Jessy Pensédent

Founder & Software Engineer @

[Jessy Pensédent @gamesofarcadia]

We are thrilled to announce the filing of a patent that solves one of chemistry's longest-standing industrial problems: room-temperature nitrogen fixation.

The Haber-Bosch process has fed half the world's population for over a century, at the cost of 1-2% of global energy consumption and significant CO₂ emissions. The reason it requires 400°C and 200 atm has never been quantitatively explained, until now.

Our method derives the exact activation barrier from first principles, and provides the complete design criteria for catalysts and electrochemical cells that operate at 25°C and atmospheric pressure.
The same principle extends to CO activation, oxygen reduction for fuel cells, and several other high-value industrial processes.

This solution comes from the current development of our molecular electronic structure calculator. Solving FeMoco was our main target to benchmark our tool and show that it was possible to do it analytically on a laptop. Unlocking room-temperature nitrogen fixation was not on our roadmap, but we love side quests!

A paper about all this will be published as soon as our tool is ready. It will cover: ORBITAL (our calculator), FeMoco as one of many benchmarks, and Nitrogen fixation as one of many results.

In the meantime, for the scientists who have been losing sleep over it: The obligate H₂ production in nitrogenase is not wasteful. The two electrons and protons that become H₂ are committed before N₂ even arrives. When N₂ binds, the only channel available to activate it is already full.
H₂ is the door opening, not energy being thrown away.
--------------
#NitrogenFixation #GreenAmmonia #HaberBosch #CleanEnergy #DeepTech

[Jessy Pensédent @gamesofarcadia]

𝐂𝐈𝐐𝐒 and 𝐂𝐈𝐐𝐀 ran a black hole on quantum hardware:
https://t.co/yIMTsDZPSc

𝐂𝐀𝐒𝐈𝐌𝐈𝐑 is now indexed by @inspirehep :
https://t.co/UP7d0kYhE1

𝐊𝐀𝐆𝐄 introduces a new type of information security:
https://t.co/HfSkViyxto

𝐀𝐥𝐥 𝐭𝐨𝐨𝐥𝐬 𝐚𝐧𝐝 𝐩𝐮𝐛𝐥𝐢𝐬𝐡𝐞𝐝 𝐩𝐚𝐩𝐞𝐫𝐬:
https://t.co/YvQOqR233U

--------------------------------------
More about the next releases:

𝐎𝐑𝐁𝐈𝐓𝐀𝐋 has already been validated on the first-tier benchmarks (H, He, H₂⁺, H₂). Ultimately, it will compute the electronic structure of any molecule, including the infamous FeMoco. No quantum computers or supercomputers required. (80% complete)

𝐒𝐓𝐑𝐀𝐓𝐎𝐒 is mostly complete. We're now validating the unified atmosphere against all cases (eccentric orbits, reentry, collision prediction, maneuvers/thrust, multi-body gravity, non-spherical drag...)

Those two are our biggest and most complicated projects to date. Surprisingly, the difficulty is not in the derivations required, but in the overall coherence. They handle far more than any tool in their category, while delivering exact results.

[Jessy Pensédent @gamesofarcadia]

For years, physics has produced increasingly specialized models, increasingly isolated fields, and increasingly fragmented tooling. What if a single foundational framework could generate efficient systems across completely different domains? Not simulations of a vision, or concepts, but actual production ready, fully analytic, and parameter-free systems producing results never seen before.

About two months ago we started to answer that question.
The result today, six shipped tools, and more to come.

People have claimed “theories of everything” for decades.
We make no such claim, we just keep shipping the receipts.

