Following the announcement of Early Access to the NVIDIA DOCA Software Framework at this year’s GTC, held in November, we launched a self-paced DOCA course to help you start working with this new framework. The NVIDIA Deep Learning Institute (DLI) is offering a free self-paced course titled “Introduction to DOCA for DPUs.” In this 2-hour introductory course, you will learn how DOCA and … Continued
Following the announcement of Early Access to the NVIDIA DOCA Software Framework at this year’s GTC, held in November, we launched a self-paced DOCA course to help you start working with this new framework. The NVIDIA Deep Learning Institute (DLI) is offering a free self-paced course titled “Introduction to DOCA for DPUs.” In this 2-hour introductory course, you will learn how DOCA and DPUs enable the development of applications that accelerate data center services. This highly anticipated training covers the essentials of the DOCA platform.
A new pillar of computing
Over the past decade, computing has broken out of the confines of PCs and servers into hyperscale data centers. With this paradigm shift comes Data Processing Units (DPUs), a new class of programmable processors that will join CPUs and GPUs as one of the three pillars of computing. DPUs are designed to offload all the virtual data center such as networking, security, and storage workloads from the CPU. In doing so, they meaningfully reduce overhead for the server CPU to focus on its primary application workload. We’re very encouraged by the prospect and are excited to foster this transformation of data center computing.
What is DOCA?
DOCA is the DPU-enablement platform of software (with libraries, drivers, runtimes, etc.) that abstracts the low-level programming requirements. It is the key to unlocking the potential of DPUs to offload, accelerate, and isolate data center workloads. With DOCA, developers can create applications that address the increasing performance, security, and reliability demands of modern data centers.
What will you learn?
In this course, you will start by getting an overview of DOCA and its features that will shape how you think about the platform and its capabilities. Through live demonstrations, you will also get familiar with the BlueField DPU hardware and the many ways to interact with it in different DOCA development environments. By the end of the course, you will be exposed to numerous examples of how DPU-acceleration can be realized.
By participating in this course, you will learn:
- Basic concepts of DOCA as a platform for accelerated data center computing on DPUs
- The DOCA framework paradigm
- BlueField DPU specifications and capabilities
- Sample DOCA applications under different configurations
- Opportunities to apply DPU accelerated computation
This course contains everything you need to begin working with DOCA. Upon completion, you will also have an opportunity to earn an NVIDIA DLI certificate to demonstrate competency in DOCA. Since one of the main focuses of DOCA is backward compatibility, you can be confident that the investment in development done today will continue to provide performance benefits as DPU hardware generations move forward. It’s never too early to get started!