Wireless

The average laptop these days comes with a wireless card which can transmit and receive two spatial streams at once, and when it’s used with 802.11ac, that gives a maximum connection speed of 866 Mbps. The XPS 15 comes with a Broadcom wireless card which is 3x3:3, meaning it can work with a third spatial stream and has 50% more bandwidth available, which gives a maximum connection speed of 1.3 Gbps. Although on paper this is higher than Gigabit Ethernet, wireless rarely gets close to its maximum transmission speed whereas Gigabit Ethernet does. But assuming you have a router capable of 3x3 Wi-Fi, this solution should have a better shot of coming close to Gigabit Ethernet.

WiFi Performance - TCP

The XPS 15 sets a new AnandTech record on this test, beating out the Intel 8260 that only just jumped ahead on the last review. But to be honest I was expecting a bit better, if only because the 2x2 Broadcom solution could already hit 500 Mbps, so I was sort of expecting this to be over 700, but it’s still a good enough result that I didn’t feel like I needed to connect Ethernet when I was moving large files around on it, which isn’t something I can say about most laptops.

Thermals

When it comes to thermal considerations, shrinking the 15.6-inch form factor down, but still using higher performance (and therefore higher wattage) parts could be a concern. The CPU thermal design power is 45-Watts, and while NVIDIA doesn’t disclose the mobile GPU TDPs, it’s likely around 60-Watts or so. In most cases, GPU intensive workloads are not too hard on the CPU, and that’s true the other way as well. Laptop gaming, for instance, is very heavy on the GPU but generally not as taxing on the CPU, so in most scenarios Dell won’t have to remove the entire combined TDPs of both parts.

To test the CPU performance with a heavy and long duration workload, I logged the frequency and temperature of the CPU while running the x264 benchmark, which takes about an hour and runs the CPU near 100% load for the duration. As you can see, there was no thermal throttling at all in this workload, but the CPU certainly got warm, peaking around 84°C.

To test the GPU, I looped the Tomb Raider benchmark for about an hour, logging the GPU data every second. As with the CPU test, there was no thermal throttling of the GPU despite the heavy load and long duration. The GPU temperature never got over 79°C during this test.

Dell did a nice job on the cooling solution for the XPS 15. Despite the smaller chassis, it doesn’t appear to throttle under these workloads.

Noise

The other side of the cooling is noise. It’s an unfortunate side effect that in order to remove the heat, we have to rely on fans, and laptops with quad-core processors and GPUs are going to need fans. As with most laptops, the XPS 15 can disable its fans completely when lightly loaded, allowing quite operation, but once you do start to do some work, the fans slowly ramp up, starting around 36 dB(A) and going all the way up to 41.9 dB(A) on sustained light workloads. The XPS 15 can be silent, but once the fans kick in they are certainly audible.

Under heavy loads, the XPS 15 is, to put it bluntly, loud. I measured 52.7 dB(A) with the SPL meter 1” over the trackpad. This was after the GPU load test, where Tomb Raider was left running for over an hour. As a comparison, the Lenovo Y700 with the same CPU and GPU, and doing the same test, was closer to 44 dB(A) after the hour, and since decibels are logarithmic, that’s a big difference. This is the down side of a slim and light form factor: the fans are louder.

Audio

The XPS 15 has two speakers on the bottom of the laptop near the front. I measured 90 dB(A) from the speakers playing music, which makes this one of the loudest laptops I’ve seen. The sound quality is typical of a laptop, with not a lot of range and a lack of low end response.

Battery Life and Charge Time Final Words
Comments Locked

152 Comments

View All Comments

  • Hulk - Friday, March 4, 2016 - link

    It's really not that hard. 1. Always use a tripod. 2. White balance when shooting (white card) or in PS. 3. Go full manual or aperture priority to get the exposure right, spot meter. 4. Shoot the sweet spot of every lens which is generally f/8. Again use the tripod and timer so you can deal with long exposures which will be necessary.