---------------------------------
Our tools are freely accessible.
No account required, no ads.
---------------------------------
#Physics #QCD #QuantumComputing #DeepTech
---------------------------------

[Jessy Pensédent @gamesofarcadia]

CASIMIR published paper:
https://t.co/2k8bemibgy

CASIMIR the online tool:
https://t.co/MHVbUpypOk
(no account required, no ads)

CASIMIR on INSPIRE-HEP:
https://t.co/JWKUFy292Z

[Jessy Pensédent @gamesofarcadia]

Introducing CASIMIR, the Quantum Chromodynamics solution in our analytic physics suite. Instant, no lattice simulations, no Monte Carlo.
It delivers:
• Phase boundary Tc(μB)
• Phase boundary curvature derivatives χ̂₂(μB), χ̂₄(μB)
• Saddle node terminus analysis
• Glueball mass spectrum
• Dense matter onset boundary
• Collider beam energy (μB, Tch, Tc) conversion
• CSV/PNG export

Key results:
• κ₂ and κ₄ match Borsanyi+ 2020 fitted values within the quoted lattice uncertainties
• Reconstructed Tc(μB) matches Borsanyi+ 2020 direct Tc points to within 0.15σ at μB = 100, 200 MeV
• Phase boundary terminates at μB* = 421.88 MeV, T* = 113.58 MeV in a saddle node bifurcation
• No critical endpoint: χ̂₂ and χ̂₄ remain finite at the terminus
• Falsifiable prediction: κ₂ₙ = 0 exactly for all n ≥ 4

This tool is designed for heavy-ion physicists, lattice QCD researchers, and anyone working on the QCD phase structure or glueballs.
We welcome your feedback.
-------------------------------------
CASIMIR has been indexed by @inspirehep, the canonical database for high energy physics maintained by @CERN, @desy, @Fermilab and @SLAClab, and will be included in the next Particle Data Group update.
-------------------------------------
All link in the comments below
-------------------------------------
#QCD #ParticlePhysics #PhaseDiagram #DeepTech

[Jessy Pensédent @gamesofarcadia]

More about Matter Vs. Photon qubits:
https://t.co/PLycm7s4p1

More on CIQA, the 1:5 analytic QEC:
https://t.co/Vnx90zBYJG

More on the million-qubit compilation pipeline:
https://t.co/38vNGduHYZ
--------------------------
https://t.co/cNfCL5A8xA

[Jessy Pensédent @gamesofarcadia]

https://t.co/dTxycfk3jS

[Jessy Pensédent @gamesofarcadia]

More on the Page Curve Experiment:
https://t.co/3a7giHvHTw

More on CIQA, the Analytic 1:5 Quantum Error Correction Code:
https://t.co/9gF76d35pp

Use CIQA freely. No account required, no ads:
https://t.co/wcnxXjmCTa

Any feedback is welcome ☀️

[Jessy Pensédent @gamesofarcadia]

A month ago we successfully ran the full Hayden-Preskill black hole circuit on quantum hardware. The goal was to validate CIQA, our analytic 1:5 error correction code.
The idea was simple: either the circuit doesn't run, and we go back to work, or it does, and we can present the most advanced quantum error correction code currently available.

Well, the circuit ran, and with it came unexpected results. We now had to present a new architecture for error correction, but also data that challenges the information recovery and the Page curve predictions within that model. The Hayden-Preskill paper angle was an afterthought, and we missed the opportunity to properly address it. We could have done better, so we did.

We published a new version of that experiment. The results are still the same, but we added new runs (including larger circuits), rounded some angles, analyzed the data further, and attached all the raw data from both the IBM side and our side in one convenient folder.

All the links in the comment below.
----------------------------------------------
cc: @preskill
----------------------------------------------
#QuantumComputing #PageCurve #HaydenPreskill #QuantumErrorCorrection
----------------------------------------------

[Jessy Pensédent @gamesofarcadia]

CASCADE is free to use, no account required, no ads:
https://t.co/Obsq15vSO2

Published paper:
https://t.co/fHCQ0Qkuuj

Any feedback is welcome ☀️

[Jessy Pensédent @gamesofarcadia]

Introducing CASCADE, the Turbulence solution in our analytic physics suite. Give it three numbers (turbulent velocity u', length scale L, viscosity v) and it instantly returns:

• The complete energy spectrum E(k)
• Structure functions S2 and S3, including the exact 4/5 law
• All cascade scales (Kolmogorov length, injection and dissipation wavenumbers)
• Reynolds numbers Re and Re_lambda

As with all our tools, CASCADE requires no simulations, no tuning, and no free parameters.

