Impact on Battery life with TWO Drives: SSD + 7200rpm HD

I cannot seem to find any information on the impact of having (powering) two drives compared to one.

From my understanding, getting a SSD + a 7200 rpm HD would greatly decrease the battery life because the harddrive is the third most power consuming component. However, I am looking for some benchmarks, research etc that would provide proof that it is one way or another and by how much it affects it.

Thank you!

 

The impact of replacing an HDD with an SSD is minimal (10 to 20 minutes in my experience over a 6 hour battery life - and it didn't matter if it was a 5400RPM or a 7200RPM HDD... as long as we were talking WD Blue and Hitachi TravelStar models).

By having two drives installed, you have effectively doubled the impact to your battery run times your storage subsystem will have.

Btw, HDD's are not the (third) most power consuming components of a mobile system.

The GPU, CPU, iGPU, Screen, M/B, and other miscellaneous circuitry (like USB 3.0, for example, and even RAM - especially 8GB+ SoDimms x 4 configurations) are all far greater power users and for most consumers, the HDD/SSD is mostly in idle mode too further reducing it's power consumption vs. other components that need constant (higher) power draws.

I would not worry about this issue one bit. But I would consider getting two SSD's instead of an SSD + HDD though.

Not so much for the power efficiencies possible (but not guaranteed... ever), but for the robustness of solid state vs. a spinning platter. And for the increased snap the system will have overall (especially if you move system folders to the HDD... as you should, for a seamless O/S navigating experience).

Hope this helps.

 

Not only do HDDs dissipate an extra 2-3W over and above that of a SSD (which can be pretty significant for run-times), but the extra heat they dissipate into the laptop's thermal envelope can cause the laptop's fan to run more. And extra drain on the battery, in and of itself, causes the power converter circuitry to run hotter, and for the battery to discharge at a faster rate.

As an interesting point of fact, the laptop quoted below started its life with a 3.25 hour battery life with its 9-cell factory battery. Between replacing the HDD with a SSD, swapping from a CCFL to a LED backlight, and upgrading the CPU -- battery life on a fresh battery is now over 5 hours. Additionally, while the factory battery only lasted me roughly a year before its capacity was depleted, the replacements are lasting more than 2 on account of the lower discharge rates. My upgrades have literally been paying for themselves simply on account of preserving my batteries!

Bottom line is that even small improvements to power efficiency can make a huge difference to run-times.

IMHO, hard drives don't belong in laptops. The advantages are just so compelling. Give me an ancient laptop (ie: see my signatures) with SSDs, any day of the week over the latest/greatest processor-containing laptops hamstrung by rotating magnetic HDDs.

 

Surely the power savings are primarily due to the back light swap. Did you take any measurements between upgrades, or just before and after all three were done?

 

That may be true of the lower powered SSD's four or five years back. Today's SSD's vs. today's HDD's are very comparable in actual battery usage (i.e. not worth figuring out) when the system is actively using the storage subsystem in any way.

The same thing for the heat. Many of today's SSD's run much hotter than I would allow an HDD to run back in 2010.

As for your batteries lasting you longer? Battery discharge rate differences are minimal between storage subsystems and do not affect longevity of the batteries in question. Battery charging rates and number of charges is what affects longevity. You've just been using the system differently.

If batteries cost ~$100 - then unless you have a tiny SSD, you haven't saved anything either.

I'll take any current platform (Broadwell i7 5700 or higher...) with a HDD any day over an ancient platform with the best SSD installed. The storage subsystem doesn't impact productivity on a computer system as much as the CPU+RAM combo does. Any day of the week.

 

Measurements are hard, because I've done the upgrades progressively over the past 4 years or so, obviously without consistent batteries. But it seemed that I gained around half an hour battery life for each of the three upgrades, give or take. With an 85Whour battery life, 3 hours of run time = 28W average, 5 hours of run time = 17W. So 11W difference. The SSD likely saved 2-3W, the CPU 4W, and the screen another 3W with the balance of savings made up in decreased running of the cooling fan and reduced power conversion losses.

Not sure what SSDs you're buying, but the only reason this might be the case is because, to reduce costs, manufacturers have pulled back on the materials they've used for enclosures. For instance, the Intel 330 uses a full aluminium case. The 335, with its cooler running components, uses a plastic case which might result in higher board temperatures due to less conductive heat transfer, but overall less power dissipation.

