There is a polynomial expression that will match the file.
I wonder if expressing it would be compressible to a lower size than original file?
[edit] I’m wrong - not all sequences can be expressed with a polynomial.
This reminds me of a data compression scheme I came up with once:
Treat an n bit file as a polynomial over the finite field with characteristic 2. Now, there are some irreducible polynomials in this field, but many polynomials have factors of x and of (x+1). Factor the polynomial into P(x)x^n (x+1)^m. and just collect these terms, storing only P, n, and m.
That reminds me of a question I was asking myself: instead of using rare gas for ion thrusters… why not ejecting very high velocity electrons ?
Their mass is much, much smaller than any atom, but why wouldn’t that lead to a thrust ?
Possibly without limit except power - no limit on a “tank”?
I thought thrust in a vacuum required ejecting mass?
... the ejected propellant mass times its velocity is equal to the spacecraft mass times its change in velocity. [0]
Electrons have very little mass; do we eject lots more of them? Is there away to store that many electrons in a small enough volume? Or does their velocity make up for the lack fo mass? Also, if we must expend mass, I don't see how the 'no limit' idea works; I'm also suspicious of any claim of free, unlimited propellant, if that's what you mean.
There's a very good chance that is all my misunderstanding of something ...
Your [0] provides the "classical" equation for momentum transfer (p=mv), which is only reasonably accurate up to speeds 40-50% of the speed of light. But TFA talks about accelerating electrons to 80% of the speed of light. Then it's p=γmv where γ=1/sqrt(1-v^2/c^2)
Basically, it's taken as a "fact" that the relative speed of light is the same for two observers regardless of one observer's velocity vs. the other observer's. So if something is going 0.8x the speed of light, but light is still somehow observed to be traveling at the same speed for both (speed of light relative to both the fast-moving observer and the slow-moving observer), this apparent inconsistency is solved by realizing that ("magically") lengths are contracted for the faster observer. So the light appears to go the same distance to both observers, because distance itself is different to the two observers.
The net effect is that from 0.5x to 0.99999x the speed of light, momentum increases asymptotically as an object approaches the speed of light. Theoretically, if your spaceship could (truly magically) capture the momentum from ejecting one single electron to 0.9999....(58 nines in total)... a 1,000kg spaceship could achieve escape velocity to leave the entire solar system just from that one electron!
Ah! That “well” is exactly what I couldn’t understand from ion thrusters, where they shed electrons as they expulsé ionized gas.
All I could find was “to keep the same charge”, but since there is no absolute ground, what’s the problem?
The service experience with both Rivian and Tesla sucks. And that is why the manufacturer should never have a monopoly on the service of your car. Rivian wait times can be as long as five months. Tesla regularly denies warranty claims.
I can't speak for everyone, but ours was great. The other day the car threw an error message about the battery; we brought it to a service center, which swapped the whole battery free under warranty and without bother. Tesla gave my wife $100 in Uber credits per day until they got a loaner a few days later.
Perhaps the best part of the "service experience" is not having to go in for oil changes, brakes, and all the other stuff legacy cars require.
Not my experience with Tesla servicing. First off, you don’t have to service the car besides tire rotation and air filter replacement once a year or so.
When I do walk in to get something checked out, I get an estimate via the app and can chat with someone before even dropping off the car, which is nice. Tesla servicing has been a good experience for me so far, and does make me want to order another Tesla when our ICE car dies a couple years from now.
I’m interested in details.
Nuclear is very expensive and not going down, while batteries are.
So with your analysis, it could be possible to estimate how low batteries have to go before meeting nuclear… It could be irrealistic but let’s see !
On the nuclear side I was looking at a GE BWRX-300 (SMR) because my province is currently looking at building a few of them.
Specs: 300MWe, nominally $1B first-of-a-kind build cost and $675M next-of-a-kind build cost. Runtime between fuel changes is 18-24 months.
--
On the battery side I was looking at the Tesla Megapack 2 XL, since our neighbouring province has a bit of capacity installed.
