sephamorr

Note the last sentence of your quote. Device-to-device variability and production yield are opposites. I can't imagine how techniques that exploit variability between devices (with no mention of temperature, lifetime drift, etc) could be reliable.

A normal worldwide USB power supply accepts 86-240v AC, so you can usually supply a DC voltage on the same terminals of the same voltage with no issue (convert from AC peak to peak to DC)

Plenty of sovereign debt interest rates are below zero, so you'd pay $101 for $100 in a year's time. Whether or not this is sane economic policy is a question though.

The scientific payloads of missions (development, design, manufacturing, operation, analysis) are usually a large percentage of the total mission's cost. It's worth looking at the exact scientific capabilities of a mission while comparing it with the cost. It's amazing how economically the Indians sent a probe to Mars, but its scientific capabilities were far smaller than the billion dollar NASA missions.

Wet Tantalum caps have terrible vibration/shock sensitivities. I would use ceramic caps in aerospace 100 times out of 100 rather than consider wet tantalums. At least ceramics are repeatably manufactured and can be mounted J-lead to reduce shock loads. The tants will just randomly break open on you. Not sure what they're currently used in, but I wouldn't consider them for new applications.

I just looked through Jackie Chan's movies for the past few years, and they're almost exclusively produced by mainland Chinese film companies. I understand he's a Hong Kong national, but it seems that he is toeing the line for the movie industry he's working for. Considering how the Chinese have cracked down on Cathay Pacific (forcing them to fire employees who have participated in the protests), I doubt a mainland movie star could get away with saying anything else.

If you'd like to read a well researched book on the subversive goals of the CCP, I suggest you read "Silent Invasion"[1]. Note that there some critical reviews of this book, suggesting it's connecting dots where there isn't true motive, but little criticism of the evidence presented. I encourage you to read it and make up your own mind and draw your own narrative around the facts.

[1]https://www.amazon.com/Silent-Invasion-Chinas-Influence-Aust...

Note that most of RAM's power consumption comes from reading and writing (a function of memory clock speed). Allocating the RAM itself is likely to meaningfully increase power draw, but accessing it will.

Regenerative braking certainly is reduced by EV's; a Tesla may never replace its brakes during its lifetime. Reading the paper, the dominant source of PM2.5/PM10 is resuspension of dust already on the road, rather than tires or brakes.

Ah yes, this is partially true: in national courts in China, foreign companies often are successfully able to defend their IP, only to find that enforcement relies on separate local jurisdictions which have little interest in enforcing rules against their own.

Whatever you do, do NOT do this with lithium ion batteries. At low voltages, chemical side-reactions start to occur which result in dendrite formation which can pierce the membrane, shorting the battery: the result is fire. LiIon batteries generally have protection that permanently blow a fuse if the voltage goes below a minimum safe voltage, but raw 18650 cells often do not. The video above refers to NiMH cells. Afaik, this is safe to do for NiMH.

Fortunately, a nuclear reactor isn't radioactive until the reactor is turned on. Enriched uranium on its own is hardly radioactive at all (U235 t1/2=~700Myears); depleted uranium has meaningful safe non-nuclear applications[0]. There are orders of magnitude more radioactivity in the RTGs already used on deep space missions. Those are very carefully protected, but the risks are deemed acceptable. So long as you don't turn your reactor on until you're well away from earth, I see very limited risk of exposure to the public, far less than from a RTG.

[0]https://en.wikipedia.org/wiki/Depleted_uranium#Civilian_appl...

This NASA document (pages 62 onwards)[0] (1974) here gives a more in-depth description of NTO materials compatibility than the link above (1970)

Titanium is practically the only metal that can store NTO for long periods of time over a reasonably large temperature range. This does come at the expense of shock sensitivity (particularly after a soak/adsorb), but the industry has to use it anyway, while working to mitigate the shock problem.

[0] https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/197200...

This appears to just be a 60v->12v isolated converter at ~2kW. I've never seen a push-pull topology used for this power level since it makes such inefficient use of the transformer and FETs (note it uses 4 parallel FETs for each branch of the primary). Looks fun though

For the utilities, absolutely. The utilities have hard requirements of delivering power to all customers while maintaining grid and frequency stability, so the ability to shape demand and supply quickly is at least one very valuable knob.

I'm confused why there is such a dissonance on the definition of "autopilot." Most people are fully aware that commercial planes have at least one human pilot in addition to an "autopilot," and I think the passengers on the plane would be very unwilling to fly if the human pilots weren't present. "chauffeur" seems like it would be misleading for partial self-driving software, but "autopilot" seems spot on: helpful, but not good enough to replace the human.

I'd like to read more about arms control negotiation, do you have a book suggestion?

The power dissipation problem of current chips (performance is almost always power/thermal limited) is correlated with 2D structures. Communication is one giant problem: 3D structures could allow for far better colocation of connected compute elements, so we might not have to spend most of our power budget feeding data in and out of compute clusters. Scaling in 2D does make this problem easier (in some ways!), much as adding a 3rd dimension would too.

