Manual
7/24/2017
PSFEM6xxxxZxxx
Viking Technology
Revision C
Page 14 of 38
www.vikingtechnology.com
Table 2-3: Random Read/ Write Input/Output Operations per Second (IOPS)
(Iometer 06)
Capacity
Flash Type
Random Read
IOPS
(4K)
Random Write
IOPS
(4K)
8GB
15nm
14200
1
1100
1
16GB
15nm
14200
1
1100
1
32GB
15nm
14200
1
1100
1
64GB
15nm
14200
1
1100
1
128GB
15nm
14200
1
1100
1
256GB
15nm
14200
1100
512GB
15nm
14200
1
1100
1
Notes:
1. Estimated Performance measured using Iometer 06 with queue depth set to 32.
2. Write Cache enabled with DDR cache.
3. Tested with VPFEM5256GZIAMTL (Toshiba MLC L die)
4.
Random IOPS cover the entire range of legal logical block addresses (LBA’s). Measurements are
performed on a full drive (all LBA’s have valid content).
5. Performance may vary by NAND type and host.
6. Refer to Application Note AN0006 for Viking SSD Benchmarking Methodology.
7.
Data is based on SSD’s using Toshiba A15nm NAND devices
8.
L95A data not currently available
2.3 Timing
Table 2-4: Timing Specifications
Type
Average Latency
Power-On-to-Ready (POR)
TBD
Command to DRQ
TBD
Time to Erase (ATA Secure Erase)
TBD
Notes:
1. Device measured using Drivemaster.
2. Sector Read/Write latency measured up to 2048 block transfers (512B/sector = 1 Block)
3. Queue depth set to 32 for NCQ
4.
Sequential IOPS cover the entire range of legal logical block addresses (LBA’s). Measurements are
performed on a full drive (all LBA’s have valid content
5. DRQ (Data Transfer Requested) bit being asserted
2.3.1 STANDBY IMMEDIATE Command
The Power-On-to-Ready time assumes a proper shutdown (power removal
preceded by STANDBY IMMEDIATE command. A STANDBY IMMEDIATE
before power down always performs a graceful shutdown and does not require
the use of the hold-up circuit. Note that SMART attribute 174 "Unexpected Power
Loss" records the number of non-graceful power cycle events.
