GPU-Accelerated Quantum Error Detection

Fri, 15 May 2026 00:00:00 +0000

GPU-Accelerated Quantum Error Detection

Overview

As quantum processors scale into the Noisy Intermediate-Scale Quantum (NISQ) era, characterization and mitigation of gate errors are essential for achieving fault-tolerant quantum computation. This research project presents a high-throughput, GPU-accelerated quantum simulation pipeline that accurately models stochastic Pauli error channels using NVIDIA cuQuantum tensor-network backends and Qiskit.

flowchart TD Q[Parameterized Quantum Circuit\n+ Noise Channels] --> T[cuQuantum Tensor Network\nStatevector Simulation] T --> G[GPU Accelerated Contraction\nNVIDIA Tensor Cores] G --> S[Statistical Sampling Algorithm\nSyndrome Measurement] S --> E[Single- & Two-Qubit\nGate Error Isolation] E --> V[QEC Verification & Mitigation]

Research & Engineering Highlights

1. High-Performance Tensor-Network Simulation

Modeled realistic quantum noise using stochastic Pauli channels ($X, Y, Z$ depolarizing noise) integrated into multi-qubit parameterized quantum circuits.
Accelerated statevector generation and tensor contractions leveraging NVIDIA cuQuantum Appliance and custom CUDA routines, achieving order-of-magnitude speedups over classical CPU simulator backends.

2. Statistical Error Syndrome Sampling

Developed a statistical sampling algorithm capable of localizing single-qubit and two-qubit entangling gate errors with minimal measurement overhead.
Reduced the syndrome extraction rounds required to verify circuit fidelity under high-depth algorithmic circuits.

3. Scalable QEC Verification

Currently drafting a journal publication formalizing this methodology as an efficient verification protocol for Quantum Error Correction (QEC) schemes on near-term hardware.

Publication Status

Scalable Quantum Error Detection via Tensor-Network Statevector Sampling (2026). In Preparation for Journal Submission. Primary Author: Zerui “Jerry” Ma.

High-Fidelity AI Video Generation Engine (LTX-2.5)

Sun, 01 Mar 2026 00:00:00 +0000

High-Fidelity AI Video Generation Engine (LTX-2.5)

Overview

Engineered an ultra-low-latency, memory-efficient video generation pipeline leveraging the LTX-2.5 Diffusion Transformer (DiT) architecture on a consumer NVIDIA GeForce RTX 5080 GPU (Blackwell architecture). By synthesizing custom low-precision quantized GEMM kernels and an asynchronous layer-wise weight offloading subsystem, this engine delivers high-resolution, multi-frame video synthesis within a 16GB VRAM budget.

flowchart LR A[Host System RAM\nPinned Memory] <-->|Async DMA Stream\nLayer-wise Weights| B[NVIDIA RTX 5080\n16GB GDDR7] subgraph GPU Computation B --> C[Custom NVFP4 GEMM Kernels\nDynamic FP4/FP8 Casting] C --> D[Spatio-Temporal DiT Layers\nFlashAttention + CUDA Graphs] D --> E[Latent Frame Synthesis] end E --> F[High-Fidelity Video Output]

Technical Architectural Innovations

1. Custom NVFP4 Quantized GEMM Compilation

Designed and compiled custom NVFP4 (NVIDIA 4-bit Floating Point) quantized GEMM kernels utilizing Blackwell tensor cores.
Implemented robust dynamic FP4/FP8 casting with adaptive block-level scale-factor management, maintaining numerical stability across extreme gradient dynamics.
Achieved a 65% reduction in model memory footprint with negligible visual fidelity degradation compared to standard FP16/BF16 baselines.

2. Asynchronous Pinned-Memory Layer-wise Offloading

Architected an asynchronous double-buffered weight streaming engine that exploits high-bandwidth PCIe Gen 5 DMA transfers.
Streams individual DiT transformer blocks into VRAM concurrently with the forward-pass computation of preceding layers, completely hiding memory transfer latency.
Enabled continuous multi-frame 1080p video generation within consumer 16GB VRAM, eliminating out-of-memory (OOM) exceptions.

3. FlashAttention & CUDA Graph Optimization

Integrated fused FlashAttention kernels tailored for 3D spatio-temporal attention matrices.
Captured static execution branches into CUDA Graphs, eradicating CPU-GPU launch overhead and cutting end-to-end inference latency by 42%.

Key Results & Benchmarks

Memory Footprint: Reduced from ~42 GB (BF16 unquantized) to <14.8 GB active VRAM.
Latency Acceleration: 42% end-to-end speedup over native PyTorch baseline implementations.
Visual Quality: Retained smooth temporal coherence and high perceptual score across complex multi-object video generations.

CUDA | Zerui "Jerry" Ma's personal website

GPU-Accelerated Quantum Error Detection

GPU-Accelerated Quantum Error Detection

Overview

Research & Engineering Highlights

1. High-Performance Tensor-Network Simulation

2. Statistical Error Syndrome Sampling

3. Scalable QEC Verification

Publication Status

High-Fidelity AI Video Generation Engine (LTX-2.5)

High-Fidelity AI Video Generation Engine (LTX-2.5)

Overview

Technical Architectural Innovations

1. Custom NVFP4 Quantized GEMM Compilation

2. Asynchronous Pinned-Memory Layer-wise Offloading

3. FlashAttention & CUDA Graph Optimization

Key Results & Benchmarks