+- PINE64 (https://forum.pine64.org)
+-- Forum: ROCK64 (https://forum.pine64.org/forumdisplay.php?fid=85)
+--- Forum: General Discussion on ROCK64 (https://forum.pine64.org/forumdisplay.php?fid=86)
+--- Thread: Rock64 4K new board; test burn-in (/showthread.php?tid=4921)
Pages:
1
2
Rock64 4K new board; test burn-in - MarkHaysHarris777 - 08-14-2017
The new Rock64 is running sweet !
... just doing burn-in now; got some python procs running; ssh terminal stuff; network and serial console right now, very pleased -- second Rock64 and many more to come.
Note: very interesting; unlike my pre release board, this production board (running the same image) is running much cooler; the cpu|gpu is cool to the touch, and the temps idling are in the high 30s ℃ ; loading the system with my PI routines drove the temps up to high 40s ℃ , and this without even a passive heatsink !
RE: Rock64 4K new board; test burn-in - KryPtAlIvIaN - 08-15-2017
Interesting. Good to know that has been addressed. How many hours are you into the burn-in test?
RE: Rock64 4K new board; test burn-in - fbms - 08-15-2017
Cool! I'm still waiting for my rock64.
RE: Rock64 4K new board; test burn-in - stuartiannaylor - 08-15-2017
Out of interest did the pre-release run @1.5Ghz and these now 1.3Ghz?
RE: Rock64 4K new board; test burn-in - MarkHaysHarris777 - 08-15-2017
(08-15-2017, 02:51 PM)stuartiannaylor Wrote: Out of interest did the pre-release run @1.5Ghz and these now 1.3Ghz?
That's an interesting question, because the answer is neither. I set the governors on-demand and most of the time the clock is either 800~ or 480~ don't remember the exact numbers but well below 1200 Ghz.
RE: Rock64 4K new board; test burn-in - stuartiannaylor - 08-15-2017
(08-15-2017, 03:30 PM)MarkHaysHarris777 Wrote:(08-15-2017, 02:51 PM)stuartiannaylor Wrote: Out of interest did the pre-release run @1.5Ghz and these now 1.3Ghz?
That's an interesting question, because the answer is neither. I set the governors on-demand and most of the time the clock is either 800~ or 480~ don't remember the exact numbers but well below 1200 Ghz.
I was watching the vids posted by Peter Feerick
https://www.youtube.com/watch?v=umLnsBlOtmo
Just noticed 1.3Ghz and just wondered if that was a safe setting he is using cpuburn to ramp up the processor.
RK3328 is rated at 1.5Ghz I think (max) but looking like 1.3 is how the Rock has been chosen at a safe lvl maybe
RE: Rock64 4K new board; test burn-in - Luke - 08-15-2017
(08-15-2017, 03:35 PM)stuartiannaylor Wrote:(08-15-2017, 03:30 PM)MarkHaysHarris777 Wrote:(08-15-2017, 02:51 PM)stuartiannaylor Wrote: Out of interest did the pre-release run @1.5Ghz and these now 1.3Ghz?
That's an interesting question, because the answer is neither. I set the governors on-demand and most of the time the clock is either 800~ or 480~ don't remember the exact numbers but well below 1200 Ghz.
I was watching the vids posted by Peter Feerick
https://www.youtube.com/watch?v=umLnsBlOtmo
Just noticed 1.3Ghz and just wondered if that was a safe setting he is using cpuburn to ramp up the processor.
RK3328 is rated at 1.5Ghz I think (max) but looking like 1.3 is how the Rock has been chosen at a safe lvl maybe
Since 4.4 build:
Quote:
- 0.4.4: dts: enable 1.4Ghz / 1.35V operating point for testing (Thanks @xalius)
Take a look in ayufan's git - description of releases. I remember that fire219 ran the board at 1.4Ghz with a hefty heatsink and large (relative to the board) fan -- it ran hot under load and throttled.
(edit) throttled without adequate cooling, is what I meant to say.
RE: Rock64 4K new board; test burn-in - MarkHaysHarris777 - 08-15-2017
I'm not pushing mine new board; for one , I don't have passive heatsink(s) for it yet!
cpuburn is a bad plan; its highly abnormal ; not a normalized test. Its guaranteed to make the system throttle, and it doesn't really simulate an actual real world environment that you might go bump in the night.
My point was not how hot can it go, my point is that in normal mode and | or idling, it runs cool; not my pre release board !
RE: Rock64 4K new board; test burn-in - MarkHaysHarris777 - 08-16-2017
I got the M3 mounting hardware today; typically I use M3 metric machine screws to mount my boards ( cased or not ) with a double set of washers|locknuts to adjust for height ( in and out of the case ); when not cased the M3 bolts make very nice stands|legs and since they are connected to ground 'bleed' very well to static desktops, mats, etc.
If you use this technique be careful because some of the components are very close to the mounts! Also, the holes on the Rock64 board are too small for M3 bolts ( just barely ). If you're careful, you can gently 'machine tap' them in with a hand driver and the legs are then very firm; frankly, the washers and nuts are not even required if all you want are legs !
Its still a do nothing machine for now; haven't really stressed it and have just been having a good time testing the images (installs and config) and doing some terminal work with C and Python.
Note:
I got to the bottom of the system running very cool; with low clock rate -- 408.00Mhz
The governor is set to interactive mode , and the serial console does not count as "interactive" because uart networking does not trigger the governor; setting the governor to "ondemand" bumps the freq both for uart and for smallish python scripts ; which gives a more realistic picture.
However, if all you care to do is lite computing via uart console using only passive cooling without heatsink, the Rock64 performs quite will and runs cool as well.
RE: Rock64 4K new board; test burn-in - korn2017 - 09-14-2017
The video jerks only at 4k @60fps but not at 4k @30 fps. Is this a hardware restriction or just a software problem?