SaaSearch

NablaTensor

Unclaimed Listing

Adjoint automatic differentiation engine on the JVM

launched today•5.0(1)
NablaTensor screenshot 1 of 1

About NablaTensor

NablaTensor is an advanced adjoint automatic differentiation engine built for the JVM, designed to record a Monte-Carlo valuation once as a flat trace of primitive operations and replay it across millions of scenarios. It enables quantitative finance developers and high-performance computing engineers to yield full first-order gradients at a fraction of the cost of central bump-and-revalue methods. The platform features generated bytecode, SIMD, and GPU support, providing deterministic path-for-path sensitivities independent of the number of risk factors.

Pricing model

Freemium (Free tier + Paid plans)

Tags & Keywords

Tech stack

Social accounts

GitHub

Discussion (1)

Community Rating & Reviews

5.0
(1)
Verified community feedback
Rate this product:
5.0
Posting as:
Sign in to rate this product or share a review.
The SaaSearch Team
The SaaSearch TeamTeamtoday

It's truly remarkable how NablaTensor records a Monte-Carlo valuation once and then replays it across millions of scenarios, delivering deterministic path-for-path sensitivities and significantly faster first-order gradients.

Visit nablatensor.com

Questions about NablaTensor

People also search for