Specs: 979kW output per pack, 3.9MWh capacity. For 300MW output capacity we need 306 units => $425M. The total capacity from fully-charged to fully-discharged is 1193MWh. Total time from full-charge to full-discharge: 3.9h.
In the middle of winter we have 8h of daylight and 16h of night/twilight. To provide 300MW overnight on a calm day we need 4.02x the storage capacity (16h runtime/3.9h discharge). That gives us 1231 units for a total cost of $1.7B.
--
Yes, the nuclear plant will be more expensive to run (trained staff, security, disposal, etc). On the other hand, the $1.7B cost doesn't include any of the devices that would actually be charging the battery packs either.
That’s interesting.
The quotes for the nuclear power plant are really low though.
France, which has a lot of knowledge and trained workforce for nuclear, is slowly building a reactor of 1.2gwe for like 20B euros.
If you use those numbers… it looks very different.
And that’s at least not idealistic price since it’s they are completing the construction.
Yeah, I'm really curious to see how the SMR thing shakes out over time. It does make sense to me that the costs between large bespoke reactors and smaller modular reactors would not necessarily be a linear scaling by MWe. Another factor that affects the price significantly is whether or not the reactor is being built at an existing site that is already licensed or whether it's a scratch-build at a new site.
https://english.hani.co.kr/arti/english_edition/e_national/8... talks about the Shin Kori 3 and 4 reactors (brought online in 2016 and 2019) costing $6.4B USD for the pair (after a 32% cost overrun), if I'm reading correctly. Those ones are 1416MW and 1418MW, or $2258/kW which is right in line with GE's estimate of a next-of-a-kind SMR build (spec sheet says $2250/kW)
That’s 2M per MW. We can spend 10, an over capacity of 5x !
And that is not accounting for the price of money, it takes very long time to build a nuclear power plant, and you will have your PV plant in the year, probably a couple for mega pack due to demand.
I’m assuming you’re looking at the Flamanville Unit 3? Reading through the history of that build is pretty bad for sure. I’m maybe jaded enough now to expect construction projects to go over budget by maybe 20-30% but lol 580% over budget is not normal at all.
There is a fundamental reason for low inflation / growth in Japan: demography.
No children, very aging population.
I think normalizing growth by population growth would give some interesting insights about "real" growth, or lack thereof.
Perhaps, but the above is not a point about Japanese growth, but rather competitiveness. I'm not sure that population decline connects directly with Japanese hotels, razors, and toothpaste becoming rapidly more globally competitive, even if it's just Japan treading water while the rest of the world gets more expensive. Usually you'd expect it from a country with a strong "demographic dividend" of a young working population. An increase in competitiveness can apparently happen with or without population growth, with or without inflation, and with or without increasing wealth for Japanese families, and with or without inflation.
For example, Japan was also getting more competitive in the 1970s-1980s, but at that time it happened alongside a population boom, young workforce, economic growth, inflation, and a strengthening Japanese consumer. This time, it's more like Japanese companies getting internationally competitive while their economy declines and their workers accept lower wages and lower standards of living, with high rates of industrial production amidst shrinking domestic demand. Again, not so good for Japanese families perhaps, but that's not what competitiveness is necessarily about.
I pretty much only work with Xilinx/AMD parts these days, and, at least on US/US+, they do indeed all have a CE on all the flops (unless there are undocumented degenerate cells with missing features).
There is a design style called RTL (register transfer level) which works better with registers that have separate clock and enable inputs. The clock goes to all registers, but on any given cycle only a subset of those are enabled depending on the current state. Since these enables are generated from the outputs of these registers which then pass through some combinational logic, they will be stable around the rising edge of the clock.
Note that people talk about Verilog or VHDL as "RTL languages". That is not correct - these languages allow hardware designs using any style to be modeled. But sticking with RTL is strongly encouraged. A design that uses the output from one part of a circuit as a clock to another part, for example, is not RTL. This might cause problems when moving from one FPGA to another or to an ASIC.
I have almost never seen hardware debouncing. It's just way better done in software. Unless you have extremely specialized switched which require HW debouncing(???), please just do it in software.
