深

深圳市大数据研究院 Shenzhen Research Institute of Big Data Tech Stack

AI research institute focused on large-model inference and biomedical data systems

Computer and Network Security Shenzhen, Guangdong 11–50 employees Founded 2016 Educational

SRIBD is a government-backed research institute in Shenzhen specializing in large-language-model optimization and biomedical AI. The tech stack—PyTorch, CUDA, vLLM, TensorRT, MLIR, and NVIDIA GPUs—reflects a deep focus on inference acceleration and model compilation. Active pain points around KV-cache optimization, edge deployment, and low-precision quantization, paired with projects on LLM pretraining and single-cell foundation models, signal a research agenda bridging cloud-scale inference and resource-constrained deployment.

Tech Stack 38 technologies

Core StackPython Linux PyTorch C++ TensorFlow Neo4j Bash DICOM NIfTI SimpleITK Transformer CUDA vLLM SGLang Triton JAX CMake Git VTune ONNX Runtime NCNN llama.cpp NVIDIA GPU OpenCL LLVM TensorRT MLIR GPT-4 LLaMA TigerGraph+8 more

What 深圳市大数据研究院 Shenzhen Research Institute of Big Data Is Building

◆Challenges

Large model inference acceleration
Building large-scale single-cell multi-omics database
Optimizing large model architecture
Optimizing edge device deployment
Low precision model performance
Knowledge distillation from large models
Compression inference optimization

▲Active Projects

Health management technology r&d
Provincial standardization demonstration project
国家重大科技专项
国家自然科学基金项目
司法部法律大数据与智能装备重点实验室
Single-cell foundation model development
Large model pretraining and fine-tuning
Cell image ai recognition project
Kvcache multi-level cache optimization
Scheduling and inference optimization

Hiring Activity

Accelerating15 roles · 7 in 30d

Department

Research

Engineering

Data

Healthcare

Ops

Sales

Seniority

Intern

Mid

Senior

Company intelligence

Find more companies like 深圳市大数据研究院 Shenzhen Research Institute of Big Data by tech stack, pain points and active projects

Get started free

About 深圳市大数据研究院 Shenzhen Research Institute of Big Data

SRIBD was established in 2016 with government approval and operates as an independent research institution under Shenzhen's Peacock talent program. Located on the CUHK (Shenzhen) campus, the institute targets precision medicine, intelligent cities, and communications infrastructure. The team comprises doctoral-level researchers with overseas experience. Current work spans large-model pretraining and fine-tuning, single-cell AI recognition, and optimization for edge devices—supported by access to the National Supercomputing Center in Shenzhen. Hiring has accelerated, with internship-heavy staffing in research roles and emerging engineering capacity.

HeadquartersShenzhen, Guangdong

Company Size11–50 employees

Founded2016

Hiring MarketsChina

Frequently Asked Questions

What is Shenzhen Research Institute of Big Data's tech stack?

Primary tools: Python, PyTorch, CUDA, Linux. Model optimization: vLLM, TensorRT, MLIR, llama.cpp, ONNX Runtime, Triton. Graph databases: Neo4j, TigerGraph. Medical imaging: DICOM, NIfTI, SimpleITK. All work targets GPU-accelerated inference on NVIDIA hardware.

What projects is SRIBD working on?

Focus areas: large-model pretraining and fine-tuning, single-cell foundation models, cell-image AI recognition, and biomedical database construction. Infrastructure: KV-cache optimization, scheduling and inference optimization for edge devices, and low-precision model compression for deployment.