We start here in the same place as our previous XPS 13 review, with SiSoftware's SANDRA 2015, or the System Analyzer, Diagnostic and Reporting Assistant. Our testing consists of four of SANDRA's built in subsystem tests: Processor Arithmetic, Processor Multimedia, Physical Disks, and Memory Bandwidth.
Sandra CPU Arithmetic and Multimedia
Sandra Memory and Storage Bandwidth
SANDRA reveals some significant gains over the previous generation XPS 13. The new Skylake Core i5-6200U handles an impressive 66.21 billion instructions per second (GIPS), a 21.77 GIPS improvement over the Broadwell Core i5-5200U. Floating point operations only see a slight increase to 31.10 GFLOPS, a 0.11 GFLOPS increase. Media performance sees even more dramatic increases with integer scores jumping from 67.88 MPix/s (Broadwell) to 98.53 MPix/s (Skylake) and floating point scores increasing from 84.61 MPix/s (Broadwell) to 111.68MPix/s (Skylake). Memory bandwidth has expanded nearly 25% to 19.419 GB/s integer and 21.342 GB/s float from 15.42 GB/s and 17.375 GB/s respectively in the early 2015 model.
Finally, we find an incredible three-fold improvement in storage bandwidth over the older mSATA Lite-On drive thanks to the blazing fast Samsung PM951 NVMe M.2 solid state drive inside. Keep in mind that the advertised read speeds of 2200MB/s are very difficult to achieve in real world usage and are only applicable to certain controlled file transfers.
|
SunSpider |
JavaScript Benchmark |
|
The Webkit Open Source Project suggest the following about the SunSpider Javascript benchmark...
Unlike many widely available JavaScript benchmarks, this test is:
Real World - This test mostly avoids microbenchmarks, and tries to focus on the kinds of actual problems developers solve with JavaScript today, and the problems they may want to tackle in the future as the language gets faster. This includes tests to generate a tagcloud from JSON input, a 3D raytracer, cryptography tests, code decompression, and many more examples.
Balanced - This test is balanced between different areas of the language and different types of code. It's not all math, all string processing, or all timing simple loops. In addition to having tests in many categories, the individual tests were balanced to take similar amounts of time on currently shipping versions of popular browsers.
Statistically Sound - One of the challenges of benchmarking is knowing how much noise you have in your measurements. This benchmark runs each test multiple times and determines an error range (technically, a 95% confidence interval).
With Sunspider, we see a massive 18% improvement over the previous generation XPS 13. There's no question that SpeedShift is part of the driving improvement here given the burst-driven workload of the test. Interestingly, these numbers also beat out the Core i7-6500U in the Lenovo Yoga 900 by 10%. While we can speculate on the reasons for this, it should suffice to say that any of these front-runners will perform as well as any other in real world usage.