Have to disagree here. The discharge rate of a Li-Ion battery (whether its in a laptop or Tesla car) affects its internal temperature, and internal temperature creates additional resistance. Heat is the enemy for Li-Ion cells during discharge. Rapidly discharged cells wear out in a shorter number of overall discharge cycles than lightly discharged cells.

They cost about $30 nowadays, but by requiring one less replacement every 2 years, that's already paid for the backlight replacement and the T9300 CPU. The SSD has more than paid for itself simply on account of its vastly superior reliability.

 

Can't really follow your logic here?

Your stats are over 4 years? There is no way you are using the system the same way over that time period... even Windows updates would guarantee you don't...

I am not questioning that your power savings are close to what you indicate - but again; I stated that the power and heat issues of todays SSD's with today's HDD's are mere minutes apart for normal workloads - regardless of the performance differences. Comparing a modern SSD to a 4 year old HDD is not exactly apples to apples.

SSD's produce more heat as their capacity and performance goes up. Materials do not come into the equation as much as that fact. Even if a thermal pad properly placed can aid greatly in reducing the interference of heat in other components. Older SSD's simply did not get good performance vs. today's examples.

The Li-Ion battery is not affected by the differences between a HDD vs. an SSD. Again, read what I originally wrote. And even taking all the upgrades you did together, the battery system is still operating within spec's, so discharge rate is not what has affected your batteries lifecycle.

Hehe, $30 for a battery? And you've saved one/two battery changes so far? And have bought a backlight replacement and upgraded the cpu too for that? Kudo's.

But the SSD was in no way involved in any battery life/lifecycle savings. And your assertion that it has paid for itself on it's vastly superior reliability is also humorous. I don't know of any HDD (except for almost any Seagate model, lol...) that wouldn't have survived just as easily. Now, if you had said the SSD is more robust than an HDD, I agree. But you didn't.

 

Not a lot has changed. Windows hasn't become any more bloated today than it was 4 years ago. I'm still running the same version of office. The same other apps.

Hard drives have not become more energy efficient over the past 4 years, and SSDs have not become less energy efficient. Not sure why you keep claiming this because its not true.

That doesn't appear to be the case. Process node shrinks have brought SSD power consumption down. HDDs have seen little improvement in comparison.

SSD performance hasn't done much in the past few years for laptop users either. Okay, the latest/greatest drives are pumping out 80k I/O's compared to the 20-40k back then, but that's mostly relevant only to servers anyways. All I was doing was offering up an explanation as to why board temperatures might be slightly hotter in some of the newer SSDs -- and its not because of overall dissipation!

An incremental 2W draw on a Li-Ion pack can make a difference over its lifetime longevity-wise, all other things being equal.

Again, not true. The SSD gave me an extra half hour of battery life the day it was installed, and reduced the rate of draw on the laptop's battery. It absolutely has contributed to battery life improvements. 2W is a big deal when you're only using 20-25W in a system on average! Every little bit counts!

Humourous? Hardly. Its a fact that SSDs are preferred for reliability. Eliminating the moving parts is a really big deal.

 

Lol... are you serious? Are you running the same version as 4 years ago? With no updates on Windows or Office either? New software=different workflow, period.

HDD's become more energy efficient with each gen introduced, this is a fact. Better manufacturing processes, better firmware, denser platters - these all combine to give us much better power efficiencies vs. older generations.

Meanwhile, SSD's have become comparatively less efficient because the Controller needs to be more powerful to fully saturate the SATA bus (as most can easily do now). And vs. older SSD's that couldn't do that even with 2 or 3 of them in RAID0, yes, they are less efficient on an absolute scale.

Idle power consumption may be lower (depending on the chipset of the platform and the configuration of the O/S), but in use, the power consumption has gone up over time. Not going to do the research for you; look at reviews yourself for proof.

HDD's power consumption has been progressively gotten better with each gen. The performance increases are mostly mechanical, not electrical in nature. For example, better bearings, higher platter density and lighter motors and platters all contribute to a more efficient component - and the small controller/cache needed for the relatively low performance HDD's offer doesn't need to be upgraded along with everything else (and of course is itself further optimized for power efficiency too - firmware and hardware).

No, increases in SSD performance are huge (until a few years ago; VRaptors were my choice of weapon for storage subsystem components), especially when capacity and the related OP (over provisioning) capability it offers.

Your 20-40K 'back then' are generous. But 4x the performance does not come with a reduction in power usage either...

