Battery Life

Perhaps more than most devices, a small laptop like the ThinkPad A285 is more likely to be used away from the desk, and therefore battery life is fairly important. Lenovo offers a 48 Wh battery inside the 12.5-inch chassis, which like the Intel based ThinkPad X280, has moved away from the dual-battery system and moved to just a single battery which isn’t replaceable by the end-user. This lets Lenovo cut some size and weight, but one of the nice factors of the older models was an optional larger battery for those that needed it, or the ability to swap out the rear battery if needed.

To test the battery life we have several scenarios. Our 2013 battery life test is very light, and with today’s PCs is almost an idle rundown. The newer battery life test is much more CPU intensive, and generally is more representative of what you’d see for basic tasks. Finally, we test the laptop on movie playback. The display is calibrated to 200 nits to keep a consistent brightness across all devices.

2013 – Light

Battery Life 2013 - Light

With just a 48 Wh battery, the ThinkPad A285 can’t compete, and sits right at the bottom with the worst battery life we’ve seen in a while, at just one minute over six hours. We’ll dig into why in a minute.

2016 – Web

Battery Life 2016 - Web - Normalized

Our newer web test taxes the CPU quite a bit more, and generally knocks down the battery life significantly on any low-powered device. That is the case here, but the difference is not as dramatic as some of the other devices.

Movie Playback

Battery Life Movie Playback

Generally movie playback is the most efficient test we have, but there’s something not right with Raven Ridge and movie playback. The decode can be offloaded to the media block, so in theory the CPU should be in its lowest power state for this test, but surprisingly the A285 fares worse here than in the web test. That result is mimicked by the other Raven Ridge laptop in our graph.

Normalized Results

Battery Life 2013 - Light Normalized

Battery Life 2016 - Web - Normalized

Finally we remove the battery capacity from the equation to see how many minutes of battery life each device gets per Wh of battery. The Raven Ridge devices sit right at the bottom.

Digging Into Power Draw

On any system there are always various components drawing power at all times. The display is generally the highest power consumer on any system at idle, and the amount of energy it draws varies on the size, brightness, and resolution. For instance, the display of a Surface Book 2 15-inch draws about 7.65 Watts at 100% brightness. Setting the display down to 200 nits drops its power draw to 4.1 Watts. Meanwhile the rest of the system pulls just 1.43 Watts at idle.

Raven Ridge, it appears, has a power draw problem. We saw it on the Acer Swift 3, and we’re seeing it again on the Lenovo ThinkPad A285. The base power draw is quite high, which makes the battery deplete much more quickly than it should. Let’s look at some numbers:

For this test, we've set the display to 100% brightness while the rest of the system is idling, with Windows set to shut off the display after a couple of minutes. The graph above includes both the Surface Book 2 (Green/Blue) and the ThinkPad A285 (Red), plotting the total power draw of each system. Since the display itself is only on for a couple of minutes, the first part of the graph is with the display at 100%, while the remaining section of the graph shows the system's average idle power draw.

Breaking things down, we see that the ThinkPad A285 starts out with an average power draw of 8.62 Watts with its display on and the system idling. Waiting a few minutes for the display to turn off, and we see that the average system power consumption drops by just over 4 Watts to 4.6 Watts. This is a significant drop, but not significant enough. The Surface Book 2 in the same scenario idles at a little more than one-third of this, meaning the A285 – and specifically, the Raven Ridge platform inside – is drawing an additional 3 watts at full idle. This is a huge amount of power to draw at idle, and leads to the less than amazing battery life results on Raven Ridge devices.

Some of the draw will be from the DDR4 memory, since most Ultrabooks leverage LPDDR3 to reduce the power usage, but that is not going to be all of it. It looks like AMD has some work to do in order to get its power usage in check if they want to better compete in the laptop space. They need their Haswell moment.

Charge Time

The flip side of the coin is charge time, and Lenovo tends to offer quite quick charge rates. The ThinkPad A285 ships with a 65-Watt AC Adapter, which is much larger than the adapter that ships most laptops of this size, and it is rated to charge 80% of a charge in just 60 minutes. It connects via USB Type-C.

Battery Charge Time

Not only does the 48 Wh battery get filled to 100% in the shortest amount of time, Lenovo even beats the 60 minute time for 80% by seven minutes.

Display Analysis Wireless, Audio, Thermals, and Software
POST A COMMENT

73 Comments

View All Comments

  • jardows2 - Tuesday, December 18, 2018 - link

    This chassis is not too thin for an Ethernet port. Our organization just ordered in a bunch of HP probook 450's that are every bit as thin as this, and they all include an Ethernet port. To make Ethernet connectivity a dongle is just a stupid business decision made by Lenovo for this class of laptop. Reply
  • HStewart - Tuesday, December 18, 2018 - link

    It not that thin of notebook, it include Type-A USB so it still fat notebook - I would think it possible to include ethernet at back part of it - probably a cost saving measure - just like the display. Reply
  • IGTrading - Tuesday, December 18, 2018 - link

    I find it ridiculous that the reviewers felt the unexplained need of including 15" laptops and notebooks with a discrete video card and compared them with a 12.5" Thin&Light portable without a discrete GPU.