    It's not like you're shooting a moving kid or portrait.
  • Beany2013 - Saturday, March 5, 2016 - link

    To be honest, the best way to get an amateur to take better photos is to hunt down a professional photographer and find out how much they'd charge for a consultation. Given that AT isn't a professional photography site, I'd wager most wouldn't mind not being the ones taking the photos, and would appreciate letting an interested amateur pick up some tips from the trade to improve their overall craft.

    I've been trying to get a semi-client of mine who operates a photobooth to do this for near three years now (I'm not a pro snapper so I don't feel I can teach him well enough) but he still doesn't do it - presumably, texting me is cheaper as I only charge in coffee and cake ;-)

    But it's definitely worth a pop. Taking a usable photo is fine, taking a really nice, professional *looking* photo is very much it's own reward.

    I agree with other commenters that the main meat of the article is definitely the words though; these photos might not be studio quality, but they're plenty clear enough for their purpose. Having them a bit nicer would be a bonus, and as a techy person and amateur snapper myself, I'm pretty certain Brett would enjoy learning how to bodge some nicer pics in - it's a lot of fun messing with this stuff with 'our' analytical kind of mindset.
  • Shadow7037932 - Saturday, March 5, 2016 - link

    Here's the thing, you don't need $10k in gear to take good photos. You can get a refurb D3200/D3300 (~$330-350) with a fast standard zoom with good sharpness (Sigma 17-50 f2.8 OS HSM) for ~$300-350. Add in an off camera flash or two (~$60-70ea). Add in a CPL filter (~$50-60 for a good one) for dealing with reflections and a light box (~$40-50 for a large one) and you're all set.

    As far as white balance and other things goes, shoot RAW, and deal with it in post. Or do it in camera using the custom white balance.
  • nagi603 - Monday, March 7, 2016 - link

    As someone who has invested into photography, you don't need 10k. Not even close. Frankly, your micron shot doesn't even have correct white balance. Fixing that would be a breeze if you shoot in RAW, use the same temperature lights, and used any post processing software. And yes, I've been there, I did shoot stuff as a journalist. Now I'm shooting events as a hobby, so I know what I'm talking about.
  • tuxRoller - Sunday, March 6, 2016 - link

    It's too bad the Verge doesn't spend as much time on article content.
  • Brett Howse - Monday, March 7, 2016 - link

    Photography is clearly not my strongest suit, but I've got some things I will try for the next review which I hope will help with the photos. Laptops are not the easiest thing for my to photograph and get a good result. Please bear with me as I try to step up my game on the images.
  • trenchtoaster - Friday, March 11, 2016 - link

    I have been reading AT for years now and this is the post which drove me to create an account and comment. I literally have no idea why the image you linked is any better than the ones from the article that you posted previously. What makes the image you linked so much better? I feel like I am missing something
  • sircod - Friday, March 4, 2016 - link

    I didn't notice it until you pointed it out, but that is pretty bad. Looks like he is shooting on a Canon EOS Rebel T4i, which apparently doesn't correct geometric distortion on its JPEGs.
  • Brett Howse - Monday, March 7, 2016 - link

    That's exactly what I am using, and apparently it does not. I, as an amateur photographer, did not even notice the distortion until it was pointed out, but I'll try to avoid it in the future.
  • close - Monday, March 7, 2016 - link

    You think someone buying a $1500+ laptop takes the decision based on how the pictures of the laptop look like on a website? Maybe people will assume the laptop is all bendy and with uneven color, right? I mean it's obvious they made one side of the laptop too bright and the other too dark.

    You go to a review site for he things you can't see in the store, not for some glamour shots. Yeah, of course they could be better but personally I don't think this subtracts anything from the value of the review.

    Also going for the IQ argument because you think the picture looks bad makes you look even dumber than your idea that "people won't buy the laptop if the picture wasn't shot RAW".

Log in

Don't have an account? Sign up now