This tool is designed for CFD Engineers, Aerodynamicists, Turbulence Researchers, Aerospace Engineers, and Wind Energy teams. We welcome your feedback ☀️

CASCADE is freely accessible online. All the links can be found in the comments below.

cc: @NASA @SpaceX @Boeing @CFDdirect @NonEquilibTurb

#Turbulence #FluidDynamics #Aerodynamics #FluidMechanics #DeepTech

[Jessy Pensédent @gamesofarcadia]

More about those results: https://t.co/00LjVYCJIf

CIQS is freely available on GitHub for non-commercial use: https://t.co/bQ682jYJeb

Happy ranking Sunday everyone ☀️
https://t.co/cNfCL5A8xA

[Jessy Pensédent @gamesofarcadia]

"What about ranking Sunday?"... When the shower thought wins, here it is 😊
If anybody is wondering about those results, yes, this is a radical difference in architecture. While everybody is investing more effort into trying to brute-force their way to scale, we decided to look at how qubits function... 👀

What we found is this:

1. There's an actual physics threshold that can be used in quantum computing.

2. The quantum field is wasting engineering power.
(seriously guys, we don't want to make you look bad, but quantum computing is supposed to be this efficient... What about a rematch? 😉)

cc: @IBM @qiskit @QuantinuumQC @BerkeleyLab

#QuantumComputing #Benchmark #DeepTech

[Jessy Pensédent @gamesofarcadia]

DYNAMO is free to use, no account required, no ads:
https://t.co/e99J3JieRx

Published paper:
https://t.co/4hs4IgUBl6

Any feedback is welcome ☀️

[Jessy Pensédent @gamesofarcadia]

Introducing DYNAMO, the newest addition to our analytic physics suite. This tool takes 8 inputs and directly computes the complete set of MHD properties without any approximations. DYNAMO has been validated against 16 benchmarks from 3 separate references:

- Solar wind data at 1 AU
- Solar corona reference values
- ITER design parameters

As with all our tools, DYNAMO requires no simulations, no tuning, and no free parameters.

This tool is designed for plasma physicists, MHD specialists, fusion engineers, and space weather scientists. We welcome your feedback ☀️

DYNAMO is freely accessible online. All the links can be found in the comments below.

----------------------------------
cc: @NASA, @esa, @ITEROrgani13436, @CEA_Officiel, @Airbus

#PlasmaPhysics #FusionEnergy #Astrophysics #DeepTech

[Jessy Pensédent @gamesofarcadia]

Every quantum computer* built today is fighting the same battle: the hardware is noisy, every gate introduces errors, and to run anything useful you need to protect logical information across many physical qubits. The current standard, the surface code, asks for around 1000 physical qubits per logical qubit to reach practical error rates. Other solutions achieve a more affordable ratio, but at the cost of efficiency, and are not reliable for any meaningful production. That's why "useful quantum computing" keeps being 10 years away...

The overhead isn't a hardware problem. It's a code problem. Surface codes were designed without analytic visibility into where the geometric protection actually comes from, so they brute-force redundancy. An analytic encoding derived from the underlying structure would change that: massive reduction in physical qubits needed, exact correction, no calibration.

CIQA is a free tool that does exactly this.

It encodes one logical qubit into 5 physical qubits with full single-qubit Pauli error correction. This is the geometric floor for error correction. The literal minimum physics needs to correct a qubit on all axes. That's also 200 times less overhead than surface codes at equivalent logical error rates. Encoding angles, syndrome thresholds, and correction operators are all derived analytically. And as always with our solutions: no fitting, no tuning, no free parameters.

CIQA was validated on real quantum hardware in two ways:

• Benchmark: Three independent runs on @IBM Heron r2, 8192 shots each. Mean corrected fidelity up to 0.847 versus an unprotected baseline of 0.015.

• Experiment: CIQA provided the error correction in the first full Hayden-Preskill black hole circuit executed past the Page time.

