Simplify Your AI Development with Intel DevCloud and DevCatalog
Table of Contents
- Introduction
- The Mission of Our Organization
- Understanding the Challenges of Developers
- Introduction to Intel Developer Cloud
- Containerization with Intel Developer Cloud
- Developing Applications on Intel Developer Cloud
- Intel Developer Cloud at Edge
- Developer Catalog in the Developer Cloud
- testing Applications on Intel Hardware
- How DevCloud at Edge Works
- Web UI
- Development Server
- Storage
- Queue
- Hardware Deployment
- Workflows in DevCloud
- Demo: Developing, Building, and Testing Applications in DevCloud
- Using Intel's Docker Hub
- Exploring Intel Developer Catalog
- Conclusion
🖥️ Introduction
Welcome to DockerCon! In this article, we'll be discussing the whole AI developer journey with Intel products like Intel Developer Catalog in the Developer Cloud and Open Window Toolkit. Our focus will be on how a containerization product from Intel can enhance developers' development experiences and help them bring their products to market easily.
🎯 The Mission of Our Organization
At Intel, our mission is to reduce the complexity that developers face while using Intel hardware and software. We work closely with customers, developers, and companies to understand the challenges they face in taking their products to market. By engineering our products to solve these challenges, we aim to help developers easily fix problems and successfully bring their products to market.
🌟 Understanding the Challenges of Developers
Working with a diverse range of developers, we have come to understand the challenges they face in the development and deployment process. Our team focuses on simplifying and easing the entire process, aligning it with cloud-native development for edge computing and cloud platforms.
💡 Introduction to Intel Developer Cloud
Intel Developer Cloud provides a platform designed specifically for developers to develop, build, and test their applications on different Intel hardware. This cloud platform eliminates the need for developers to procure and benchmark hardware on their own. With Intel Developer Cloud, developers can evaluate the performance of their applications on various hardware options before making informed decisions on which hardware to purchase for their deployment.
📦 Containerization with Intel Developer Cloud
One of the challenges faced by developers worldwide is benchmarking their applications' performance on different hardware. Traditionally, developers would need to buy and test different hardware configurations to make the right decision. However, Intel Developer Cloud offers a solution to this problem. We have hosted different devices from Intel in our data center, ranging from Core processors to Xeon processors with various accelerator cards. Developers can come to Intel Dev Cloud, test their applications, evaluate performance, and make informed decisions about the hardware they need to purchase.
🚀 Developing Applications on Intel Developer Cloud
Intel Developer Cloud provides developers with a comprehensive software lifecycle. Developers can develop containerized or application-based software solutions using the platform. For this discussion, we will primarily focus on the development and deployment of containerized solutions.
🌌 Intel Developer Cloud at Edge
Developers working on edge computing and cloud platforms can leverage Intel Developer Cloud at Edge. With a wide range of devices available from Intel, such as CPUs, GPUs, VPUs, and FPGAs, developers can create tailored solutions for different devices. They can develop their solutions directly on the platform without the need for separate hardware or installations. The browser-based application allows developers to interact with the platform and build their solutions effortlessly.
Developers using Intel Developer Cloud at Edge can build containerized solutions for various verticals, including Healthcare and more. We also provide pre-built solutions for developers to explore and try out.
📚 Developer Catalog in the Developer Cloud
Intel Developer Cloud offers a Developer Catalog, a web UI catalog that lists all the consolidated Intel software tools and SDKs. The Developer Catalog serves as a centralized resource for developers, providing easy access to the software they need for their development projects. Developers can explore the catalog, find the right tools and SDKs, and leverage them to enhance their development experiences.
🔬 Testing Applications on Intel Hardware
Once developers have developed and built their applications on Intel Developer Cloud, they can test them on various hardware configurations. Developers can test their applications on different Intel hardware, ranging from Core processors to Xeon processors with different memory footprints and accelerator options. By testing their applications on multiple hardware configurations, developers can Gather performance metrics and evaluate the best hardware options for their deployment models. Intel Developer Cloud provides a telemetry interface where developers can view and analyze the performance metrics, helping them make informed decisions about hardware purchases.
⚙️ How DevCloud at Edge Works
DevCloud at Edge consists of five key building blocks that provide developers with an efficient and seamless experience:
-
🌐 Web UI: The browser-based interface allows developers to access the platform without the need for additional software installations. They can interact with the Developer Container Playground, which serves as the development and testing environment.
-
💻 Development Server: The Development Server, equipped with Jupiter Lab, provides developers with a space to develop their applications. It enables them to write and test code efficiently.
-
💾 Storage: As developers work on their applications, all the resources and code are stored in the storage system. This ensures that developers can easily access and manage their files and projects.
-
⏩ Queue: Once developers have built their applications, they are added to the queue system. The queue ensures a fair deployment process and waits for the specified hardware to become available for the application's deployment.
-
🖥️ Hardware Deployment: Once the required hardware becomes available, the application is deployed to the specific hardware selected by the developer. Developers can then observe and analyze the application's performance on the deployed hardware.
