[Alex Rissato] proudly reports that he now holds the record for highest benchmark score on HWBOT (machine translation); something he sees not only as a personal achievement but admirably, of national pride. Overclocking a Raspberry Pi is not as simple as achieving the highest operational clock rate. A record constitutes just the right combination of CPU clock, memory clock, GPU clock and finally the CPU core voltage. If you’ve managed to produce that special sauce, the combination must be satisfactorily cooled and most importantly be stable enough to pass an actual performance benchmark.
[Alex] realized that the main hurdle to achieving the desired CPU clock was the internally generated and hence restricted, CPU core voltage; This is externally LC filtered and routed back to the CPU on a stock Pi. [Alex] de-soldered the filter on the PCB and provided the CPU with an externally generated core voltage.
Next, the cooling had to be tended to. Air cooling simply wouldn’t cut it, so a Peltier based heatsink interface had to be devised with the hot side immersed in a bucket of salt water. All of this translated to a comfy 16C at a clock speed of 1600 MHz.
Was all the effort justified? We certainly think it was! Despite falling short of the Pi zero CPU clock rate record, currently set at 1620MHz, [Alex] earned the top spot in the HWBOT Prime overclocking benchmark. Brazil can now certainly add this to its trophy cabinet, arguably overshadowing the 129 Olympic medals.
Hackers have a long history of overclocking CPUs ranging from desktop computers to Arduinos. [Jacken] wanted a little more oomph for his Pi Zero-Raspberry Pi-based media center, so he naturally wanted to boost the clock frequency. Like most overclocking though, the biggest limit is how much heat you can dump off the chip.
[Jacken] removed the normal heat sink and built a new one out of inexpensive copper shim, thermal compound, and super glue. The result isn’t very pretty, but it does let him run the Zero Pi at 1.5 GHz reliably. The heat sink is very low profile and doesn’t interfere with plugging other things into the board. Naturally, your results may vary on clock frequency and stability.
Some people are never happy. [Jackenhack] got hold of a couple of shiny new Raspberry Pi 3s, and the first thing he did is to start overclocking them. Fortunately, he knows what he is doing, so none of the magic smoke escaped, but it seems not all Pis are happy with the process.
For one of the three seemingly identical Pi 3, adding heat sinks let him push the CPU from the native 1.2GHz up to 1.45GHz. That did involve a bit of overvolting (increasing the voltage to the CPU), but that can be easily done in software. He also experimented with adding heat sinks to the memory, then bumping up the speed of the memory to increase throughput. Again, he was able to make some impressive gains, bumping the speed up from the native 400 Mhz to 500 Mhz. Both of those are stable overclocks: he was able to run the system at 100% CPU load for an extended time, and has incorporated the overclocked Pi into his system that contributes to the NTP pool project.
However, when he tried the same overclock with the second of the Pi 3 victims test subjects, it failed due to the CPU overheating. So, it seems that there is a lot of variation in the individual bits of silicon on the Pi 3. Perhaps some liquid nitrogen would help? It did for an Arduino…
Some guys build hot rods in their garage. Some guys overclock their PCs to ridiculously high clock frequencies (ahem… we might occasionally be guilty of this). [Nerd Ralph] decided to push an ATTiny13a to over twice its rated frequency.
It didn’t seem very difficult. [Ralph] used a 44.2 MHz can oscillator and set the device to use an external clock. He tested with a bit-banged UART and it worked as long as he kept the supply voltage at 5V. He also talks about some other ways to hack out an external oscillator to get higher than stock frequencies.
We wouldn’t suggest depending on an overclock on an important or commercial project. There could be long term effects or subtle issues. Naturally, you can’t depend on every part working the same at an untested frequency, either. But we’d be really interested in hear how you would test this overclocked chip for adverse effects.
Now, if you are just doing this for sport, a little liquid nitrogen will push your Arduino to 65 MHz (see the video after the break). We’ve covered pushing a 20MHz AVR to 30MHz before, but that’s a little less ratio than [Ralph] achieved.
Give your valentine an analog love note on the big day. [Tom’s] LED heart chaser design does it without any coding. It’s a 555 timer with CD4017 decade counter. The nice thing about the setup is a trimpot adjusts the chaser speed.
While we were prowling around DP for the last link we came across [Ian’s] post on a new version of Bus Pirate cables. We’ve got the old rainbow cables which are pretty convenient. But if you’ve used them you’ll agree, hunting for the correct color for each connection isn’t anywhere near a fool-proof method. The new cable uses shrink tube printed with probe labels. They sound like a huge pain to manufacture. But this makes connections a lot easier. In our experience, when it doesn’t work its always a hardware problem! Hopefully this will mean fewer botched connections.
Make your tiny LiPo cells last longer. Not capacity wise, but physically. The delicate connections to the monitor PCB break easily, and the plug is really hard to connect and disconnect. [Sean] shows how he uses electrical tape for strain relief, and a bit of filing to loosen up the connector.
KerbalEdu: Kerbal Space Program for education. That’s right, you can play Kerbal as part of school now. Some may shake their heads at this, but school should be fun. And done right, we think gaming is a perfect way to educate. These initiatives must be the precursor to A Young Lady’s Illustrated Primer method of education. Right?
Some of our younger readers will never have experienced this before, but back in the day your video games would slow way down if there were too many moving objects on the screen. The original Castlevania comes to mind, but many will remember the problem while playing the fantastically three-dimensional Super Nintendo game Starfox. [Drakon] isn’t putting up with that hardware shortfall any longer, he hacked this cartridge to run at 42 MHz, twice as fast as the design spec.
We only occasionally look in on the cart hacking scene so it was news to us that three different versions of a pin compatible chip were used in this hardware. The first two suffer from the slowdown problem, but the final revision (SuperFX GSU 2) doesn’t. It can also be overclocked as high as 48 MHz but because of the video frame rate you won’t see added improvement with the extra 6 MHz.
[Drakon] used a Doom cartridge as a guinea pig because it offers the most RAM, and set to work rerouting the traces for the ROM chip to an EEPROM so that the hardware can be used with different games. He also took this opportunity to patch in the faster clock signal.
[Fred] likes to squeeze every cycle possible out of his graphics card. But sometimes pushing the clock speed too high causes corruption. He figured out a way to turn a knob to adjust the clock speed while your applications are still running.
The actuator seen above is a Griffin Powermate 3.0. It’s a USB peripheral which is meant to be used for anything you can imagine. [Fred] uses an AutoHotKey script that he wrote to capture the input from the spinner, process that information, then adjust GPU clock speed in the background. Since the clock on his ATi Radeon 5800 can be adjusted using the AMD GPU clock tool, it’s an easy choice for this application. Now better graphics are at the tips of his fingers. See for yourself in the video after the break.