Ah, that explains it - gaps. It did seem odd without context. Thanks.
I’m thinking up new test protocols, which you may or may not want to do, but would be more thorough:
1A (“idling”) - A “baseline” dataset, with everything on (modems, wireless, BT - but no useage), 100% screen brightness (on all the time), idling: let it go from battery from 100% to 0 (shutdown). Only the data logging script running.
1B (“idling with more realistic screen”) - Same as 1A, but screen at some lesser brightness that looks usable (75-85%?).
2A (“idling securely”) - Same as 1A, but HW killed with switches.
2B (“idling securely with more realistic screen”) - Same as 1B, but HW killed with switches.
3A (“some use”) - Same as 1A, but this time some usage, like a really long youtube video (no screensaver, Firefox)
3B (“some use with more realistic screen”) - Same as 1A, but this time some usage, like a really long youtube video (no screensaver, Firefox)
4A (“actual use in offline/secure mode”) - same as 2A, but this time some usage, like a really long video from memorycard (offline)
4B (“actual use in offline/secure mode with more realistic screen”) - same as 2B, but this time some usage, like a really long video from memorycard (offline)
I think those should give an idea of what to expect from current Dogwood and give context to measure against when future improvements arrive. These, in a matrix, should let us to approximate the effects of the screen (brightness), HW switches (online/offline) and light usage to battery life. Possibly to heat also, if we can bet those measurements too. I’m sure these can be refined - they need to be reproducible and simple, yet still cover the different profiles.
A bit dull to do, I’ll admit, but that’s base data for you. And I’ll understand if it’s a bit too much to ask.
I was thinking the same thing, I’m not sure how I could hold the screen on without use but I can try. I can tell you things seem to be moving very fast.
This is the data from my usage tonight, the first dip was it updating to the latest software, the period after that was full idle in lockdown mode, the lowest power drawl you can expect at any given time. The next period where it gets more steep was it in full use with decent load, 100% brightness, all switches on as if I had a carrier, and needed the camrea and mic, active use of the bluetooth (keyboard again) and wifi. (I was using a website based chatting service with my GF and decided to do this to get more data while I did)
For the screen, see: Prevent suspend & keep the screen on or then there is this hack. Or set up a propeller and tape a pencil on the other end of the axle and have that tap the screen…
That heavy-ish 1 hour use with gf seems to have shortened about 1.5 to 2 hours away from the battery, compared to idle.
Also, the phone properly got fairly hot for once during this due to charging and this being fairly intensive at the same time. Still not so hot that it was uncomfortable or anything, this is the hottest I’ve ever had the phone. So, guess it’s not advice able to run a benchmark while charging.
I did a totally unfair comparison with my Thinkpad T450s laptop (2-core/4-thread Core CPU), and the Librem 5 doesn’t do that bad in the comparison, when we consider the differences in power consumption and the cost of the processors.
I ran those same on a couple of years old work laptop (i5 4x2.53Mhz M540 2core/4thread) and it looks even better: 5.59 / 245.3 / 0.9 / 7.62 / 0.38 / 3.24 / 2.92 It’s not the killer of the latest laptops, but the difference is small compared to the weight and volume.
My guess is that @amosbatto’s table shows timings in seconds. In that case smaller is better, and the “Times better” column means how much faster the ThinkPad was compared to the Librem 5 for that particular test.
It would have to be timings in seconds for some rows and work achieved (e.g. hashes performed or bytes encrypted) in fixed amount of time in other rows to make sense to me.
I would run it myself but the program appears to hang on my computer - but that’s another problem …
The results (if you want to run your own tests) and keys are from hardinfo - or in my case the GUI version “System profiler and Benchmark”.
CPU Blowfish: lower number is better
CPU Cryptohash: higher
CPU Fibonacci: lower
CPU N-Queens: lower
CPU Zlib: higher
FPU FFT: lower
FPU: Raytracing: lower
Also, at the end of @dylanlesterpvcs data (and the GUI) there are reference results to compare also (although a bit older CPUs but still).
The guy on the PinePhone forum re-ran hardinfo in Mobian Nightly using a recent version of hardinfo, and now the results seem reasonable. Here are the results:
Benchmark
Thinkpad T450s (4x 2.7GHz Core i5-5200U)
Librem 5 (4x 1.5GHz Cortex-A53)
PinePhone (4x 1.152GHz Cortex-A53)
i5 vs L5 % difference
L5 vs PP % difference
CPU Blowfish
4.94
8.81
12.58
78%
43%
CPU CryptoHash
432.63
105.57
104.85
310%
1%
CPU Fibonacci
0.82
3.36
3.36
310%
0%
CPU N-Queens
7.01
5.31
10.27
-24%
93%
CPU Zlib
0.47
0.18
0.12
161%
50%
FPU FFT
2.10
9.86
17.81
370%
81%
FPU Raytracing
2.10
5.08
9.80
142%
93%
Based on the 30% difference in clock speed between the Librem 5 and PinePhone, and the much faster RAM in the Librem 5 (LPDDR4-3200 vs LPDDR3-1264), I expected a 40%-50% difference in performance, which is what we see in Blowfish and Zlib, but everything else surprised me.
For CryptoHash and Fibonacci, the Librem 5 and PinePhone were the same. For some reason, the i.MX 8M Quad is really good at N-Queens–even better than a fifth-gen Core i5.
It isn’t clear to me whether the floating point benchmarks are being run in the CPU or the GPU. I expect that the Vivante GC7000Lite in the Librem 5 will have twice the graphics performance of the Mali-400 MP2 in the PinePhone. @dos’s gaming tests (1 23) showed a big difference in the GPU performance. I wish that we had some more GPU benchmarks to compare the two phones.
For an individual (pre)order, @antpanlinuxmade a nice approximation on how fast they can be assembled, so it could be made into a formula which includes your number in queue and how many people are assembling the phones. Plus the shipping and such delays.