Overclocking discussion
-
@thelostsoul which emulator/plugin are you using for Yoshi's story? There is a know issue right now with mupen64-glide on the pi right now that is bugging games out after about 10-20 minutes of playtime. It's currently being investigated.
Edit: Just tested Yoshi's story. This is a different issue than the one I mentioned above. Yoshi's story must be on the mupen64 config blacklist because no matter which plugin I pick it reverts to using mupen64-glide. I tested using lr-mupen64 (which doesn't look at the blacklist) and was able to get past the glitch. However just a little further into the game, the glitch occurred and it froze the emulator.
I was able to force Yoshi's story to use mupen64-rice by disabling the mupen64 compatibility check. There are some visual glitches using rice but it doesn't have that same gamebreaking bug that glide does.
-
You could always try to delid the IHS and apply liquid metal.
-
@quicksilver I tried different emulators for quick testing. Don't remember which one exactly, but the recommended one from compatibility list in documentation, the mupen64 auto setting and one random.
Its frustrating, because it starts working so good, almost perfect until the glitch happens. So, which problem is it? Is there a thread for? So i don't off topic here anymore. :D -
@thelostsoul it doesn't matter which emulator you pick to try to run Yoshi's story. It forces mupen64-glide (if you see the fps counter it's using glide). You can use mupen64-rice but only if you remove the compatibility check. Mupen64-rice has some visual glitches running Yoshi's story but that glitch doesn't occur.
I don't know if there is an existing thread for this issue. There are quite a few n64 games that have issues like this only on the pi. It's usually due to the fact that the Pi's GPU drivers are old garbage and don't have certain features.
-
@quicksilver Ok then, I thank you, you saved me a lot of time. I am glad I asked here.
-
@rion said in Overclocking discussion:
You could always try to delid the IHS and apply liquid metal.
That was an interesting video. It looked like some hefty glue. I think direct die cooling definitely works. It's as stripped down as you can get before making contact.
It seems the BIQU and Flirc use the same basic cooling technique. Which is "case heatsinking" I would call it for short by making direct contact with the chips. The BIQU does 3 things different/better.
It comes in direct contact with all 3 main chips (Flirc only 1 chip), it uses solid aluminum alloy columns (Flirc uses hollow aluminum alloy columns), The BIQU uses grease (Flirc uses thick tape or a very thick layer of grease). These differences make a 10 to 20 degrees Celsius difference.
The BIQU and "open case" are kind of go about it entirely different. The BIQU of course "case heatsinking". The open case relies more on open air dissipation and the use of fans to move the heat away quicker. I used tape on the open case but grease might drop those temps possibly.
I think that would be a good test. Thin tape vs. thin grease. We know thick tape vs thick grease is no contest except it does hold much more solid. Which might be more important than cooling in some scenarios. At a minimum it might be a mute point of a degree or two. I will test that at some point.
The only way I see to improve the open case is a bigger heat sink because the gap between the chip heatsink and fan ( which is air) has a higher resistance vs. metal on the BIQU which has proven to be more effective. Even when the case is fully heated up and under load.
The BIQU might run a bit cooler with passive venting on the top (like those cool raspberry Pi cut outs--also avoiding the use of fans). That way "general" heat buildup from other components that has to dissipate through the walls can just rise up naturally through the venting, as heat likes to rise. The BIQU does have a good size slot for a ribbon on the side that probably does allow a good amount of heat to escape but the Flirc is closed up solid on the sides and open on the bottom only.
-
@riverstorm said in Overclocking discussion:
The BIQU uses grease
I just want note something. My BiQU doesn't come with grease and so I used it without. Didn't know it comes with, thought you applied it from somewhere else. When I applied thermal paste from my desktop pcs cpu, it made the difference (20 degrees). Just in case someone gets the BiQU without grease or something else.
-
@thelostsoul said in Overclocking discussion:
I just want note something. My BiQU doesn't come with grease and so I used it without. Didn't know it comes with, thought you applied it from somewhere else.
You're right, that's a good point. It would need to be ordered separately and it would raise the overall price of the BIQU. I must have about 10 syringes at home so I almost take it for granted when I grab one.
-
@riverstorm said in Overclocking discussion:
The only way I see to improve the open case is a bigger heat sink because the gap between the chip heatsink and fan ( which is air) has a higher resistance vs. metal on the BIQU which has proven to be more effective. Even when the case is fully heated up and under load.
I've been linking to this before but this is the best passive cooling solution i have seen so far.
This modification with an acrylic sheet would've be possible to do on the official case and the nespi for example. But instead of using super glue you could attach the heatsink with epoxy to the sheet instead.
-
@rion Looks massive. But is it better than BiQU, where the whole case is a heatsink, touching also the other two chips. Oh btw, I see I saw that video already and commented it.
-
@rion said in Overclocking discussion:
I've been linking to this before but this is the best passive cooling solution i have seen so far.
That's impressive. I think he was only using thermal grease. I like how he made the sheet plastic shield to protect against shorting. Also the hot glue is smart. It holds it in place but very easy to remove.
The no sink, small heatsink and copper plate were all close within a few degrees. Temps definitely scale with the size of the heatsink. 30C difference between none/small to large. It seems more important to remove it from the immediate chip and once on the heatsink/case the Pi just runs cooler.
-
@riverstorm At some point though, I think some of these cooling setups start to get into overkill territory. I have seen people liquid cool their pi's, but what is the point exactly? Even at the max over voltage setting, I dont think you can heat your pi up enough to need something like that.
-
@quicksilver I think those guys are overclocking for science or like sports/hobby/fun or for their YouTube channels. There is no practical reason to do extreme overclocking with extreme cooling it. Same goes for desktop pc cpu and gpus.
-
@quicksilver - I definitely agree and ask myself sometime why the hell am I sitting here doing this but as a hobby I enjoy it quite a bit. I know guys that spend crazy amounts of money and time hunting (guns, ammo, clothing, food, lodging, etc.), same as fishing, camping, RVs, etc. Sorry for the examples I grew doing mostly outdoor types of things but the point is because we can! ;) No actually probably no practical reason except enjoyment. Instead of pushing speed in cars we like to do it without wheels! ;)
-
-
@riverstorm I meant why bother with such crazy cooling methods. I understand why someone would do extreme overclocking, heck I am all for that. But once your cooling needs are met what is the point of further cooling?
My comment was more about the crazy stuff people do on youtube etc, not about the tests that you have been running (I have found those to be most interesting). There seems to be this misconception that a lot of people have that if you cool your pi enough you can overclock it further. And that is only partly true. The pi can only be overvolted up to a value of 6 (8 if you force turbo), so there is a limit to the max overclock you can achieve on your pi regardless of how much cooling you have.
-
@quicksilver said in Overclocking discussion:
There seems to be this misconception that a lot of people have that if you cool your pi enough you can overclock it further.
Yeah I agree the returns on cooling diminish to the point of pointless. Even running 60 to 70C is well below 90C where it would throttle and I am good with that. I can't imagine running at 50C is going to extend the life of the Pi by that much or if it does I prefer the simple case to an elaborate water cooling loop. I wouldn't be against trying it but I have so much Pi paraphernalia laying around and really need to just clean house.
-
@riverstorm said in Overclocking discussion:
I can't imagine running at 50C is going to extend the life of the Pi by that much
Especially if you overvolt so much, then. Not sure whats worse, high voltage or high temperatures near the limit?
-
@thelostsoul said in Overclocking discussion:
Especially if you overvolt so much, then. Not sure whats worse, high voltage or high temperatures near the limit?
Yeah it seems high temps are a killer and a by product of overvolting but once force turbo is in play all bets are off and you frying the board with voltage vs. temps is a possibility.
Probably why they flip the bit and void the warranty. The thing I find interesting about that is I have never heard of such a thing when it comes to CPU's and even then when a CPU is fried you can't boot it. I've blown a CPU or two over the years and Intel has never sent me a sorry buddy no can do letter. I always wondered how they pull that information or if it's smoke and mirrors. Especially if you did something like send it through a degausser. Pop and it's all done.
I am ok overvolting right up to 6 and leaving it there. I think I would be more concerned sitting around 90C vs. overvolt 6 but with force tubro bumped to 8. I don't know which I would find more concerning.
-
@thelostsoul Both can technically shorten the life of your pi. But the raspberry pi engineers designed the pi with limits in place to protect it. That is why by default you cannot over volt past a value of 6 and the pi is designed to throttle itself once it hits a certain temp. Even if the overall lifespan of my pi is shortened a little bit by overclocking I dont really care since it is a $35 computer that I will most likely replace due to obsolescence not equipment failure.
Contributions to the project are always appreciated, so if you would like to support us with a donation you can do so here.
Hosting provided by Mythic-Beasts. See the Hosting Information page for more information.