Browsed by
Author: matt

Computational Lithography

Computational Lithography

Optical lithography hit an interesting point in the 1990’s. Once circuit paths shrank to 180nm, we hit the physical limits of what the light wavelength and the optics in the lithography system could imprint on the silicon. You simply could not print silicon circuits any smaller. This was a huge problem for chip companies.

What some clever researchers did was come up with a way to bend these limits. Circuits are printed on silicon using a mask that creates shadows where circuit lines should be, and allows UV light to mark the areas that should be cut away. What they discovered is they would modify the circuit images by adding little ‘corrections’ to the circuit paths. These corrections would cause optical interference patterns that would actually generate the smaller features in the silicon. They started with simple ‘dog ears’ at the ends of circuits. As the features got smaller and smaller, the optical tricks they used got more and more complex.

This image below shows you an example. Each of these 5 lithography pictures generates the same circuit ‘plus sign’ on the left marked ‘no correction’. However, the ones to the right are able to do so at tremendously smaller circuit sizes. The image on the far right, ILT, can only be generated using complex mathematical models that work backwards from the desired final circuit image to the mask that must be put into the lithography machine to generate it. This, however, takes a lot of compute power.

Fast forward to today, and nVidia has released GPU optimized software to compute lithography patterns. Not only that, but they released the software called cuLitho. Using GPU’s, they claim it is 40x faster and uses 1/9 the power of traditional methods to calculate the computed lithography masks

Top 10 Airports in the US

Top 10 Airports in the US

How do domestic airports rank? This year, my Midwest home of Indianapolis gets some love as the #2 airport in the US.

  1. Savannah/Hilton Head International Airport
  2. Indianapolis International Airport
  3. Rhode Island T.F. Green International Airport
  4. Palm Beach International Airport
  5. Tampa International Airport
  6. Hartford Bradley International Airport 
  7. Minneapolis St. Paul
  8. Long Beach Airport
  9. Portland (Oregon) International Airport
  10. Detroit Metropolitan Airport

Sadly, Portland falls to #9 after having lead the list in years past, but things haven’t been very good at Portland airport in the last few years. We had protesters/counter-protesters assault each other a few times, and then of course a recent woman that fired a few shots at the airport.

360 VR experiences – black holes

360 VR experiences – black holes

ScienceClic English created this 360 experience of approaching and falling into a black hole. You can move your camera around and see in all directions. This is interesting because it’s a different take on some of the older ideas of what would happen if you fall into a black hole.

If you want to know what you’re seeing, here’s an explanation of what’s going on in this video and what the different visual phenomenon are caused by.

Reading 50 year old rope core memory

Reading 50 year old rope core memory

Mike Steward decided to recover the original Apollo guidance computer programs that landed man on the Moon in the 1960’s. Unfortunately some of them seem to have been lost to history.

It turns out, chunks of the original hardware still exist – such as the rope core memory which contained the programs. The next question is, how do you read these programs off 50 year old rope core memory hardware? This video below tells you how he did it!

He even wrote a web app that simulates how core memory works. We do a decent job recording history’s events, but I think it’s extremely cool that this kind of historical technical information is not being lost to the ages.

You can also check out the many other videos about the Apollo guidance computer in the other parts of his videos – or a previous article which has a super-awesome description of how Apollo computers work by Robert Wills.

Bored Ape NFT party severely injuries attendees

Bored Ape NFT party severely injuries attendees

A number of people have reported experiencing eye pain, vision problems, and sunburnt skin after attending ApeFest, a Bored Ape Yacht Club NFT collection event Nov 3-5th in Hong Kong. One person posted he woke up at 4am and could not see anymore. They rushed to the hospital where they are hoping to make a full recovery but were diagnosed with UV eye damage. It turns out, someone was almost certainly using full-spectrum UV-C (often called germicidal) lights instead of black lights.

The same thing happened at the Hypebest event in 2017 in Hong Kong. It’s not like this kind of harm is unique. Like the Jagermeister event that left a person in a coma, you better know your science.

BigClive on Youtube (who does amazing videos about extremely dangerous and counterfeit electronic devices you can buy and should be cautious of) recently uploaded a video and found the likely culprit. In the photo of the area with the toilets there were fluorescent tubes which are the characteristic teal-blue of mercury vapor discharge which emits quite a bit of UV-C and ozone as well.

How much radiation?

How much radiation?

