The XPS is using the new Broadwell-U Intel processor, which is a die-shrink, and lower power usage is expected from this.
I'm not sure why you think it's fishy. As long as Anandtech has kept their benchmark tests consistent across different units tested, it should be fine. I'm not sure how you got HD video viewing as a light benchmark test, I've never heard of that before. Generally, decoding a highly compressed chunk of data and using it to constantly update the display seems close to a worst case scenario to me, unless you are doing serious data crunching (e.g. encoding) which while more common than it used to be, I wouldn't consider common enough to include as a regular benchmark in a general purpose review.
According to Anandtech's article, the 13-inch Macbook Air has a 54 Wh battery and the Dell has a 52 Wh battery, so the XPS battery is slightly smaller. The hardware drawing power from the battery is different through (Broadwell-U @ 14 nm vs Haswell @ 22nm), so extrapolating numbers in that way is not likely to yield very accurate results.
HD video decoding is a fixed workload so higher end CPU's spend less time working and get to down clock. There are plenty of games for example where they simply max the CPU so it's a good idea for your heavy load useage to scale.
Anandtech then just displayed numbers for useage levels vs hours:minutes.
HD video decoding is a fixed workload so higher end CPU's spend less time working and get to down clock. There are plenty of games for example where they simply max the CPU so it's a good idea for your heavy load useage to scale.
Anandtech then just displayed numbers for useage levels vs hms.
I'm not sure why you think it's fishy. As long as Anandtech has kept their benchmark tests consistent across different units tested, it should be fine. I'm not sure how you got HD video viewing as a light benchmark test, I've never heard of that before. Generally, decoding a highly compressed chunk of data and using it to constantly update the display seems close to a worst case scenario to me, unless you are doing serious data crunching (e.g. encoding) which while more common than it used to be, I wouldn't consider common enough to include as a regular benchmark in a general purpose review.
According to Anandtech's article, the 13-inch Macbook Air has a 54 Wh battery and the Dell has a 52 Wh battery, so the XPS battery is slightly smaller. The hardware drawing power from the battery is different through (Broadwell-U @ 14 nm vs Haswell @ 22nm), so extrapolating numbers in that way is not likely to yield very accurate results.