You know how this goes. I am going to purchase a new Chromebook but I don’t want to spend too much and I don’t want to be disappointed with my decision (buyer’s remorse). Fanning the flames of indecision is the vast array of current, new, and pending products. To add further cause for procrastination are the design choices which are across the board with traditional clam shell, 2-in-1, and rumored Chrome OS tablets with attachable keyboards.
And then there’s the performance dilemma. If I have to choose a low end processor to satisfy my budget, will I be satisfied with the performance? Many say yes, if the product yields a satisfactory benchmark score.
I am replacing my 2015 Chromebook Pixel and I want as many of the same features obtainable at the lowest possible price. In this case my budget is $500.
Common Benchmark Tools
Comparing benchmarks is one way to rank choices with similar features as most buyers want the biggest bang for the buck.
The following is a short list of the more popular core benchmark tools primarily used to test cell phones and tablets.
Google Octane v2
Not that the above tools can not be used to evaluate Chromebooks, but in the Chromebook world, the tool most widely referenced is Google Octane. Why is this?
As Chrome OS is browser centric and delivering web content is a complicated task, making sure the browser delivers the desired performance is critical.
Google describes the Octane benchmark as follows:
The key word in the above quote is “suite”, specifically Octane consists of 17 benchmarks to cover the most common use cases encountered on the internet.
Low, Medium, High Octane Score Comparison
The following table compares the Jide Mini Android PC, Acer R13, and the Chromebook Pixel 2015. From a SoC perspective I am comparing the quad-core Allwinner A53, quad-core MediaTek M8173C, and the Intel dual Core i5.
|Benchmark||Remix Mini||Acer R13||Pixel 2015|
|13. GB Emulator||3334||19840||55987|
|14. Code Loading||1665||5246||16812|
A couple of interesting observations about the data. Even though the Jide score was consistently low, the Allwinner 53 handled every benchmark with the least amount of deviation. The Chromebook Pixel scored high but the benchmark scores are all over the place. It’s best in show was the graphic GB Emulator benchmark.
It is not obvious from the chart but the Acer R13 MediaTek SoC was a huge step up from Allwinner. That is to say in 12 out of the 17 benchmarks the improvement in performance gained by MediaTek over Allwinner was greater than the improvement gained by Intel over MediaTek. My assumption is this trend would continue if the Intel dual Core i7 were added to the mix.
Living with the Baseline
It would be interesting to know the volume of sales of Chrome OS devices ranked by Octane. My assumption is there are more low end devices being sold (e.g. students) than high.
So how good or bad is it to use a device which scores on the low end of the Octane spectrum? I decided to find out by using the Jide as my daily driver. The following are my observations.
Most Everything Works
This is truly amazing. By selecting a low end device you will not remove large pieces of usability from the table. Although not blazingly fast, Google docs, Google sheets, Google drawing, YouTube, browsing the web all work. Granted, if your circumstance requires a dozen or more active browser tabs there may not be enough Octane at this level to fuel this scenario. Regardless of the number of open tabs, I found web sites which blast out ad upon ad or numerous multi media streams to be problematic.
I didn’t test Google Sheets with a large data set (e.g. 20000 rows or 2000000 cells) but I did test a document with multiple sheets, conditional formatting, and derived cell values from moderately complex calculations. This scenario worked but I can see a big data scenario which might challenge a device with a low Octane score. In my mind big data equals power user.
Code Load vs Initialization
It is interesting apps like Google docs load fairly quickly (code load benchmark) but then there is an additional delay as the app initializes. In the case of Google docs, the UI becomes dimmed and a small box displays with the message “Working”. I am not sure how to benchmark this but I would guess it is either “Richards” or “Typescript”. Once initialized, the user experience is good.
Good vs Good Enough
The Samsung Plus and Pro present a perfect opportunity to test the Chromebook good vs good enough theorem. My assumption based on the anticipated Octane scores is the Pro will be good (23000 Octane) and the Plus will be good enough (10000 Octane).
Some concern has been expressed about the ability of the Plus to deliver a snappy user experience while driving a 2400 x 1600 display. With advertised support for external 4K monitors (3840 × 2160) and as several of the benchmarks (Raytrace, NavierStokes, Mandreel, GB Emulator, Box2dWeb) test the ability to render graphics, I assume screen resolution will not be an issue.
I consider myself a demanding user as I log time on my Chromebook just about every day but not a power user as I seldom exceed four or five open tabs. With that said; I am going to save the $100, stay in budget, and give the Plus a chance to shine.