"Throughout the test, the Arm Cortex-M0+ processor drew an average current from the compact Al-BPV of 1.4±0.4 μA with a voltage of 0.72±0.14 V. ... As the average power drawn from the compact Al-BPV during the 27 weeks of experimental run (1.05 μW) was larger than the minimum power required to run the CPU (i.e., 0.3 μW), it is likely that a smaller power-generating system could be used, or that more computationally intensive algorithms could be performed."
and:
"We programmed the Cortex-M0+ CPU to calculate a sum of consecutive integers as an example compute workload, and assess the correctness of the computation by verifying the sum against a precomputed value."
Pretty neat -- although it looks like they were clocking the chip at only 10 KHz.
I think the algae is great but this also shows how very, very little power you need for small embedded devices these days. You could probably run this CPU off of almost anything. Moldy old pop tart? Go for it!
The energy is not coming from the light or the algae, despite the fact that they tried to test it by blocking the light and trying it without algae.
The energy is coming from oxidizing the aluminum, and the algae are oxidizing it. Specifically the algae is making oxygen, which is reacting with the aluminum.
I would like to see them do with experiment using gold next time (i.e. non reactive metal) and I bet there will be no electricity.
> In the new work published in Energy & Environmental Science, researchers at Cambridge University harnessed algae’s ability to produce electricity during photosynthesis. They put blue-green algae, also called cyanobacteria, into a see-through case made of plastic and aluminum. Then they placed it on a windowsill and connected it to the microprocessor, which they had programmed to run for cycles of 45 minutes followed by 15 minutes of rest.
> Abstract: Sustainable, affordable and decentralised sources of electrical energy are required to power the network of electronic devices known as the Internet of Things. Power consumption for a single Internet of Things device is modest, ranging from μW to mW, but the number of Internet of Things devices has already reached many billions and is expected to grow to one trillion by 2035, requiring a vast number of portable energy sources (e.g., a battery or an energy harvester). Batteries rely largely on expensive and unsustainable materials (e.g., rare earth elements) and their charge eventually runs out. Existing energy harvesters (e.g., solar, temperature, vibration) are longer lasting but may have adverse effects on the environment (e.g., hazardous materials are used in the production of photovoltaics). Here, we describe a bio-photovoltaic energy harvester system using photosynthetic microorganisms on an aluminium anode that can power an Arm Cortex M0+, a microprocessor widely used in Internet of Things applications. The proposed energy harvester has operated the Arm Cortex M0+ for over six months in a domestic environment under ambient light. It is comparable in size to an AA battery, and is built using common, durable, inexpensive and largely recyclable materials.
Do not query the database upon each and every request (and use a CDN for media assets, but it's secondary) and you'll basically never die from being hugged.
Heck, a few month ago someone shared their website hosted on an old Android phone and it didn't have any issues handling the load from the HN crowd (which isn't really that big anyway).
Hugged is an euphemism that describes when demand is higher than supply for server/bandwidth resources.
Solution: increase supply and/or reduce demand.
Reducing the dynamic content (reduce the number of HTTP requests and the time taken for the server to respond to each request) goes a long way, but CDNs are good for increasing supply if that is ultimately necessary.
While CDNs are useful, you really shouldn't need them to handle HN traffic.
Someone posted what the otherside of a hug looks like[0], and it's somewhere around a unique visitor every 1-2 seconds, which really shouldn't be too much of a problem for... well anyone. The poster themselves got through it with a $7/month heroku server (albeit not without some manual intervention).
And, y'know, if sometimes a hobby project gets taken down, that's fair. You could end up getting 100 requests in a single second on a server you strung together off blog posts as a learning exercise over a weekend. But any time any professional organization gets taken down by the amount of traffic HN generates, I'm always a little embarrassed for them.
Yeah, at such a load, it's almost always something like too many DB connections/queries.
If you somehow use one DB connection per query and make multiple queries for every user, no wonder your website struggles under even a tiny bit of load.
The number of times I've been bitten by an orm and the n+1 problem...
Which, I guess is fine for some internal thing that needs to handle 8 requests a day.
I've kinda stopped using orms at this point. The sql is usually clearer than whatever dsl has been shoved into the orm. If you ever change databases, it requires slightly more work (assuming you haven't gone too crazy with db specific functionality), but burning a day or two to update the sql in your project is usually still quicker than orchestrating the migration.
The first computers were humans that computed things. This is thousands of times faster than a human at the task it is doing. It’s 1/10th or so the speed of the guidance computer that landed Apollo rockets on the moon. (Maybe as fast, depends on some things I’m not going to look up)
"Throughout the test, the Arm Cortex-M0+ processor drew an average current from the compact Al-BPV of 1.4±0.4 μA with a voltage of 0.72±0.14 V. ... As the average power drawn from the compact Al-BPV during the 27 weeks of experimental run (1.05 μW) was larger than the minimum power required to run the CPU (i.e., 0.3 μW), it is likely that a smaller power-generating system could be used, or that more computationally intensive algorithms could be performed."
and:
"We programmed the Cortex-M0+ CPU to calculate a sum of consecutive integers as an example compute workload, and assess the correctness of the computation by verifying the sum against a precomputed value."
Pretty neat -- although it looks like they were clocking the chip at only 10 KHz.