Hennessy and Patterson's Computer Architecture: A Quantitative Approach is the definitive textbook. They also have a 2nd textbook Computer Organization and Design. The former is more focused on people who might go after a Computer Engineering degree while the latter is more an exploration of microprocessor architecture relevant to a wide audience of programmers.
Previous editions are available free online and just as good for learning the big picture imo. Actually the older editions might be better for learning microarchitecture specifically because the more recent editions have cut some material on that in order to cover mobile and cloud computing.
It's even less true for Forth, given that it has separate data and return address stacks, and neither is addressable. There is some dedicated Forth hardware that relies on this fact to separate the stacks physically in hardware and e.g. use SRAM for them.
I still find it expensive for a thermal imager. I personally bought a Chinese Guide PC210 for €216 shipped, and that includes 21% VAT, for a whole unit and not just a sensor. For another €15 I got a ZnSe lens for macro use.
Not only is it cheaper, on paper it's also much better:
- 45 mK vs 50 mK thermal sensitivity
- ±2 °C or ±2% vs ±10°C or ±10% accuracy
- 256x192 vs 160x120 resolution
- 25 vs 8.7Hz framerate
- -20 to 550 °C vs -10 to 450 °C range
The kicker is that (being European) I can't even get a FLIR equivalent to some of these specs, as they would violate ITAR export restrictions.
Emphasis being "on paper". When I worked for a research company we got a cheaper Korean company to demo their thermal cameras. They were just way way worse than FLIR and the software was abysmal (FLIR's software is surprisingly good).
Ended up going for the £4k FLIR camera (not my money tbf). On the plus side I didn't even know about the 9fps thing at the time and they didn't ask me about it, but we still got a 25fps camera. Not exactly sure how that happened.
I intentionally mentioned the on paper part because I can't attest to the absolute accuracy, but in the aspects that are easily comparable I would say it's way ahead of the FLIR C5 that cost 4x as much I use at work.
Both are used for diagnostics of electronics, so the resolution and framerate make a noticeable difference here. On top of that mine boots up and is ready to go in under 5 seconds, while the FLIR takes close to a minute to get ready for some reason.
Likely because it is self-calibrating or waiting for some internal temperature to stabilise.
For a lot of uses that doesn't really matter tbf. IIRC in the FLIR software you can disable the periodic self calibration they do (which is quite annoying). Maybe that would improve startup time.
One of the European sellers of the camera (Eleshop) sells a lens and 3d printed holder. I just bought the same lens and 3d printed a holder myself, from what I remember someone on the eevblog linked a suitable one.
Separately, there's a notion of purchases large enough you have to plan for them. $523 is a few weeks of pay for some people even in the US. The headline is set up to suggest a low-cost (because it's DIY) relative to the status quo of devices an average student couldn't really buy to tinker with. The result is more expensive than an off-the-shelf thermal camera I bought awhile ago without substantially better features (other than tinkerability).
You are right, there are cheaper solutions like smartphone addons, but as a standalone thermal imager including display, storage, etc., there are not many options that are less expensive and provide the same flexibility in terms of open-source software and hardware. This is a comparable all-in-one device from FLIR itself: https://www.amazon.de/FLIR-89401-0202-C5-Kompakt-W%C3%A4rmeb...
I have zero background in this area, just following the rabbit hole -- there seems to be open access to ESA's Sentinel missions: https://scihub.copernicus.eu/
The Sentinel data is excellent, and spans several satellite platforms and a huge number of modalities. In some significant cases (surface deformation, imaging spectroscopy, for example) Sentinel surpasses (current!) NASA data. (I use both in various projects.)
In general there are a lot of trades, including ground sampling distance, temporal revisits, and spectral coverage. So it’s good to have options.
Both NASA and ESA do exceptionally well at offering free data for science purposes. All the NASA science data, Earth and otherwise, is free.
People think of this data as a camera that’s pointed at a target, but that’s not how it works now. There are calibration and inversions that have be done to transform the observed electromagnetic data into physically-relevant quantities.
I’m not a commercial user, but it is possible that there are charges for some commercial purposes, or for niche applications that require high volumes or low latency. (Although many missions offer near real time feeds for free as well as slower calibrated feeds.) The data egress costs are a concern for NASA Earth science…I don’t remember the specific numbers but they have been increasing rapidly.
It’s easy to get confused. If you are a potential science user, be sure to get in touch with another savvy user in your domain, or with someone from the instrument team, before you write it up.
Hey, thanks for this link, it looks interesting but I’ll have to find out what kinds of measurements they provide. The User Guide seems to have a complete list:
There are several interesting things there, like CH₄, SO₂, etc.
They also include data derived from “Suomi-NPP VIIRS Clouds” which is the NOAA instrument used by the Gas Flare Tracker, matched to the ESA data.
It would be great news if they changed their policies in these years.
(A couple of minutes later, checking while writing this comment...)
Oh... “Please login to access our services...” Ok, I might take a look at it later.
Thanks anyway for the link, I might have seen a previous version of this page but I don’t remember seeing the detailed product (measurements) list. It does look more promising that it used to.
Meanwhile in USA: “The VIIRS Nightfire (VNF) digital data access will be transitioned to restricted access for organizations that have approved data user license agreements with EOG. Commercial users can also be approved, but will be expected to make annual payments to cover a portion of the VNF program costs.”
https://eogdata.mines.edu/products/vnf/
The ESA is just as open as nasa, and the freely available Copernicus images benefit US citizens as well. This pissing contest that you try to start is ridiculous. Americans and Europeans are very happy with each other in this regard.
I’m sorry I wasn’t clear: the last paragraph quoting the current conditions for using the VIIRS Nightfire data, where as you correctly note the conditions of use are similar, is a change that happened afterwards. I should have written “Meanwhile in USA, things have changed: ...”.
During the years when I worked on the Oil Spill Monitor and Gas Flare tracker, from 2013 to 2019, NOAA gave the data completely without conditions: anyone could go to the Suomi-NPP website and just download the CSV files without login or anything else, and not just the current one (the satellite does one pass per night over the whole world) but also historic data.
Now things have converged, and your statement that "ESA is just as open as NASA" is, at least when comparing the Sentinel vs. the VIIRS Nighfire data, absolutely correct.
Many companies, like mine, are somewhere in the middle of a 10 year lease with very steep early exit penalties. People think that having everyone working from home saves the company money, but when it comes to the office, chances are it’s already a sunk cost. In my company’s case, we’re paying for quite a bit of space, but currently allowing people to choose whether to come in or not. Most of the time only a handful of people do, but the office is still there whether 10, 100 or 1,000 people are using it. Plus, there’s a lot of capitalized equipment set up there, already paid for and waiting for people to use it. Moving out would be expensive (because of the lease break fees), and a hassle. So instead, the people in charge of a lot of companies probably rationalize that they might as well get some benefit out of those sunk costs by forcing people to use them again.
Software (you may or may not need it for HA): https://github.com/dresden-elektronik/deconz-rest-plugin