    Why the heck are we comparing plums to grapes and apples to pineapples ?!

    The most basic journalistic "rules" for an initial review are to compare products priced similarly and of course, products with comparable specifications.

    Does anybody care to compare a Lamborghini to a Ford K ? Sure those are both cars, both use petrol and both have 4 wheels ... but really !?!

    You want to have an expert comparison between similar products from DIFFERENT product lines .... to show just how much extra performance does a larger chassis bring ... or a higher price ... ok, do it and present it as such.

    Here's a 12" ThinkPad with X CPU vs a 15" ThinkPad with the same X CPU (or similar) and here is the extra performance the larger and more expensive product gets you.

    But why the heck would you compare a 15" Microsoft Surface 2 with a discrete GPU costing 2300 USD with a 12.5" ThinkPad with just an APU (no discrete GPU) costing half the price ?!?!

    Is this to bury the good test results of Lenovo's AMD-based ThinkPad and make it look mediocre ?!

    If Ford launches a new Ford K ... what kind of reviewer will compare it and test it against the newest Ferrari ?!

    This is being forced (on all reviewers) by Intel's PR team. You either do it, or next time you have ZERO access to anything Intel-related. (happened to me too, many years ago)
    Reply
  • nico_mach - Tuesday, December 18, 2018 - link

    It doesn't seem to have a great analogue in these numbers, true. But it is very close to the Surface Pro i5 results, which is much thinner and lighter, so it still doesn't look good. Unfortunately, the prices aren't listed in these performance results - so the dGPU and i7 that look so good aren't really in the same league, but we don't know by how much.

    The results are unfortunate, but this is still progress by AMD, considering where they were 2 years ago. And getting Intel to respond the way they have - buying AMD graphics, matching their core counts, moving on with a hybrid 10nm scheme - really shows how competitive they've become. And when was the last time Intel had to resort to dirty tricks? Unfortunately I think both Intel and AMD should worry more about Quallcomm and Apple than each other. This comparison is ultimately irrelevant.
    Reply
  • DanNeely - Tuesday, December 18, 2018 - link

    Adding the 1060 was arguably gratuitous; but all the fanboiing on the Ryzen mobile had been about how it was going to crush the competition so Intel's mainstream parts and NVidia's entry level MX150 are entirely appropriate. And Lenovo's performance here is just awful. With 8 clusters of GPU cores instead of 10 it should be able to do at least 80% of the performance of the 15W Ryzen 7 in the Swift 3 (and a bit closer should be expected since it can use its TDP to clock the GOU higher); but instead it's coming in around 2/3rds. Reply
  • StevoLincolnite - Tuesday, December 18, 2018 - link

    There is allot of issues in regards to Ryzen mobile, drivers and TDP.
    In some instances, limiting the CPU will allow the GPU in the 2500u to run faster than the 2700u.

    I have the Lenovo E585 myself with a Ryzen 2700u. - It does Overwatch, 60fps, 1080P at 56% scaling all settings low.

    The driver situations is certainly bullshit though, my notebooks drivers are months out of date... AMD doesn't offer any newer revision... And if I "work around" the drivers and install unofficial ones, windows update will override it at a later date and downgrade them.

    I honestly expected more out of it... Going to upgrade the Ram to 16GB with lower latencies and hope that might make a small difference. My next notebook will certainly have a discreet GPU.
    Reply
  • IGTrading - Wednesday, December 19, 2018 - link

    Yes, the mobile driver for Ryzen needs work and needs it quick. If OEMs don't give a crap about their customers and offer no decent support, AMD should step in with a Beta mobile driver. Doesn't need to be WHQL, doesn't need anything special ... just offer it.

    As for testing in exactly the same conditions, HP x360 has Ryzen and Intel models in the same chassis. When tested these two, AMD Ryzen trounced the Intel model.
    Reply
  • Hul8 - Tuesday, December 18, 2018 - link

    It's helpful to have the *one* token example with a discrete GPU, so readers can tell how much you give up by going integrated. Otherwise the review would be in a bubble of similarly handicapped systems. Reply
  • HStewart - Tuesday, December 18, 2018 - link

    My comment was nothing about Intel - your response falsely assume it was.

    Most notebook users use notebook for word processing, spreadsheets and such and don't need descret CPU.

    What does the GPU matter in comment about USB Size? I had to look up the specs to see what specs it had - is Vega 8 even descrete GPU? it not part of discussion - just falsely assume it was bias statement.
    Reply
  • MonkeyPaw - Tuesday, December 18, 2018 - link

    I agree when it comes to some of the performance tests. The dGPU model really skews the graphs by being so much better than iGPU models. Makes a comparison of similar models harder to do. Reply

Log in

Don't have an account? Sign up now