.03ns is a frequency of 33 GHz. The chip doesn't actually clock that fast. What I think you're seeing is the front end detecting the idiom and directing the renamer to zero that register and just remove that instruction from the stream hitting the execution resources.
For a 32 bit number you're looking at going from using 256 to ~1800 transistors in the operation itself. A modern core will have roughly 1,000,000,000 transistors. Some of those are for vector operations that aren't involved in a xor or sub, but most of them are for allowing the core to extract more parallelism from the instruction stream. It's really just a dust mote compared to the power reduction you could get by, e.g., targeting a 10 MHz lower clock rate.
There's a structure called a carry-bypass adder[1] that lets you add two numbers in O(√n) time for only O(n) gates. That or a similar structure is what modern CPUs use and they allow you two add two numbers in a single clock cycle which is all you care about from a software perspective.
There are also tree adders which add in O(log(n)) time but use O(n^2) gates if you really need the speed, but AFAIK nobody actually does need to.
That's all true, but on any modern x86 processor both the single pair of gates for the xor and the 10 or so for a carry-bypass 64 bit wide subtraction both happen with a single clock cycle of latency so from a programmer's perspective they're the same in that sense. There's still an energy difference but its tiny compared to what even the register file and bypass network for the operation use, let along the OoO structures.
The question isn't whether they both take a clock cycle, but rather whether any future implementation of the ISA might ostensibly find some sort of performance advantage, even if none do right now. From that standpoint, xor seems like a safer bet.
There's been a lot of churn over the years but additions being done in the same timeframe as XORs has been pretty constant. The Pentium 4 double pumped its ALU but both XORs and ADDs could happen in a half cycle latency. The POWER 6 cut the FO4s of latency in stage from 16 to 10 and kept that parity as well. When you need 2 FO4s for latching between stages and 2 to handle clock jitter at high frequencies the difference between what a XOR needs and what an ADD need start looking smaller, particularly when you include the circuitry to move the data and select the instruction. Maybe if we move to asynchronous circuits?
The blog post is about why this is idiomatic not whether it needs to be done that way today. It’s idiomatic because once upon a time none of that existed and xor gates did. The author apparently never took intro to digital logic.
There's only a few grams of hydrogen in the reactor's plasma, it's reaction with oxygen wouldn't be much more exciting than just losing containment. There are engineering challenges that have to be addressed but no worse than the 6 MW research reactor I used to walk by every day to my college classes in the middle of a dense city.
The proliferation risk of someone using the neutron flux to produce an atomic or dirty bomb are real but that exists no matter where it is.
I think the proliferation risks will be in future the reason, independent of technological obstacles or costs, why US will not allow to build fusion power plants in all countries around the world.
Hybrid nuclear fusion–fission power plants have been already proposed and studied in theory.
"In general terms, the hybrid is very similar in concept to the fast breeder reactor, which uses a compact high-energy fission core in place of the hybrid's fusion core. Another similar concept is the accelerator-driven subcritical reactor, which uses a particle accelerator to provide the neutrons instead of nuclear reactions."
> Hybrid nuclear fusion–fission power plants have been already proposed and studied in theory.
I have a hand-wavy hard sci-fi universe I've been rolling around my head for years and I eventually came to the conclusion that fission-fusion drives would be really handy for spacecraft, since it would be much easier to start a fission reaction in a cold/dark ship than fusion because of the power requirements. Otherwise you need some other way to generate 10s or 100s of MW to start the fusion reaction.
Using the energy of fission for spacecraft propulsion has been studied and some prototypes have been already constructed. Most of these nuclear engines should be used only outside of Earths atmosphere.
In the text of the treaty there are promises to NPT signatories, such as:
Article IV
"1. Nothing in this Treaty shall be interpreted as affecting the inalienable right of all the Parties to the Treaty to develop research, production and use of nuclear energy for peaceful purposes without discrimination and in conformity with articles I and II of this Treaty."
