arXiv subjects

Gilbert Bernstein

Publications and source records attributed to Gilbert Bernstein.

9 recordsLinked to original sources

arXiv2026

Piper: A Programmable Distributed Training System

Large-scale model training increasingly relies on composing multiple parallelism strategies, such as data, pipeline, and expert parallelism, together with memory-saving optimizations like ZeRO. Deployed systems for foundation model pretraining often rely on human experts to manually design a high-level parallelism strategy then implement the corresponding low-level execution strategy, making it difficult to adapt the system to new strategies. Meanwhile, many general-purpose frameworks are more flexible but their implementations are still tied to a fixed set of common parallelism strategies, making it challenging to integrate state-of-the-art strategies. We present Piper, a user-controllable distributed training system that decouples the strategy from the runtime implementation. Piper allows users to declare a comprehensive distributed training strategy with a small set of model annotations and scheduling directives. Each directive applies a transformation on Piper's intermediate representation (IR), a unified global training DAG that represents all computation and communication. Using this IR, Piper compiles per-device execution plans and executes them with a distributed runtime agnostic to the strategy. We show that the combined system maintains performance parity on commonly available strategies such as ZeRO, while also enabling additional performance and memory efficiency gains through joint scheduling of compute and communication in composed parallelism strategies such as DeepSeek-V3's DualPipe.

cs.DC↗

arXiv2026

YASPS: A Symbolic Framework for Extensible, High-Performance IPC Simulation

Incremental Potential Contact (IPC) enables robust, contact-rich simulation by casting elasticity and contact as a single energy minimization problem, but high-performance IPC pipelines are typically built from specialized kernels and assembly logic tied to fixed energies, primitive types, and parameterizations, making extensions costly and combinatorial. We present YASPS, a GPU-oriented framework that removes this extensibility bottleneck by making structure explicit in a differentiable intermediate representation. YASPS introduces two first-class relational operators: JOIN, which composes dependent quantities across user-declared relations (e.g., element-to-vertex connectivity), and UNION, which represents alternative parameterizations within a relation (e.g., mixing free vertices with affine-body or other parameterizations without fragmenting the program). Because JOIN and UNION are part of the symbolic program, YASPS differentiates through them using dedicated rules and an efficient second-order procedure that reuses intermediate Jacobians and reduces Hessian-projection cost. From the same relational description, YASPS derives the global gradient/Hessian sparsity and block layout, enabling structure-aware block-sparse storage and compression, and JIT-compiles CUDA kernels for evaluation, derivatives, assembly, and solving. Across IPC-style examples, including layered cloth-on-bunny, mixed rigid/deformable bunnies, and a caged deformation model, YASPS supports rapid front-end extensions with minimal back-end changes while achieving competitive end-to-end performance; its Hessian compression yields near 10x faster CG iterations in our benchmarks.

cs.GR↗

arXiv2026

Rethinking Collision Detection on GPU Ray Tracing Architecture

Discrete Collision Detection (DCD) is a fundamental task in several domains including particle-based physics simulations. Efficient DCD uses indexing structures such as Bounding Volume Hierarchy (BVH), but accelerating irregular BVH traversals demands meticulous efforts to achieve performance. Modern GPUs feature Ray Tracing (RT) architecture that provides hardware acceleration for BVH traversal and optimized drivers for BVH construction. Recent work has attempted to exploit RT architecture to accelerate DCD on spherical particles by reducing DCD to fixed-radius neighbor search. However, this reduction breaks down for particles with different radii, necessitating the use of large bounding boxes that result in a higher number of duplicate collisions and poor performance. To address these limitations, we present Mochi, a new reduction that reformulates DCD on RT architecture by exploiting the symmetry of collision relations to support both uniform and non-uniform spherical particles efficiently. Mochi introduces per-object proxy spheres that decouple BVH bounding volumes from the collision search radius, enabling significantly tighter bounding boxes without sacrificing correctness. Mochi is provably sound and guarantees that all true collisions are detected. We integrate Mochi into an end-to-end particle simulation pipeline and evaluate it across large-scale particle workloads, showing consistent speedups over state-of-the-art BVH-based and RT-based DCD implementations. Mochi generalizes prior RT-based neighbor search formulations while avoiding their fundamental limitations for non-uniform spheres.

cs.GR↗

arXiv2025

Jenga: Responsive Tiered Memory Management without Thrashing

A heterogeneous memory has a single address space with fast access to some addresses (a fast tier of DRAM) and slow access to other addresses (a capacity tier of CXL-attached memory or NVM). A tiered memory system aims to maximize the number of accesses to the fast tier via page migrations between the fast and capacity tiers. Unfortunately, previous tiered memory systems can perform poorly due to (1) allocating hot and cold objects in the same page and (2) abrupt changes in hotness measurements that lead to thrashing. This paper presents Jenga, a tiered memory system that addresses both problems. Jenga's memory allocator uses a novel context-based page allocation strategy. Jenga's accurate measurements of page hotness enable it to react to memory access behavior changes in a timely manner while avoiding thrashing. Compared to the best previous tiered memory system, Jenga runs memory-intensive applications 28% faster across 10 applications, when the fast tier capacity matches the working set size, at a CPU overhead of <3% of a single core and a memory overhead of <0.3%

