Bapco MobileMark 2012 1.5
Version and / or Patch Used: 2012 1.5
Developer Homepage: http://www.bapco.com
Test Homepage: http://www.bapco.com
MobileMark 2012 1.5 is an application-based benchmark that reflects usage patterns of business users in the areas of office productivity, media creation and media consumption. Unlike benchmarks that only measure battery life, MobileMark 2012 measures battery life and performance simultaneously, showing how well a system design addresses the inherent tradeoffs between performance and power management.
The MyDigitalSSD BP4 2.5" gives the longest battery life we've tested to date with a 2.5" SSD, but 2.5" battery life doesn't always relate to mSATA battery life. Strap in because you're about to get six months of lab testing in a few educational paragraphs.
At this time there are very few notebooks with mSATA slots running the SATA III spec. I found a grand total of two, both using AMD processors and chipsets. This will change with the introduction of Haswell, Intel's 4th generation Core processor. Haswell brings new chipsets with it and the SATA III port count increases from two to (up to) six, depending on the chipset used. We don't have any firm Intel Haswell data, nothing confirmed, but the popular rumor says up to six SATA III ports. That number may be just flagship desktop boards but, well you get the idea. It's not released yet, but it's pretty safe to say you won't have to worry about your mSATA port running at SATA II speeds, like they do now.
There are two methods of testing mSATA products. The first is with a native mSATA slot. On current notebooks and desktops that means SATA II limitations. The second method is with an mSATA to 2.5" adapter bracket to get SATA III speeds. The adapter kind of destroys the purpose of the mSATA form factor.
The mSATA to 2.5" adapter brackets available in the wild, i.e. retail, all use the 5-volt power line and converts it to 3.3 volts, the standard voltage for mSATA. Intel made a very limited number of special adapters that accept 3.3 volts and forward it directly to the mSATA drive. The chart below shows accurate 3.3 volts power consumption data with SATA III speeds, tested on a desktop motherboard.
The chart above, the Notebook Battery Life chart, also shows native 3.3 volts, but at SATA II speeds. This is a limit dictated by our Lenovo W530 and every other existing Intel based notebook on the market today that we know of. There are ultrabook products with mSATA SATA III ports, but we've yet to find one that also accepts 2.5" form factor HDD/SSDs. We wanted to use the same notebook for testing both mSATA and 2.5" products, and we got it wrong. We should have went with two separate test beds, a notebook for 2.5" SATA III and an ultrabook for mSATA. With Haswell just around the corner, we're not going to change the test system until after May when Haswell arrives.
PCMark Vantage HDD Tests - Power Draw
Running an mSATA drive at SATA II speeds increases what's called disk busy time. A drive is idle, gets sent a command, completes the task and then returns to idle. The time it takes to complete the task is the busy time. In the Trace Based Power Over time chart we clearly see the disk busy time and the return to idle by the power usage. Over SATA II, the busy time increases because the bus is slower.
The more time a SSD can spend at idle power, the longer the battery life. Several events can take a SSD out of an idle power state, the user requesting an action or a SSD performing background tasks.
Let's dive into these two charts starting with the second chart first. We've automated the test from the very beginning. Both software tools are open and a script starts upon requesting an action. The data logger starts for 10 samples and then the Vantage HDD test starts. All of the latency comes from the SSD and not in the actions of starting the tests.
In the chart, you can see the Intel 525 starts before the other drives. The MDSSD BP4 and Kingston prototype with a LSI SandForce B02 controller start at the exact same time and the Mushkin Atlas with asynchronous flash starts much later.
We can use the distance between the spikes and idle state drops to measure performance and disk busy time. There is little backend activity happening in the trace power chart, but in the Notebook Power chart, which runs for up to five hours, background activity becomes a factor.
In our notebook test, the MyDigitalSSD BP4 240GB delivers less battery life than the Intel 525 240GB and Kingston Prototype with a LSI SandForce B02 stepping controller. This is while limited to SATA II speeds, though. I suspect there will be significant changes when the mSATA drives are connected to SATA III. Also, Haswell based systems bring a new SATA spec feature called DEVSLP or Dev Sleep. We know the B02 controller takes advantage of DEVSPL.
By the time you read this, we'll have a Haswell based notebook in the US lab, but we'll be bound by NDA's from disclosing our findings until June 4th.