SSD power consumption (vs spinning HDD)

I've caught SSD fever, and I've already decided to upgrade to one in my work notebook. While I'm waiting on Intel G2 availability, I was wondering about the effect that this will have on power consumption (and battery life). It looks like it'll have a neutral or possibly negative effect.

It seems like the main benefits are speed and security (stored data more robust to shocks). Power consumption is actually higher than with my current drive. I know the SSD will do its work faster, and will be at idle a higher % of the time. But by enough to compensate for the higher power usage?

Is my math off? Has anyone lost battery life by upgrading to an SSD?

Intel G2
1.5W read/write
0.75 idle

OEM Fujitsu drive
1.3 read/write
0.6 idle
0.13 standby

 

it won't make much of a difference on battery life because HD's or SSD's don't consume much power with respect to the overall power consumption of the notebook. the motherboard, CPU, fan, screen, and memory take up much more power so i'd say at max you'll only gain or loser a few minutes of the battery.

 

I guess that's a good way to look at it. The worst-case scenario for the comparison is when the spinning drive is in standby. The difference of .62W translates to a difference of 16 minutes of battery life with 90Wh battery.

Since I do actually use the HDD, the possible cost to battery life will necessarily be < 16 minutes. Maybe even no difference if I use the drive enough.

Guess I'll find out for myself in a couple weeks.

 

You seem very keen on getting the Intel drive and it is a very good performer. Drives with a Samsung controller at least were the ones to beat up until a few weeks ago when it came to power consumption.
http://techreport.com/articles.x/16848/13
I gained battery life by upgrading, how much exactly is a bit difficult for me to say. I have Corsair P128 drive installed in my laptop. I also have a Samsung SSD installed in my travel laptop.

More links
http://www.tomshardware.com/charts/2009-flash-ssd-charts/Power-Requirement-at-Idle,921.html
http://www.tomshardware.com/charts/...Power-Requirement-at-Max.-Throughput,922.html

 

isn't the drive idle most of the time instead of standby? this translate to a difference of only 0.15w, which is about 4 minutes of battery. but since you know you'll be reading/writing stuff, it'll probably be closer to 5-8min of battery. going by this logic, you'll lose these minutes based on the power consumption number that you gave.

what i don't understand is how a SSD (no moving parts) use more battery than a HD that's spinning. and to add on to this, SSD produce almost no heat compared to the HD, so i really think SSD consume less power no matter what. going by this logic, you should be gaining battery using an SSD.

i think these two points counter-balance any gains or losses you'll be seeing with the battery, especially when it's already only a few minutes.

 

@WFD & Rachel: I agree with you both.

I calculated the "< 16 minutes" figure just to try to establish the upper limit on cost to battery life. I'd expect the observed cost in any realistic setting to be much lower.

@Rachel
When I got the Fujitsu, I didn't really know enough to appreciate how hard drives could bottleneck system performance. I also didn't know that my machine is limited to SATA 150 when shipped with a spinning hard drive. On top of that, my hard drive is definitely geared more toward power savings than performance. New 5400 drives by WD are noticeably faster.

Long story short, the solution is either to ditch the notebook or find a way to maximize performance w/in limitations of SATA 150. I don't regret buying the Lifebook, and don't want to sell it. So that means finding the SSD that's best for SATA I. The Intel seems to be that drive because its strength is random read/write performance. The fact that its sequential write speed is < 100Mb/s is no concern to me. The only thing I miss out on is sequential read.

I took a good look at the Corsair 128 that Hexus reviewed, and I liked what I saw. But its main advantage (along with the Indilinx Barefoot SSDs) is write performance. And I have no way to take advantage of those speeds in this notebook. So it doesn't have as much value for me as it might if I had a different notebook. The Intel just seems right for me.

 

I think the power consumption is measured differently. The Intel will still give you more battery life than any HDD - this can be confirmed in the SSD thread from users w/ the Intel G1. I doubt the G2 will be much different in terms of power consumption.

 

Those are manufacturers specifications and they have little to do with real life power consumption.

Real life power measurements:
http://techreport.com/articles.x/17269/10

You may be able to find real life power measurements of the Fujitsu sotckdrive on Tom's Hardware.

I agree.

 

There's one thing you forget to take into account. Because the drive is faster, it will finish its job a lot quicker than a spinning HDD. As the result, most of the time the power consumption of an SSD will be its idle power consumption, whereas for a spinning HDD it will mostly be active read/write power consumption.
To prove my point, when using Vista with a HDD, the harddrive activity light is always blinking, whereas with an Intel SSD, the light almost never goes off. I also make sure that it is that way, by telling the SSD to turn off after 1 minute of idling. I can't do that with HDD, because performance will be degraded during the spinning up and spinning down of the spinning platters.

 

This one of the better real life tests, imo.

Source: http://www.laptopmag.com/review/storage/intel-x25-m.aspx?page=7

It's Intel gen 1 but gen 2 is nearly identical in power consumption.

Scorpio Blue is comparable to the Fujitsu stockdrive.

 

Ah, that's definitely a nice set of results Phil. It shows the advantage you get from the Intel's faster random write speed, since web surfing involves downloading a lot of small files into the browser cache. Meanwhile the Samsung still wins on the video test, because it's still more efficient in sequential reads.

Now what would be interesting to see is a rerun of the websurfing scenario, where the browser is configured to only use RAM for its cache (or putting the cache on a RAMdisk). In that case, none of the drives should actually be active for any write operations. (I have my browser configured to use /tmp for its cache, and have a tmpfs mounted on /tmp...) Then the I/O activity will be limited to paging in executable code from the disk, and the drive should be mostly idle.

Likewise for the Video rundown test it would be interesting to see what readahead settings were used (if any). When you've got a lot of free RAM, you might as well set the readahead as large a possible. That again capitalizes on the sequential read efficiency
and allows the drive to go idle for longer periods of time.

Anyway, based on the TechReport results I concluded the Samsung is still the most power efficient. It has extended my runtime by 30 minutes at least, but the drive it replaced was also an obvious power hog. And, I've tweaked my system cache settings so that the drive can stay idle for very long periods of time...

(Btw, "NBR Impulse Buyer" ? I love it...)

 

Toms Hardware also give measure power consumption data. SSDs and HDDs. Samsung and Intel seem to be the best for low power consumption. Some other SSDs are less good.

John

 

The last test that Phil linked to did not use the more power conservative PBJJ-2 drives in their tests. Obviously those drives were not released yet when they did these tests.

When i have the time i will try to conduct my own tests.