NVIDIA-Certified Professional

Accelerated Data Science

(NCP-ADS)

About This Certification
Exam Preparation
Exam Blueprint
Contact Us
Stay Informed

About This Certification

About This Certification
Exam Preparation
Exam Blueprint
Contact Us
Stay Informed

About This Certification

The NCP Accelerated Data Science certification is an intermediate-level credential that validates a candidate’s proficiency in leveraging GPU-accelerated tools and libraries for data science workflows. The exam is online and proctored remotely, includes 60–70 questions, and has a 90-minute time limit.

Please carefully review our certification FAQs and exam policies before scheduling your exam.

If you have any questions, please contact us here.

Certification Exam Details

Duration: 90 minutes

Price: $200

Certification level: Professional

Subject: Accelerated data science

Number of questions: 60–70

Prerequisites: Two to three years of hands-on experience in accelerated data science. Strong foundation in machine learning and GPU-accelerated computing. Experience in GPU-based optimization strategies and accelerated data manipulation techniques. Deep understanding of end-to-end data science workflows, from data preparation and cleansing to model development and deployment, with a focus on leveraging GPU acceleration for enhanced performance and efficiency.

Language: English

Validity: This certification is valid for two years from issuance. Recertification may be achieved by retaking the exam.

Credentials: Upon passing the exam, participants will receive a digital badge and optional certificate indicating the certification level and topic.

Exam Preparation

Topics Covered in the Exam

Topics covered in the exam include:

Data analysis
Data manipulation and software literacy
Data preparation
GPU and cloud computing
Machine learning
MLOps

Candidate Audiences

Data scientists
Data engineers
Data analysts
Machine learning engineers
AI DevOps engineers
Applied data scientists
Software engineers
Solution architects
Deep learning performance engineers
Researchers

Recommended Training

Accelerating End-to-End Data Science Workflows

A self-paced course where you’ll learn how to build and execute end-to-end GPU accelerated data science workflows, enabling you to quickly explore, iterate, and get your work into production.

Learn More

Fundamentals of Accelerated Data Science

This course covers the same material as “Accelerating End-to-End Data Science Workflows” but is taught as a live, instructor-led workshop. In it, you’ll learn how to use GPU-accelerated tools to conduct data science faster, leading to more scalable, reliable, and cost-effective results.

Learn More

Enhancing Data Science Outcomes With Efficient Workflow

In this instructor-led workshop, you’ll learn how to develop and deploy an accelerated end-to-end data processing pipeline for large datasets.

Learn More

Exam Blueprint

The table below provides an overview of the topic areas covered in the certification exam and how much of the exam is focused on that subject.

Topics Areas	% of Exam	Topics Covered
Data analysis	14%	Detecting anomalies in a time-series dataset Conducting time-series analysis Creating and analyzing graph data using something like cuGraph Identifying how much data is big data (or when to use which acceleration method) Performing exploratory data analysis (EDA) Visualizing time-series data
Data manipulation and software literacy	19%	Designing and implementing extract, transform, and load (ETL) workflows using accelerated ETL processes Implementing data caching to reduce shuffle Using distributed data processing frameworks for processing big data Implementing data parallelism using Dask for multi-GPU scaling Profiling deep learning models using tools such as DLProf Determining the optimal data processing library to use for varying dataset sizes
Data preparation	17%	Performing data cleansing and preprocessing using CuDF and pandas Transforming and standardizing data Standardizing data as needed to ensure uniformity across features Generating synthetic data to augment datasets using cuDF and NVIDIA RAPIDS™ Identifying and acquiring datasets Monitoring data processing pipelines to recognize bottlenecks Processing, organizing, and storing datasets
GPU and cloud computing	16%	Analyzing graph data using GPU-accelerated tools like cuGraph Optimizing performance of the data science process through GPU acceleration Describing, following, and executing the CRISP-DM process Utilizing dependency management frameworks, such as Docker and Conda to manage software-versioning conflicts Determining the optimal data type choice for each feature Comparing frameworks' performance by designing and implementing a benchmark
Machine learning	15%	Feature engineering Identifying how much data is big data (or when to use which acceleration method) Performing rapid experimentation to find the balance between model accuracy and inference performance Optimizing hyperparameters of machine learning models Training machine learning models, comparing single-GPU and multi-GPU scenarios Using GPU memory-optimization techniques, such as batching and mixed precision, to train machine learning models
MLOps	19%	Determining the optimal data type choice for each feature Assessing and verifying the memory size of a dataset Comparing the required memory with the available memory on a device Performing benchmarking and optimizing different GPU-accelerated workflows Deploying and monitoring models in production

Exam Study Guide

Coming soon.

Get in Touch

NVIDIA offers training and certification for professionals looking to enhance their skills and knowledge in the field of AI, accelerated computing, data science, advanced networking, graphics, simulation, and more.

Stay Up to Date

Get training news, announcements, and more from NVIDIA, including the latest information on new self-paced courses, instructor-led workshops, free training, discounts, and more. You can unsubscribe at any time.