We all know the high performance characteristics of solid state drives (SSDs), but what about their energy usage when compared to traditional spinning disk technologies? With the recent release of a new storage benchmark from the Storage Performance Council (SPC) that looks at not only performance, but real world energy use, and the release of the first two tests, we have a chance to find out.
The SPC Benchmark 1C/Energy (SPC-1C/E) benchmark extension for storage components was released recently. It includes the existing SPC-1C workload, which is designed to measure performance of small storage subsystem configurations, but now includes ‘energy’ extensions to provide insights into energy use in addition to performance.
In what is a very nice touch, not only does the benchmark measure power consumption of the system under test during the peak performance period, it also requires the storage to idle for a minimum of half an hour, which gives insights into the power consumption across the subsystem’s performance spectrum from idle to peak usage and what happens as the subsystems migrate from one state to another.
This contrasts to the Environmental Protection Agency’s (EPA) recent release of specification 1 of the ENERGY STAR rating for servers, which measures power consumption of the server at idle and doesn’t yet provide insights into power consumption under load. This will be handled by a future V2 specification expected late 2010.
The first two results under the SPC’s SPC1-C/E benchmark were released along with the announcement of the new benchmark. One was Seagate 300GB 10k rpm SAS drive subsystem and the other was an IBM 69GB Solid State Disk SAS configuration. As luck would have it, one result uses traditional rotating disk drives and the other uses solid state drive technology. This gives us an initial insight into the power consumption characteristics of these very different technology types, along with the more traditional price and performance aspects.
It is worth noting that the two solutions are not directly comparable. For example, the IBM eight (8) drive SSD offering represents a 548GB subsystem, compared to the 24 x 300GB solution representing 3,600GB of addressable capacity. However, as these are the first two test results released it is interesting to compare and contrast the outcomes all the same.
The table below outlines the key metrics and outcomes.
|
|
IBM |
Seagate |
|
Power (W) peak |
165.4 |
220.9 |
|
Power (W) idle |
160.8 |
171.9 |
|
Drive type |
SSD |
Traditional spinning magnetic disk |
|
IOPS peak |
45,004 |
8,004 |
|
Cost in $USD |
$87,468 |
$15,209 |
|
Capacity in Gigabytes |
548 |
3,600 |
|
Cost/IOPS in $USD |
$1.94 |
$1.90 |
|
Cost/GB in $USD |
$159.61 |
$4.22 |
|
GB/Watt |
3.31 |
16.30 |
|
IOPS/Watt |
272.09 |
36.23 |
From an energy point of view it is interesting to see the power consumption at idle compared to peak performance. On servers, idle power consumption can be anywhere from 60% to 90% or more of the peak performance energy consumption. The IBM Solid State Drive solution shows its idle power consumption is 97% of peak usage power consumption. This contrasts with the traditional disk solution from Seagate, which has idle power consumption at 78% of peak. This makes sense when you think about it. Solid State drives, as the name suggest, don’t have moving parts and so don’t have rotating platters, motors and breaks to ramp up and ramp down the disks. The mechanical parts can draw a lot of current. Without them it therefore makes sense that the power usage differences between idle and peak performance of the subsystem is not that great either.
But, what about other perspectives? The chart below plots a head to head comparison of four metrics.
The above chart is analysis by IDEAS based on data supplied by the Storage Performance Council (SPC) in the two Executive Summaries of the two test results so far. Note that the segment for each bar with a green label is better than the red label – e.g. the lower cost per GB is better, even though it is the smaller segment of that bar on the chart.
- The first is the price per unit of performance (Cost / IOPS) where the two solutions are neck and neck. Both solutions have a cost per transaction of around $1.90. 2.
- The next is the cost per Capacity (Cost / GB) where the spinning drive technology has a big advantage over the solid state disks. The Seagate solution costs just $4.22 per GB, compared to the IBM SSD $152 per GB.
- Next is one measure of ‘greenness’, (Capacity / Watt). That is, how much capacity is available per unit of energy? The spinning disk wins here. This is primarily because the traditional disk solution offers relatively high-capacity (3,600GB) compared to the IBM SSD’s 548GB, whilst both are in the same ballpark in terms of total power consumption. The result is the spinning disk solution can support around 16GB per Watt, compared to the SSD 3.30GB per Watt.
- The other ‘greenness’ measure is of course performance per Watt (IOPS / Watt) and here the tables are reversed and it is the SSD solution which comes out on top. With peak performance of 45,000 IOPS compared to 8,000 IOPS for the traditional disk, the SSD technology can process more transactions per unit of energy, given that both subsystems have similar power draws.
Of course, these are just the first two results for this new benchmark and we expect more will soon follow and it will be interesting what new insights these will bring.
But in the meantime:
- The outcomes certainly back up the perception of the speed of solid state technology over traditional spinning disk.
-
But as the Cost per IOPS shows, the relatively higher cost of SSD technology means that the price performance of the two is on par.
-
And if capacity rather than performance is your key requirement, then from an energy use perspective at least, spinning disk technologies would appear to still be in play.
The SPC is a non-profit corporation founded to define, standardize and promote storage benchmarks and to disseminate objective, verifiable storage performance data to the computer industry and its customers. The organization's strategic objectives are to empower storage vendors to build better products as well as to stimulate the IT community to more rapidly trust and deploy multi-vendor storage technology.
Gary Burgess, Senior Vice President, Research & Operations, Ideas International Limited
With 20 years of experience at IDEAS and 28 years of experience in the IT industry overall, Gary specializes in IT hardware research, including features, performance, Green IT, and pricing. He also leads the IT Infrastructure Support Services research for IDEAS International.