An online version of CIQA is available. No account required, no ads.
Supports @qiskit, OpenQASM 2.0, and JSON gate list input.
4 operations available:
• Encode
• Syndrome
• Correct
• Concatenate*

Quantum hardware engineers, error correction researchers, and anyone tired of waiting for million-qubit machines: this tool is for you.
Feedback is welcome.

Use CIQA:
https://t.co/SjgcYB1cvp

Read more about CIQA:
https://t.co/Bc83965cRw
--------------------------------------
*This is true only for matter qubit architectures. We show in the following paper that photon qubits do not have a decoherence channel:
https://t.co/kACjgmcyGx

*Concatenation scales cleanly: 1:5 → 1:25 → 1:125 reaches logical error rates of 10⁻¹⁷ enough to run Shor's algorithm.
-------------------------------------
https://t.co/YvQOqR2ATs
-------------------------------------
#QuantumComputing #ErrorCorrection #DeepTech

[Jessy Pensédent @gamesofarcadia]

In 2025, @IBM published a study where they tested 7 quantum compilation SDKs. Among those 7 SDKs, 5 could perform full transpilation:
• Qiskit (IBM),
• QTS (IBM),
• Tket (@QuantinuumQC),
• BQSKit (@BerkeleyLab),
• Staq (@softwareQinc).

Among those 5, only @qiskit passed all 892 tests.
It took 17 hours and 15 minutes.

We ran the exact same IBM suite, with the exact same IBM parameters, using CIQS, the first fully analytic compilation pipeline.

CIQS passed all 892 tests, in 1 hour and 15 minutes, on an older computer.

Our take on the results is two-fold:

1. Under IBM’s own framework, CIQS is ahead of every SDK in the benchmark: up to 15 times faster than Qiskit, while scaling beyond one million qubits, three orders of magnitude past where the IBM benchmark ends. 

2. Standard benchmark metrics such as 2Q gate count and depth are misleading. They are insufficient to characterize the real quality of a compiled circuit on actual hardware.

"Benchmarks do not tell the whole story."

Discuss?

Full paper (7 pages + raw data):
https://t.co/BRtkcaz3uy

Free download (for non-commercial use)
https://t.co/DUUgNije4d

--------------
https://t.co/YvQOqR233U
--------------
#QuantumComputing #Benchmark #DeepTech

[Jessy Pensédent @gamesofarcadia]

UPDATE: This tool has been indexed by INSPIRE-HEP, the canonical database for high energy physics maintained by CERN, DESY, Fermilab and SLAC, and will be included in the next Particle Data Group update.
https://t.co/JWKUFy292Z

[Jessy Pensédent @gamesofarcadia]

CASIMIR is live 🔥 (updated description)
Analytic QCD phase diagram. SU(3), Nf = 3.
It delivers:
• Phase boundary Tc(μB)
• Phase boundary curvature derivatives χ̂₂(μB), χ̂₄(μB)
• Saddle node terminus analysis (CEP exclusion)
• Glueball mass spectrum
• Dense matter onset boundary
• Collider beam energy (μB, Tch, Tc) conversion

Key results:
• κ₂ and κ₄ match Borsanyi+ 2020 fitted values within the quoted lattice uncertainties
• Reconstructed Tc(μB) matches Borsanyi+ 2020 direct Tc points to within 0.15σ at μB = 100, 200 MeV
• Phase boundary terminates at μB* = 421.88 MeV, T* = 113.58 MeV in a saddle node bifurcation
• No critical endpoint: χ̂₂ and χ̂₄ remain finite at the terminus
• Falsifiable prediction: κ₂ₙ = 0 exactly for all n ≥ 4

Instant, no lattice simulations, no Monte Carlo. CSV/PNG export.
Free online tool. No account. No ads.
https://t.co/PPRWlxK284

Open to feedback from heavy-ion physicists, lattice QCD researchers, and anyone working on the QCD phase structure or glueballs.

cc: @BrookhavenLab, @CERN, @Fermilab, @desy

#QCD #ParticlePhysics #HeavyIonPhysics #PhaseDiagram