aadc-quantlib-tracing 1.41.0 on PyPI
QuantLib 1.41 built with AADC source tracing, so a Python script can ask an unmodified C++ pricer which constants, branches and inputs produced a number. Linux x86-64.
PyPI
Releases, talks, awards and papers, newest first.
QuantLib 1.41 built with AADC source tracing, so a Python script can ask an unmodified C++ pricer which constants, branches and inputs produced a number. Linux x86-64.
PyPIDmitri Goloubentsev presented at the 22nd WBS Quantitative Finance Conference. When a pricing model runs on AADC, the whole calculation is recorded with every value tied to the source file and line that produced it. The talk showed what a validator gets from one such recording of an unmodified QuantLib swaption: the hardcoded constants with their coordinates, the frozen branches, which market quotes a discount factor depends on and which it provably does not, and where a replay stops being valid. The evidence is laid out against PRA SS1/23, SR 26-2 and the ECB guide, and the measured figures on the product page reproduce from a pip install. Earlier that day Dmitri joined Alexander Sokol, Nicole Königstein and Achintya Gopal on the opening panel, Is Active Deception Part of AI Model Risk?
About the AADC TracerThe 2.x line is a new engine: record once, replay the plain tape for a one-off valuation, or compile it for repeated use and batches. Wheels for Linux x86-64 and aarch64, Windows x64 and, since 18 September, Apple Silicon, so a model developed on a Mac runs unchanged on the grid. The Community Edition is free for non-commercial and academic use and runs the interpreter and every JIT backend at full speed; an Enterprise licence adds the C++ SDK, AVX-512, the ahead-of-time code generators and production use. Intel Macs are not supported.
pip install aadcEnergy Risk named MatLogica innovation of the year in the technology category at the 2026 awards, announced at the Energy Risk USA dinner in Houston on 7 May, for the gas storage work in which a neural policy is trained through the recorded simulation.
Risk.netDmitri Goloubentsev and Natalija Karpichina publish SNAPO (Smooth Neural Adjoint Policy Optimization): a neural policy inside a differentiable simulator with smooth constraints and exact adjoint gradients, demonstrated on natural gas storage, pension fund asset-liability management and pharmaceutical manufacturing.
arXiv:2605.06570At the 21st WBS Quantitative Finance Conference, Dmitri Goloubentsev presented a production-ready way to compute autocallable Greeks with smoothing and AAD, cutting the computational cost by 90% without losing accuracy.
Read the write-upChartis Research again named MatLogica Category Leader in automatic differentiation, with five category awards in total and an overall ranking of 13th in the Quantitative Analytics 50.
AwardsOn matched quant workloads AADC ran more than ten times faster than the three machine learning frameworks.
Read the benchmarkOn the final evening of QuantMinds International, Dmitri Goloubentsev and George Petropoulos of Delta Capita hosted a roundtable joined by Jesper Andreasen, Peter Jäckel and Serguei Issakov.
Risk.net named MatLogica technology newcomer of the year in its Asia Risk Technology Awards.
Risk.netDmitri Goloubentsev presented how live risk can be run in a cloud environment without the cost usually attached to it.
Read the architectureA free workshop on keeping Python for prototyping while getting compiled performance and AAD in production. The recording is on YouTube.
WatchIntel's developer site published a walkthrough of accelerating simulations and backpropagation with AADC from Python and C++ analytics, using vectorisation and adjoint differentiation.
Read on intel.comA target architecture for cloud-based live risk that uses code generation AAD to make sensitivities fast and cheap enough to compute continuously.
ReadMatLogica was ranked 10th in the Chartis QuantTech50. Dmitri Goloubentsev's presentation, Leveraging Automatic Adjoint Differentiation and Cloud for Real-time Risk, is on YouTube.
WatchMatLogica was shortlisted for FinTech start-up of the year by Informa Connect.
ShortlistTwo conference talks used MatLogica's code generation AAD: Stephan Bosch of ING on comparing AAD techniques and performance, and Svetlana Borovkova of Probability & Partners on estimating expected shortfall sensitivities with AADC.
PresentationsChartis Research gave MatLogica four awards: innovation, AAD, data-parallel programming, and innovation in computational frameworks, with an overall ranking of 10th out of 50.
ChartisMatLogica graduated from the 2023 Accenture FinTech Innovation Lab London programme, one of fifteen companies in the cohort.
AccentureDmitri Goloubentsev's Barcelona talk on the automatic implicit function theorem, based on the paper by Goloubentsev, Lakshtanov and Piterbarg published on Risk.net and SSRN.
WatchMatLogica signed a memorandum of understanding with Tachyum to bring the AADC toolkit to the Tachyum Prodigy processor.
CUDA analytics can now be accelerated by AADC on a CPU, with AAD as an option.
How it worksA visual demonstration of QuantLib accelerated by AADC, running live.
DemosProf. Roland Olsson used AADC to train custom neural network architectures for time series analysis. The paper reports better accuracy than the methods it compares against and several times lower training time.
Paper on arXivAdopting AADC gave the bank 15 to 20 times faster risk, cut overnight portfolio risk from more than 8 hours to 2, and intraday risk from more than 30 minutes to a few.
Client resultsAn online environment with the multi-curve fitting and AAD risk demo on QuantLib and an external Levenberg-Marquardt library, and a video on reaching a 350× performance gain for XVA pricing on Intel AVX2 with 5 threads.
DemosDr Antoine Savine, author of Modern Computational Finance (Wiley, 2018) and best known for his work on volatility and interest rate models, became an advisor to MatLogica.
Dmitri Goloubentsev on AAD integration strategies for top performance and ease of use, presented at QuantMinds in Barcelona.
WatchChartis Research named MatLogica a Category Leader in the XVA components quadrant of its RiskTech100 research.
AwardsDmitri Goloubentsev on implementing adjoint differentiation efficiently for Longstaff-Schwartz.
PresentationsBased on results with QuantLib and ORE, a demonstration of how the library works and the idea of integration complexity.
SIAMLeaders from Intel, MatLogica and Quantifi on how Intel Xeon Scalable processors and Intel software improve the performance of financial risk models.
WatchMatLogica's approach to top performance, shown on QuantLib: 150× faster XVA on a single core.
WatchDmitri Goloubentsev introduced MatLogica's technique for speeding up repetitive calculations.
WatchThe first virtual, worldwide workshop on automatic differentiation.
ProgrammeMatLogica spoke at Risk.net's Quant Summit Europe.
SpeakersMatLogica presented at Intel's developer workshop.
WatchJoint results of Intel and MatLogica on two key XVA benchmarks, with gains of up to 1000× on Xeon Scalable CPUs.
WBSThe talk behind the paper AAD: Breaking the Primal Barrier.
Paper (PDF)First public presentation of the idea behind the AAD compiler, in the XVA and AAD stream.
WBS