A battery pack's lifecycle is stated in 1000's of charge/discharge cycles - even if I accept your 2WHr number - it is irrelevant over the a nominal 3 year period - especially as the discharge rate is still within spec's of the battery pack in question.

What SSD? From what HDD? On what O/S? In what workload? On what platform?

Reliability is not the same thing as durability/robustness.

"Humourous' is humorous though.

 

I've heard plenty of claims of SSDs significantly reducing power draw and extending battery life...but, never seen any hard data to prove it, or evidence in my own experiences.

SSDs are great for any number of reasons, but power savings relative to hard drives is not one of them as far as I can tell. Not to say they use more or less, just not a significant enough amount more or less that it's even noticeable.


Sent from my iPhone using Tapatalk

 

What are you laughing about? Win7 hasn't changed much in the past 4 years. Sure, they've added a few patches and tweaked a few frameworks, but the code that has been introduced, if anything, is more optimized today than it was in the past on account of the industry's emphasis on mobile platforms where power and memory are legitimate considerations.

You're dramatically over-stating the efficiency gains. Yes, benchmarks might show relative equivalency when driven into saturation for significant periods, but a SSD can perform the same workload at a dramatically lower duty cycle than what you see for the HDDs.

"Better", but not "much better". I just compared a 8-year-old Seagate 2.5" 5400rpm design to the current shipping Seagate 2.5" 5400rpm hardware, the difference was literally 0.2W, or around 10%. Most of which came from density -- they only were spinning one platter around instead of 2!

Modern SSDs are more energy efficient on the same workloads than the previous generation. Not sure where you're coming up with this 'comparatively less efficient' claim. RAID-0, who cares, nobody uses
that in a laptop configuration.

No, idle power consumptions have been going down on account of better coding, and the push towards running on light mobile platforms. As hardware matures, the kernel developers (Windows/Linux) have been utilizing power states more effectively. For the practical user, a lot of supplied systems software is far more aware of the power state of the machine and will take measures accordingly.

Less than 10%. If even. Hardly worth writing home about. The big gains in power efficiency for storage has been the move to flash memory in the form of SSDs.

Not really. The big jump was going from a hard drive with its 150-250 (maybe 350 if we're being generous and have a decent 15krpm drive) IOPS to the SSDs which can do thousands. Once you're into SSD storage, upgrading doesn't really buy you a lot of improved performance. Sure, it might show up on a benchmark that a drive is "double" the speed, but to the end-user, that's pretty meaningless. Not like the Earth-shattering improvements that nearly all SSDs (except the most poorly implemented/buggiest models) provided over HDDs.

It absolutely does. If you keep the workload constant, the SSDs use less energy than their predecessors. But in some cases, they are capable of doing much more, albeit using slightly more energy in the process.

Rate of discharge determines the number of cycles a battery pack will be able to withstand. Anything that can be done to reduce the rate of discharge will give you more cycles, and more run-time within each cycle. Basically any sort of energy efficiency improvement that can be achieved in a Li-Ion powered system will have a significantly positive impact on overall battery serviceability and the customer experience.

Its actually significant, and not irrelevant at all. Li-Ion cell life decreases with increasing discharge rates and discharge severity. Every EE who has ever designed Li-Ion systems and observed them over their lifecycle will verify this. I'm not aware of a rule of thumb that's been studied with scientific scrutiny, but its fairly safe to say that its even small reductions in drain rate can have very positive impacts on cell longevity and number of cycles achievable.

There are few 'specs' given by the manufacturers for cell cycle longevity versus drain rate. But its just common sense that minimizing drain rate either by reducing load, or by providing more cells will allow a battery pack to run cooler, and withstand more cycles.

 

Thank you all for your inputs and discussions. However, I think only tilleroftheearth and pitz actually answered the question. I am not really interested in the power usage or savings between SSD and HDD. I am more interested in the power usage between 1 Drive versus 2 Drives in general. That could be SSD + HDD, or it could be SSD + SSD, it doesn't matter, what matters is the difference between 1 vs 2 drives on power / temp / battery life comparison.

From what I gathered so far, having two hard drives is not really an issue because the power usage is quite low anyway, especially if it is not used constantly (i.e. storage only drive vs system files drive). Also, there is a significant difference between SSD vs HDD that it would be better to have TWO SSD's, although it has been argued that the power usuage between SSD and HDD is very similar today than it was five years ago.

TLDR: What justification is there for not getting a second hard drive due to the reduction in battery life, or an increase in temperature with 2 drives, that is significant enough to warrant the 1 drive only theory.