I wrote a little while back about radiation from nuclear sources and how to detect them. But what about getting voluntary radiation for medical procedures: like X-rays, CT scans, and MRI’s?

Annual background radiation

Did you know you’re getting about 3.00 mSv of radiation every year if you live in the US? This breaks down to about 0.0082mSv per day.

There’s actually quite a bit of variation in the US depending on where you live. The types of rock in your area, position relative to the planet’s poles, elevation, and wide variety of other factors can affect your daily background radiation dosage.

Approximate effective radiation dose:Comparable to natural background radiation for:
One day’s background radiation0.0082 mSv1 days
A year of background radiation (US average)3.0 mSv1 year/365 days
Cross-country flight from New York to Los Angeles0.04 mSv4.87 days

X-rays

X-rays are something most people are familiar with. It turns out, however, modern x-rays actually give you a very low dosage. The lowest doses are dental X-rays. For a full panoramic dental x-ray, you’ll get about 0.007 mSv (0.7mrem). How little is that? One full day’s background radiation is about 0.0066 mSv (depending on where you live).

Still, there does appear to be a link to dental X-rays and certain types of thyroid and laryngeal, parotid gland, and salivary gland cancers.

Moving up, a full chest X-ray gives you a 0.1mSv dose. That’s about 10 days worth of radiation.

MRI’s

MRI’s use magnetic imaging, so they do not use radiation at all. They use a very powerful rotating magnet to generate their images.

Is it perfectly safe? Well, MRI’s can be done with contrast to help identify certain issues, and those chemicals can carry risk. Chuck Norris’ wife (yes, the Chuck Norris) even filed a $10 million lawsuit when his wife experienced health issues from the contrast used in an MRI.

CT scans

CT scans are particularly troublesome because they actually give you a pretty substantial dose of radiation. How much? Ex: A chest X-ray gives you the equivalent of 10 days of natural background radiation (0.1mSv). This is a very low dosage and highly unlikely to cause permanent or long-term damage.

On the other hand, a chest CT scan gives you 2.6 YEARS of radiation dosage – the equivalent of 77 chest x-rays. See some examples below or click them to see even more dosages based on different body part:

If that wasn’t bad enough, CT’s often also involve the use of contrast chemicals that may carry their own risks.

Beyond first class

Beyond first class

First class has seen some pretty big improvements the last few years, but that’s small peanuts these days.

The new A380’s offer suites – but even those with their full-sized recliners, full bed, in a private room doesn’t really touch what’s available.

Etihad Airways started offering something they’re calling The Residence on some of their highest end flights. Private airport entrances with concierge service, a private 3 room suite on the plane with living room, double bedroom, and full bathroom with shower. Add private large screen tv’s, cognac and turndown services along with private high end meals and you have a recipe for luxury. The price? $20,000 per ticket (vs $5000 for first class).

2023 Update: Unfortunately, it seems like The Residence were largely discontinued when Etihad retired most of their A380’s during covid. Nonstop Dan shows why they ran into problems selling them. Not because there weren’t customers, but because it turns out people that can afford that kind of luxury usually find it is barely any better deal than just renting a whole private jet. The price for a one-way private jet from Abu Dhabi to London also costs about $40,000 – or about the same as 2 Residence tickets. So if you’re flying with at least one other person, the private jet lets you take many more friends/business partners, have much more flexibility in schedules, and avoid big airports all together.

Bonus points for mentioning Abu Dhabi is a huge hub because it is located in a spot that makes most of the world’s population just 6 hours away – from China to Europe.

Chihuahua or muffin

Chihuahua or muffin

Free code camp compares various AI-based image recognizers to see how well they can identify if a picture is a chihuahua or a muffin. It’s surprisingly harder than you think and has a history of being used to determine the quality of the recognizer.

The author compares solutions from Amazon, Microsoft, IBM, Google, Cloudsight, and Clarifai. They also discuss the per-image cost as well as the quality of tags and other considerations. Definitely worth looking at if you’re trying to find an image classifier system.

Final results are on Topbots.

Links:

Photogrammetry/NeRF/Gaussian splatting compared

Photogrammetry/NeRF/Gaussian splatting compared


Matthew Brennan is not a computer scientist, but he takes 335 frames from a video and then processes them 3 different ways to compare the results. He creates a 3D mesh out of it for Photogrammetry, the processes it into a NeRF, and finally Gaussian Splatting.

What’s cool is that he shows how each works and how to process the data yourself. He also gives you access to the data to try it yourself.

Here’s the software he uses: