KB random write speed up to 20,000 IOPS.
1000 delivers consistent, sustained I/O across all capacities to meet the demands of enterprise applications
and hyperscale workloads. The performance metrics for a 960GB capacity Saber 1000 include read bandwidth
of up to 550 MB/s, write bandwidth of up to 470 MB/s, 4KB random read speed of up to 98,000 input/output
operations per second (IOPS), and 4KB random write speed up to 20,000 IOPS.
Power Loss Management
In the event of a sudden power loss, the Saber 1000 Series uses a technique referred to as Power Failure
Management Plus (PFM+) which holds up the SSD circuitry long enough to ensure the integrity of the device
so that it can be fully operational again once power is restored. To achieve this, each Saber 1000 SSD uses
an internal 16 Byte PFM+ log that describes a single action performed on the NAND and this log contains
only the essential data needed to rescue the SSD from metadata corruption. As a result, the PFM+ logs
are accumulated and saved frequently with no impact on overall SSD performance. As most enterprise and
hyperscale datacenter servers are deployed in clusters, PFM+ ensures that once the power is restored, and the
server cluster stabilizes again, all Saber 1000 SSDs will return back to the condition they were in at the time of
failure.
Management Plus (PFM+) which holds up the SSD circuitry long enough to ensure the integrity of the device
so that it can be fully operational again once power is restored. To achieve this, each Saber 1000 SSD uses
an internal 16 Byte PFM+ log that describes a single action performed on the NAND and this log contains
only the essential data needed to rescue the SSD from metadata corruption. As a result, the PFM+ logs
are accumulated and saved frequently with no impact on overall SSD performance. As most enterprise and
hyperscale datacenter servers are deployed in clusters, PFM+ ensures that once the power is restored, and the
server cluster stabilizes again, all Saber 1000 SSDs will return back to the condition they were in at the time of
failure.
Central Management
OCZ’s StoragePeak 1000 SSD Management console enables IT managers to monitor and administer Saber
1000 SSDs as well as other OCZ enterprise SSD resources from a unified web-based user interface. This
network-accessible management console securely connects to multiple hosts (running Linux or Windows
operating systems) providing a cross-platform view of the OCZ SSDs installed in servers, storage arrays or
appliances, and includes a configurable alerting feature that enables IT to launch corrective actions to remedy
any issues at an early stage.
1000 SSDs as well as other OCZ enterprise SSD resources from a unified web-based user interface. This
network-accessible management console securely connects to multiple hosts (running Linux or Windows
operating systems) providing a cross-platform view of the OCZ SSDs installed in servers, storage arrays or
appliances, and includes a configurable alerting feature that enables IT to launch corrective actions to remedy
any issues at an early stage.
Supported Applications
The Saber 1000 Series is targeted for read-intensive applications such as read cache & indexing, Video On
Demand (VoD), Virtual Desktop Infrastructure (VDI), media streaming, front-end Web server, cloud infrastructure,
online archiving, Customer Relationship Management (CRM), and Enterprise Content Management (ECM), to
name a few.
Demand (VoD), Virtual Desktop Infrastructure (VDI), media streaming, front-end Web server, cloud infrastructure,
online archiving, Customer Relationship Management (CRM), and Enterprise Content Management (ECM), to
name a few.
Specifications
240GB
480GB
960GB
Sustained Performance
Sequential Read (128KB, QD32)
550 MB/s
550 MB/s
550 MB/s
Sequential Write (128KB, QD32)
515 MB/s
500 MB/s
470 MB/s
Random Read (4KB, QD32)
86,000 IOPS
95,000 IOPS
98,000 IOPS
Random Write (4KB, QD32)
20,000 IOPS
23,000 IOPS
20,000 IOPS
Random Mixed R65%/W35% (4KB, QD32)
31,000 IOPS
33,000 IOPS
30,000 IOPS
Random Read Latency (4KB, QD1)
121 µs
92 µs
59 µs
Random Write Latency (4KB, QD1)
52 µs
27 µs
28 µs
Drive Writes Per Day (DWPD)
0.5
0.5
0.5
Sustained performance measured using IOmeter with Queue Depth 32. Measurements are performed using SNIA based metrics.