 

pitz, there is no point in continuing this conversation with you. You simply cannot or will not read for comprehension.

When you can argue that Win7 hasn't changed in 4 years... what's the point?

See:
https://en.wikipedia.org/wiki/Windows_7


10% improvement over an already efficient product is significant (even if you're using Seagate as your example).

If you can't understand why SSD's are comparatively less efficient today because of their more powerful controllers.... zzzzz....

VRaptors in RAID0 was my storage benchmark back then - nothing to do with RAID0 on a noteobook. Again; reading for comprehension is a pre-requisite to having an intelligent conversation.

Batteries are most stressed when charged - as usually they take (much) longer to discharge than they do to charge. But, keep insisting the discharge cycle is where most of the wear happens (we know better).

4x the performance is not huge to you? The original SSD's (circa 2009) that I tested that were already 'hundreds' of times faster than a hard drive in synthetic 'scores' were not even able to keep up with a HDD as far as I was concerned in my workflows - and certainly not worth it for the price. Today's SSD's are; because they are finally better overall than a HDD. Those early examples? At $600 a pop? Yawn. And to get there the SSD's power sipping capabilities was silently shot, buried and forgotten long ago on a cold winters night.

Even you yourself contradict your own conclusions about the effect the O/S and Program updates have had on power consumption (see the bolded part of your quoted text below) - and, more importantly, you totally disregard to give the specifics as I asked of you.

The Intel X-25M had great performance (for 2008) and power efficiency back in 2008 when it was introduced.

See:
http://www.anandtech.com/show/2614/16


And yeah, it offered 6% improvement of battery life (over a 7 hour base from a 7200RPM HDD). But those days of SSD's offering increased battery life in anything other than idle conditions is long gone.

Long gone as in over half a decade ago.

I'm sure my posts can still be found on this forum where I've stated a few times that Intel just needs to increase the power requirements to introduce the fastest performing SSD (at that time I wrote the posts).

And in 2015 they did: with the 22W+ Intel 750 PCIe SSD.

 

The one drive/two drive argument could be:

One Drive:
1) A larger drive is usually faster (currently limited to ~1TB currently - the SanDisk Extreme Pro 960GB model).

See:
http://forum.notebookreview.com/threads/samsung-adds-2tb-capacity-to-its-ssd-850-family.778276/

(Please see my posts in the thread above).

2) A single drive will use less power than any two drives will.

3) A single drive may be less expensive if your workflow can be limited to the capacity selected.

4) A single drive will create less heat than two will.

Two Drives:
1) Two drives can give you effectively twice as much capacity.

2) Two drives can be setup more optimally for specific workflows.

3) Two drives can be setup to backup important WIP files (work in progress).

4) Two drives can relieve a potential bottleneck if copying files back and forth many times a day in a normal workflow (for example; I copy RAW images to two locations on the field and edit one version while leaving the other untouched, for backup purposes...). When I'm finished with a folder, I copy the edited tiff or jpeg files along with the original RAW files from the WIP folder to the WC (work completed) folder - and this is from one drive to another for maximum performance.


The better question would be to state your normal or expected workflows with your system - then the answer will be tailored to your wanting maximum performance or maximum efficiency.

The obvious facts that more heat will be generated and also less battery life with a 2 drive system is not an issue to me if the performance delta over a 1 drive setup was significant enough for the workflow expected.

 

Sorry for being way too theoretical with that other user. Anyways, an extra HDD will add roughly 2.5-3W of power consumption to a laptop. So if you have a laptop that consumes 20W average that lasts 4 hours (on a 80W-hour battery, ie: 9 x 18650 cells), then adding the extra hard drive will reduce you to 80/23 = 3.5 hours.

You'd have to use numbers as applicable to your specific system in the calculation. Higher drain rates themselves increase I^2R losses inside the battery itself (cells have internal resistance!). So the actual outcome will probably be slightly worse than your calculations. If a battery system is particularly under-sized or overloaded for a given laptop design, it might be significantly worse.

In the example I gave which is a fairly realistic scenario, it is a pretty significant hit -- half an hour of battery life, or a 1/8th reduction of run-time just to run that 2nd hard drive.

His "argument" is nonsense and isn't supported by the available data. Power consumption for SSDs to perform a fixed workload is falling all the time and likely will continue to fall.

 

The nonsense argument is that SSD's perform a fixed workload. But that is not how they are used by us or the O/S's and Programs they run. Yeah, they do each step faster and with usually more power required until they hit idle, but they are still active and drawing power unless the system is turned off.

