> It's better to use a constant time algorithm, but that's harder to do in a curve generic way and has a pretty significant performance impact (particular before the safegcd paper).
Crypto noob here, but isn't modular inverse the same as modular exponentiation through Fermat's little theorem? I.e., x^-1 mod n is the same as computing x^{n - 2} mod n which we know how to do in a constant-time way with a Montgomery ladder.
Or is that too slow?
I personally did not know about classes prépa before the last year of high school. I will forever be thankful for my maths teacher who told me about it that year, since I would have probably slacked off at university.
- Make them good at science, i.e., Maths and Physics.
- Get them into a decent high school, e.g., Henri 4 or Louis Le Grand in Paris.
- Hope they have good grades and manage to get into a good preparatory class [1], e.g., Henri 4, Louis Le Grand in Paris, or Hoche and Sainte-Geneniève in Versailles.
- Make sure they don't slack off, and hope they get into a good engineer school, e.g., Ecole Polytechnique, Ecole des Mines, Ecole Nationale des Ponts et Chaussés, CentraleSupelec.
(Lists are not exhaustive)
If they manage to get into one of these schools, they will most likely end up not have any difficulty to find a somewhat well-paid job in France.
> Google put in significant engineering effort into "Ryu", a parsing library for double-precision floating point numbers: https://github.com/ulfjack/ryu
> In other words, CHERI can reduce memory-unsafety related slowdowns by two thirds, which is pretty incredible. AMD CPUs are even starting to have hardware support for it, bringing its run-time overhead down to 6.8%.
Source on AMD CPUs having support for CHERI-style capabilities ? Afaik, there is only the Arm Morello prototype out right now and FPGAs.
You might be interested in [Catala](https://catala-lang.org/en/), a domain specific language for implementing legislative texts. Iirc, they have worked on the french tax code, and are starting on the us one now.
Would virtual memory still be necessary on a capability machine architecture [0] ? My understanding is that would not be the case since all programs can only access memory that's been allocated to it. By definition, capabilities cannot be created out of thin air like pointers can be ?
> ... leakage by timing side-channels depends in parts on how accurate your time-measurements are (e.g. Javascript's timer resolution was degraded, in order to make transient failure attacks like Spectre harder [1]).
But that doesn't matter if how long it takes for your instructions to execute is data independent, no ?
So what kind of formal verification is it ? Is it proof assistant, model checking ?
And what does it verify ? It's not really clear from a first glance.
Crypto noob here, but isn't modular inverse the same as modular exponentiation through Fermat's little theorem? I.e., x^-1 mod n is the same as computing x^{n - 2} mod n which we know how to do in a constant-time way with a Montgomery ladder. Or is that too slow?