Sun SSD Server Platform Bandwidth and IOPS (Speeds & Feeds)
Significance of Results
The Sun SSD (32 GB SATA 2.5" SSD) is the world's first enterprise-quality, open-standard Flash design. Built to an industry-standard JEDEC form factor, the module is being made available to developers and the OpenSolaris Storage community to foster Flash innovation. The Sun SSD delivers unprecedented IO performance, saves on power, space, and cooling, and will enable new levels of server optimization and datacenter efficiencies.
- The Sun SSD demonstrated performance of 98K 4K random read IOPS on a Sun Fire X4450 server running the Solaris operating system.
Performance Landscape
Solaris 10 Results
| Test | SSD Result | ||
|---|---|---|---|
| X4450 | X4140/X4440 | T5240 | |
| Random Read (4K) | 98.4K IOPS | 33.8K IOPS | 71.5K IOPS |
| Random Write (4K) | 31.8K IOPS | 16.6K IOPS | 14.4K IOPS |
| 50-50 Read/Write (4K) | 14.9K IOPS | 13.3K IOPS | 15.7K IOPS |
| Sequential Read (MB/sec) | 764 MB/sec | 463 MB/sec | 1012 MB/sec |
| Sequential Write (MB/sec) | 376 MB/sec | 470 MB/sec | 531 MB/sec |
Windows 2003 Server Results
| Test | SSD Result | |
|---|---|---|
| X4140 | X4440 | |
| Random Read (4K) | 66.9K IOPS | 63.1K IOPS |
| Random Write (4K) | 17.6K IOPS | 29.7K IOPS |
| 50-50 Read/Write (4K) | 14.9K IOPS | 16.9K IOPS |
| Sequential Read (MB/sec) | 832 MB/sec | 844 MB/sec |
| Sequential Write (MB/sec) | 121 MB/sec | 408 MB/sec |
Results and Configuration Summary
Storage:
- 4 x Sun SSD
32 GB SATA 2.5" SSD (24 GB usable)
2.5in drive form factor
Servers:
- Sun SPARC Enterprise T5240 - 4 internal drive slots used (LSI driver)
Sun Fire X4450 - 4 internal drive slots used (LSI driver)
Sun Fire X4440 - 4 internal drive slots used (LSI driver)
Sun Fire X4140 - 4 internal drive slots used (LSI driver)
Software:
- Window 2003 Server
OpenSolaris 2009.06 or Solaris 10 10/09 (MPT driver enhancements)
Vdbench 5.0
Benchmark Description
Sun measured a wide variety of IO performance metrics on the Sun SSD using Vdbench 5.0 measuring 100% Random Read, 100% Random Write, 100% Sequential Read, 100% Sequential Write, and 50-50 read/write. This demonstrates the maximum performance and throughput of the storage system.
Vdbench profile:
- wd=wm_80dr,sd=sd*,readpct=0,rhpct=0,seekpct=100
wd=ws_80dr,sd=sd*,readpct=0,rhpct=0,seekpct=0
wd=rm_80dr,sd=(sd1-sd80),readpct=100,rhpct=0,seekpct=100
wd=rs_80dr,sd=(sd1-sd80),readpct=100,rhpct=0,seekpct=0
wd=rwm_80dr,sd=sd*,readpct=50,rhpct=0,seekpct=100
rd=default
###Random Read and writes tests varying transfer size
rd=default,el=30m,in=6,forx=(4K),forth=(32),io=max,pause=20
rd=run1_rm_80dr,wd=rm_80dr
rd=run2_wm_80dr,wd=wm_80dr
rd=run3_rwm_80dr,wd=rwm_80dr
###Sequential read and Write tests varying transfer size
rd=default,el=30m,in=6,forx=(512k),forth=(32),io=max,pause=20
rd=run4_rs_80dr,wd=rs_80dr
rd=run5_ws_80dr,wd=ws_80dr
Key Points and Best Practices
- All measurements were done with the internal HBA and not the internal RAID.
See Also
- Vdbench is publicly available for download at: http://vdbench.org
Disclosure Statement
Sun SSD delivered 71.5K 4K read IOPS and 1012 MB/sec sequential read. Vdbench 5.0 (http://vdbench.org) was used for the test. Results as of June 17, 2009.

The x4140/x4440 results seem surprisingly low compared to the x4150 (1/3 the random reads and 1/2 the random writes). Are you sure there wasn't an issue with the tests? If not, any idea why there was such a performance delta?
Posted by Bill Hathaway on June 25, 2009 at 10:16 AM PDT #
We did some digging with this with the platform engineers and this is not an issue with the tests. This is the real performance. This is based on the LSI drivers used on each platform has caused this variability they all don't perform equally. The Cross platform engineering team is taking a closer look at this to level the playing field.
Posted by Leah Schoeb on June 25, 2009 at 11:45 AM PDT #
That was straight-up Windows 2003 Server, no service packs? How about Windows 2008 Server? That there are no results for Windows on the T5240 is understandable :) But why no results for Windows on the X4150?
A few things that might have made the post stronger:
Given that each of those system models have a choice of processor/frequencies, it would be good to include what those were.
Reporting the CPU utilization at those IOPS and MB/s rates would be goodness so people know how much CPU they have left-over for the "real work" :)
[Perhaps the frequencies and CPU utilization might help answer Bill's question about the difference between the x4140/x4440 vs the x4150]
Also, since the post asserted power savings it would be further goodness to show the power delta between using the SSDs and using platters.
It would also help to expand on how using SSDs results in the asserted space savings particularly since the drives were all internal to the system chassis.
Posted by rick jones on June 26, 2009 at 08:30 AM PDT #