Then the power savings are equivalent for both HDD's and SSD's.

The Intel X25-M gave 27 minutes of extra time (6%) on a 7 hour nominal (w/7200RPM HDD) battery life. Today's SSD's won't give that kind of difference on a nominal 4 hour battery life.

pitz acknowledged that the O/S he uses has become more efficient - yet still attributes this to the solid state hardware.

We're not there yet (power-wise) and most likely won't ever be. Why? Because SSD's are deemed good enough for power usage and performance isn't. So while they keep improving for a theoretical fixed workload at the same or lower power levels required - their absolute power requirements in a computer that is not just sitting there idle is effectively the same as a modern HDD. And it has been that way for the last few years.

 

Hard drives, like the WD Blue idle at about 0.5-0.7W, and only about 1.5W at load. Modern SSD's vary wildly, and some consume more than 0.5W at idle, as much as 1W, but those with power saving features idle at well under 0.5W. The advantage of an SSD is that they're faster so even though they may consume more power at load, they also get the work done ten times as quickly. But I wouldn't necessarily say as a blanket statement that SSD's consume less power than an HDD.

 

Exactly. It all depends on what SSD you're looking at. Samsung drives tend to be very power efficient, especially at idle, but drives that contain enterprise-class features, such as the Intel 730, will consumer more juice than a platter drive in pretty much all circumstances.

 

The last time I tested (a very long time ago) the idle power improvements of an SSD over a HDD, the results seemed spectacular (almost 2 hours longer, if I'm remembering correctly; WD Blue 1TB vs. Crucial M4 512GB SSD with a base time with the HDD of about 6 hours).

However, when I actually took the system with me out and about in the real world (not to do my 'heavy' workloads, but just to have a system with me for my notes, web research and accounting program), that two hour magical savings dissipated to less than 10 minutes. Not impressive w/regards to battery life. But still impressed because the system stayed responsive in all storage subsystem aspects, without taking a step backwards at anytime (as many previous SSD's I tested had). The extra 10 minutes battery life was simply a bonus, and I did notice a cooler running system too.

I repeated that test a couple of times more with different client's systems - and found the same thing each time; increased idle time in no way increased the true run time of those systems significantly. But, as long as the battery life didn't take too much of a hit (and it didn't), it didn't matter if the increase was minimal either.

In my own use, a mobile system is not on and idling for longer than a few seconds anyways. I turn it on, do what I need and then it is off again.

If others use their systems in a more lackadaisical fashion, they may think they have gained more battery life. But system idle time does not translate to extending the usefulness of a system to me, and I'm pretty sure, not to anyone else either.

To compare this to phones with the 60+ hour standby times, yet using it brings it down to a few hours or minutes. The extended idle time in this context is important and useful. On a notebook, not so much.

That is why I am ignoring the great/impressive idle power levels SSD's may offer. They simply do not mean much in the real world with the systems I operate in.

 

Idle power consumption *IS* important because while you're typing away or even web browsing (especially if your cache is set to RAM) your SSD is idle 99% of the time. Now how all this actually translates into actual use entirely depends on the system it's being used in and what fashion.

 

HTWingNut, I guess that could be true for most.

But if I'm so bored to be typing/browsing, that means that PS/LR is churning away in the background and I'm just sitting there watching the wheels go 'round and 'round...

 

But then the appropriate comparison is comparing the Intel 730 against an array of, say, 15k rpm SAS drives which are obviously quite power hungry. But if you keep the applications and workloads relatively consistent, SSDs offer pretty significant overall energy savings. Perhaps not at peak loads, but simply by virtue that they can burn through the workload of a HDD in a mere fraction of the time and go back to idle.

 

Power needed per task? No comparison; an SSD should usually win because of it's race to idle design philosophy.

Real world power usage? No point comparing (unless a handful of minutes count...); when a computer is on the workload is not a fixed .

If we have a faster system; we tend to do more (work) with it (because we can) for a given amount of time/battery life.

If we leave the system alone; the O/S will optimize itself and run it's internal maintenance routines.

If all of this goes away (which it never really can...)? The SSD will run it's deep GC routines.

An idle system is a myth (especially a mobile one).

Until idle time can be measured in weeks and months on our more powerful mobile platforms, increases in idle time while on battery power is just theoretical and another 'score' to ignore when comparing components that suggest they increase it in laboratory only usage.