"The world’s leading nuclear exporters should ensure that states have reliable access at reasonable cost to fuel for civilian reactors, so long as those states renounce enrichment and reprocessing."
"The 40 nations of the Nuclear Suppliers Group should refuse to sell enrichment and reprocessing equipment and technologies to any state that does not already possess full-scale, functioning enrichment and reprocessing plants."
For example when United Arab Emirates wanted to build the Barakah nuclear power plant, (supplied by Korea Electric Power Corporation, not by an U.S. company), it had to sign an the Section 123 Agreement with United States of America. As part of the agreement, the UAE committed to forgo domestic uranium enrichment and reprocessing of spent fuel.
To be fair, it's not only U.S. who want's the control access to nuclear technology and nuclear materials. For example India wanted to become a member of Nuclear Suppliers Group for a long time. As of 2019, China has thwarted every attempt of India's inclusion into NSG and has made it clear that status quo will remain citing "lack of consensus" among NSG members.
Another example is South Korea. South Korea is constrained in its nuclear power policy by the 1974 Korea-US Atomic Energy Agreement. Only in November 2025 did the USA formally affirmed support for South Korea’s civil uranium enrichment and spent fuel reprocessing for peaceful uses.
Good points, but in all of that they are specifically targeting enrichment and reprocessing, while allowing nuclear power plants.
A fusion reactor is a power plant. It produces neutrons but so does fission; in fact, conventional fission plants get a third of their power from plutonium that they breed from U238, and plutonium accounts for most of the long-term radioactivity in the waste.
By comparison, a fusion plant would have no uranium present for any legitimate reason, and assuming it's D-T it would need those neutrons to breed tritium. Tritium has a use in thermonuclear weapons but not without highly enriched fissiles, and tritium is also a byproduct of fission plants.
Precisely because fusion reactor could be used for plutonium breeding they will be subject to the same controls and political pressures as fission power plants.
Operation of all fission power plants, outside of countries with nuclear weapons, is subject to very strict controls by International Atomic Energy Agency, which reports to United Nations Security Council. Also because each civilian nuclear power reactor breeds enough plutonium to produce multiple nuclear weapons each year, spent nuclear fuel is subject to the same very strict controls. (I know the the isotopic composition of this plutonium is miserable for weapon production, but nuclear weapons using reactor-grade plutonium have been build and tested).
I don't know any thermonuclear weapon using highly enriched uranium. Highly enriched uranium is used only in few designs of nuclear weapons: nuclear weapons which you build when you don't have access to weapon grade plutonium, or nuclear weapons which have to survive strong external shocks - nuclear bunker busters. Plutonium Pu-239 based nuclear weapons are physically smaller and lighter, easier to place into missile nose cone.
For the most part I think we're in vehement agreement. I wrote above that nuclear plants are permitted with inspection under the NPT, and I would expect D-T fusion to be the same. Your links also say that reactors are allowed in practice.
When I said "highly enriched fissile" I meant that to include plutonium. Weapons generally use Pu239, while nuclear waste plutonium is 58% Pu239 mixed with other isotopes, plus of course a lot of U238 and various transuranics.
There is one area where fusion reactors are clearly being treated differently, and that's by regulators such as the NRC, which has already decided to treat fusion reactors like particle accelerators and medical devices, rather than with the much more difficult process they use for fission reactors.
Fission-fusion or accelerator-driven fission is pure BS. It combines the disadvantages of _both_ and none of the advantages.
Modern fission power plants are designed with a reactor vessel to last a century and to withstand high pressures and temperatures. It's built and emplaced permanently in a large concrete shielding structure.
In a hybrid design this just won't work. Fuel will need to be right next to a high-vacuum chamber that will need periodic maintenance.
With NVidia/OpenAI actual graphics cards did change hands. Vendor financing, like when a car dealership gives you a loan to buy a new car, is actually pretty normal.
reply