Having seen the trash coming down several of these rivers in person, everything I could identify seemed to be of local origin. I certainly didn't see non-local languages written on plastic bottles, for example.

RTL-SDR refers to repurposed USB TV tuners using RTL2832U chipsets. They do not have any transmit ability.

There's probably plenty of bias, but a large part of the China2025 plan is to achieve dominance in semiconductors. Since the domestic chip industry needs to catch up and the intelligence services are not above IP theft, this isn't surprising to me.

There was a project that did exactly this, I'll see if I can find the link, most of the work was spent stripping HDCP. The key heuristic in ad detection is that (at least in 2014) there is usually a single black frame between content and advertisements, and ads are rarely transmitted with more that stereo audio, while most content comes as 5.1 or 7.1 channels, even if it is a naive upscale.

Several lines of micro controllers include primitive "GPUs" to give hardware acceleration to the operations usually required for LCD interfaces. The capabilities aren't necessarily 3D rasterizers, but do make it a lot easier to draw interfaces, interface cameras, etc. Take a look at the graphics cores on some of the higher end STM32 parts. Certainly sprite drawing would be quite faster, not sure how much else could be accelerated with this type of core.

While there is certainly some truth to your statement, the subsidies being referred to in the article are on installed panels. Prices on panels were held high by subsidies, and their removal resulted in a market supply glut. This 10-15% raise is just the correction from the previous ~30% fall.

Something that's usually missed about nuclear waste is that the most radioactive wastes have the shortest half lives so burn themselves out quickly. The waste that lasts for thousands of years is far less radioactive. For example, in a uranium mine, you don't even care about the dose you get from uranium (~4 billion year half life), but you do care about the fission products with far shorter half lives. Your point stands though.

Let me shed some light on how you can drive LEDs from an AC power source. Naturally the worst techniques are usually the cheapest. To add to the complexity, directives like EnergyStar are aiming for power factor >0.9 in lighting above 3W. Maintaining high power factor (sin^2 input power) while having low lighting ripple (constant output power) requires some method of of storing energy. In methods 1,2,3 (and 4 depending), neither high power factor nor good lighting ripple is achieved, but very low cost is.

Note that this info is for fairly low power LEDs. When you start having high power LEDs, regulations may require better power electronics, and the cost of the electronics might be amortized against the cost of the thermal cooling solution, etc, so you can make a more expensive electronics package.

1) By stringing ~ 180v of LEDs in series, and adding a current limiting resistor, you can drive the whole thing from the AC line. This has terrible variability (short light spikes every 60hz, and the thermal effects will change the LED properties so much that you could easily have 2x light intensity change as they warm up) and poor total illumination (as most of the cycle is spent with the AC waveform lower than the needed voltage. I haven't seen this in the wild, as option 2 isn't much more expensive and far better.

2) Use a single diode + capacitor rectifier to generate ~180v, and then have a similar string as in (1). This gives far better illumination (the capacitor will probably stay in the 160-180v range), but still has substantial 50/60hz ripple based on the size of the capacitor

3) Similar to (2) but use a full bridge rectifier instead of a single diode. As we are now driven by a rectified sine wave, the ripple voltage will ~half, and the ripple frequency will now be 100/120hz.

4) Many applications don't have ~60 LEDs in series, so you can't drive directly from the rectified AC line. The first stage is a rectifier, and a second stage most likely is a buck converter. The quality of the light out of this depends on how much you care about power factor. Chips like NCL30288, for example, maintain high power factor and fairly low lighting ripple. Most people won't notice this.

5) Particularly in higher power lighting, a separate power factor correcting stage achieves high power factor, and a second stage generates a constant current that should have exceptionally low luminous intensity ripple (if designed right). The ripple frequency should be in the hundreds of kHz as well, so I don't think any eye could notice this.

And don't get started on dimming, I haven't found a product on the market that does non-phase cut dimming in <20W lights.

I watched it with some friends (all of us SpaceX engineers), and our universal comment on the way out was "so about those crazy [vibration] levels..."

A pencil point is a tapering edge, so I'm assuming the author is considering <1mm for "fully capable computer the size of a pencil point" being unreasonable.

Microcontrollers are not a huge amount larger than that any more:

-Cortex-M4f in 1.6x1.6x0.65mm (MAX32660)

-Attiny20UUR in 1.56x1.4mmx0.50mm

Bare dies may be a bit smaller than the above.

You're spot on here. As I understand, the Russian state was quite surprised at the capabilities of Western detection science. It was certainly not common knowledge that low-concentration alpha emitters (still plenty poisonous) could be detected in blood, nor that the assassins could be tracked all the way back to Russia just from their limited exposure.

The 3g/4g analogy works here too. LTE has categories (0 to 19, with some extras) that increase bandwidth (or another metric) from Rel8 to Rel13.