cs.ET↗

arXiv2024

Exo 2: Growing a Scheduling Language

User-schedulable languages (USLs) help programmers productively optimize programs by providing safe means of transforming them. Current USLs are designed to give programmers exactly the control they want, while automating all other concerns. However, there is no universal answer for what performance-conscious programmers want to control, how they want to control it, and what they want to automate, even in relatively narrow domains. We claim that USLs should, instead, be designed to grow. We present Exo 2, a scheduling language that enables users to define new scheduling operations externally to the compiler. By composing a set of trusted, fine-grained primitives, users can safely write their own scheduling library to build up desired automation. We identify actions (ways of modifying code), inspection (ways of interrogating code), and references (ways of pointing to code) as essential for any user-extensible USL. We fuse these ideas into a new mechanism called Cursors that enables the creation of scheduling libraries in user code. We demonstrate libraries that amortize scheduling effort across more than 80 high-performance kernels, reducing total scheduling code by an order of magnitude and delivering performance competitive with state-of-the-art implementations on three different platforms.

cs.PL↗

arXiv2024

Learning to Move Like Professional Counter-Strike Players

In multiplayer, first-person shooter games like Counter-Strike: Global Offensive (CS:GO), coordinated movement is a critical component of high-level strategic play. However, the complexity of team coordination and the variety of conditions present in popular game maps make it impractical to author hand-crafted movement policies for every scenario. We show that it is possible to take a data-driven approach to creating human-like movement controllers for CS:GO. We curate a team movement dataset comprising 123 hours of professional game play traces, and use this dataset to train a transformer-based movement model that generates human-like team movement for all players in a "Retakes" round of the game. Importantly, the movement prediction model is efficient. Performing inference for all players takes less than 0.5 ms per game step (amortized cost) on a single CPU core, making it plausible for use in commercial games today. Human evaluators assess that our model behaves more like humans than both commercially-available bots and procedural movement controllers scripted by experts (16% to 59% higher by TrueSkill rating of "human-like"). Using experiments involving in-game bot vs. bot self-play, we demonstrate that our model performs simple forms of teamwork, makes fewer common movement mistakes, and yields movement distributions, player lifetimes, and kill locations similar to those observed in professional CS:GO match play.

cs.LG↗

arXiv2022

Differentiable Rendering of Neural SDFs through Reparameterization

We present a method to automatically compute correct gradients with respect to geometric scene parameters in neural SDF renderers. Recent physically-based differentiable rendering techniques for meshes have used edge-sampling to handle discontinuities, particularly at object silhouettes, but SDFs do not have a simple parametric form amenable to sampling. Instead, our approach builds on area-sampling techniques and develops a continuous warping function for SDFs to account for these discontinuities. Our method leverages the distance to surface encoded in an SDF and uses quadrature on sphere tracer points to compute this warping function. We further show that this can be done by subsampling the points to make the method tractable for neural SDFs. Our differentiable renderer can be used to optimize neural shapes from multi-view images and produces comparable 3D reconstructions to recent SDF-based inverse rendering methods, without the need for 2D segmentation masks to guide the geometry optimization and no volumetric approximations to the geometry.

cs.GR↗

arXiv2020

Differentiating a Tensor Language

How does one compile derivatives of tensor programs, such that the resulting code is purely functional (hence easier to optimize and parallelize) and provably efficient relative to the original program? We show that naively differentiating tensor code---as done in popular systems like Tensorflow and PyTorch---can cause asymptotic slowdowns in pathological cases, violating the Cheap Gradients Principle. However, all existing automatic differentiation methods that guarantee this principle (for variable size data) do so by relying on += mutation through aliases/pointers---which complicates downstream optimization. We provide the first purely functional, provably efficient, adjoint/reverse-mode derivatives of array/tensor code by explicitly accounting for sparsity. We do this by focusing on the indicator function from Iverson's APL. We also introduce a new "Tensor SSA" normal form and a new derivation of reverse-mode automatic differentiation based on the universal property of inner-products.

cs.PL↗

arXiv2016

Opt: A Domain Specific Language for Non-linear Least Squares Optimization in Graphics and Imaging

Many graphics and vision problems can be expressed as non-linear least squares optimizations of objective functions over visual data, such as images and meshes. The mathematical descriptions of these functions are extremely concise, but their implementation in real code is tedious, especially when optimized for real-time performance on modern GPUs in interactive applications. In this work, we propose a new language, Opt (available under http://optlang.org), for writing these objective functions over image- or graph-structured unknowns concisely and at a high level. Our compiler automatically transforms these specifications into state-of-the-art GPU solvers based on Gauss-Newton or Levenberg-Marquardt methods. Opt can generate different variations of the solver, so users can easily explore tradeoffs in numerical precision, matrix-free methods, and solver approaches. In our results, we implement a variety of real-world graphics and vision applications. Their energy functions are expressible in tens of lines of code, and produce highly-optimized GPU solver implementations. These solver have performance competitive with the best published hand-tuned, application-specific GPU solvers, and orders of magnitude beyond a general-purpose auto-generated solver.

cs.GR↗