🔄 Workflows in DevCloud
DevCloud offers four different workflows to cater to developers' diverse needs:
-
Self-Development: Developers who want to develop their own applications from scratch can leverage the Jupyter Lab environment. They can utilize the OpenVINO toolkit, specifically designed for Intel hardware, to optimize inference on CPUs, GPUs, VPUs, and FPGAs.
-
Docker File Import: Developers with existing projects hosted on Git repositories or Docker Hub can import their Docker files directly from these sources. They can build Docker images and deploy them on Intel hardware.
-
Existing Application Testing: If developers already have their applications available in the form of helm charts, Docker Compose files, or Docker Hub/GCP/AWS registries, they can easily test these applications on different Intel hardware configurations using DevCloud.
-
Exploring End-to-end Solutions: Developers who want to understand how DevCloud works or explore various use cases can access pre-built solutions and use them on the DevCloud at Edge infrastructure. This provides a hands-on experience with real-world applications.
🎬 Demo: Developing, Building, and Testing Applications in DevCloud
To illustrate the process of developing, building, and testing applications in DevCloud, let's walk through a quick demo.
To access DevCloud, visit devcloud.intel.com and log in using your credentials. Once logged in, you will be directed to the Container Playground, where you can develop, build, and test your applications.
In the Container Playground, you have multiple options for deploying your applications. You can import resources from containers in the registry, directories, Docker files, Git repositories, or even launch Helm charts. For this demo, we'll focus on deploying a pre-built solution from the DevCloud Marketplace.
By exploring the DevCloud Marketplace, you can find a wide array of examples organized by different verticals such as retail, industrial, and healthcare. For example, let's choose the "People Counter System" as our application. We can launch it and select the desired hardware for deployment.
Once launched, the application will go through the queueing system and eventually be deployed on the specified hardware. You can then view the application's output, such as videos showing people counting or performance metrics like frames per Second (FPS) and latency. DevCloud provides extensive logging and telemetry interfaces for analyzing application performance.
🐳 Using Intel's Docker Hub
Intel collaborates with Docker Hub to provide an extensive collection of Docker images as a verified publisher. On hub.docker.com, you can find various images for AI, HPC, IoT, and edge computing workloads.
For example, let's search for "video analytics serving" on Docker Hub. We can find Intel's Docker image for video analytics serving based on the OpenVINO Toolkit. This image allows developers to leverage inference as a service for Intel CPUs, GPUs, VPUs, and FPGAs. By copying the container URL and importing it into the DevCloud Container Playground, developers can easily deploy and test their applications using this image.
📚 Exploring Intel Developer Catalog
Intel's Developer Catalog provides a wide range of software tools, libraries, and SDKs tailored to different verticals and developer focuses. From AI to high-performance computing, the Developer Catalog offers standalone software, containers, and Helm charts.
To access the Developer Catalog, visit developer.intel.com/devcatalog. You can search for specific software or browse through different categories. For example, if you search for "intelligent traffic management," you will find applications and solutions designed for traffic management using Intel technologies. These solutions are ready for deployment on various Intel-based platforms.
✅ Conclusion
In conclusion, Intel Developer Cloud and the Developer Catalog are powerful tools that enable developers to tackle complex challenges in their development journeys. With the ability to develop, build, and test applications on Intel hardware, developers can optimize their workflows and bring their products to market more efficiently. Whether it's leveraging Intel's Docker Hub or exploring the Intel Developer Catalog, developers have access to a rich ecosystem of resources to enhance their development experiences. So, start exploring Intel Developer Cloud and unleash your potential!
🌟 Highlights
- Intel Developer Cloud provides a platform for developers to develop, build, and test applications on Intel hardware.
- Intel's mission is to reduce the complexity that developers face while using Intel hardware and software.
- The DevCloud at Edge offers developers a space to develop applications for edge computing and cloud platforms.
- The Developer Catalog is a web UI catalog that lists all the consolidated Intel software tools and SDKs.
- Developers can test their applications on different Intel hardware configurations to evaluate performance and make informed decisions.
- The Container Playground in DevCloud enables developers to deploy, develop, and test their containerized workloads.
- Intel collaborates with Docker Hub to provide a wide range of Docker images for various workloads.
🙋♀️ FAQ
Q: What types of hardware can developers test their applications on in Intel DevCloud?
A: Intel DevCloud supports a range of hardware options, including Core processors, Xeon processors, GPUs, VPUs, and FPGAs.
Q: Can developers import Docker files from their own Git repositories?
A: Yes, developers can import Docker files from their Git repositories directly into the DevCloud Container Playground.
Q: Is there a limit to the number of applications developers can deploy in the DevCloud?
A: There are no specific limits on the number of applications developers can deploy in the DevCloud. However, it operates on a queuing system, so applications may wait for hardware availability.
Q: Can developers access performance metrics and logs for their applications in the DevCloud?
A: Yes, DevCloud provides developers with access to performance metrics and logs for their applications, helping them analyze and optimize performance.
Q: Is the DevCloud platform suitable for both individual developers and large organizations?
A: Yes, the DevCloud platform caters to the needs of both individual developers and larger organizations. It provides flexibility and scalability for different project sizes.
🔗 Resources
No difficulty
WHY YOU SHOULD CHOOSE TOOLIFY
