AI developer: Why you should be using Intel DevCloud if you want to make the most of your Intel hardware

From tutorials to prototyping and onto cost-effective scaling, DevCloud gives you the advantages your business needs

Programmer working with program code

You and your managers want to know that the AI models you're testing will run efficiently and cost-effectively in production – and with the power of Intel DevCloud, you can.

DevCloud is an all-in-one online destination to test the performance of your AI models on Intel processors. It's a dedicated space that allows developers and engineers to innovate in a production-ready space. What's great about this platform is that it gives you access to dedicated Intel technology – without having to invest in hardware upfront. 

You can bring your AI model to DevCloud and then test it across a range of Intel processors. This development sandbox allows your business to try AI workloads remotely before purchasing hardware, so your business can be sure it's buying the right processors.

How you can use the DevCloud to find the right hardware setup

With DevCloud, developers and engineers don't need to have the actual hardware in their hands. They can test their models from anywhere in the world as long as they have an internet connection. 

First, go to DevCloud to sign up and receive approval. Once you're signed up, you can select the CPU you are going to use in your machine, upload your model and data, and test your AI model online. This testing provides real-life performance data so that you can see if the processors you're intending to use will give you the results your business needs. 

This access makes it easy to transition your AI models from testing to inference – and to make that move on a hardware platform that gives your business the processing performance power it can trust.

DevCloud has another crucial benefit – it includes written tutorials to get you started. Provided in the form of Jupyter notebooks, these tutorials include a series of code tips that help and guide developers. These tutorials make it much easier to start writing great code. That's a huge benefit for the people developing AI solutions on Intel hardware.

DevCloud is remote, it's secure and it allows your business to ensure its AI models will scale in a cost-effective manner. Developers get immediate access to the most current inferencing technologies Intel has to offer. Key features include:

  • The latest Intel hardware, including Intel Core processors, Intel Xeon Scalable processors, Intel IoT developer kits and Intel FPGAs

  • Intel-optimized frameworks – the latest version of OpenVINO

  • Key software tools, such as Intel Parallel Studio XE Cluster Edition, Intel System Studio, and Intel MKL-DNN

  • 50 GB of file storage

Dedicated and always available: What makes DevCloud unique

If you don't use DevCloud, the only other option is to test your models on on-premise hardware. That requires a heavy capital commitment in terms of both the hardware you will need to either rent or buy and the software stack that will be required to test your models.

Using virtual machines isn't a sensible alternative, either. If you test your models on virtual machines, you'll only prove how your model runs and behaves in a virtualised environment, rather than on real hardware.

DevCloud gives you the opportunity to run your performance tests on a dedicated machine. You have online access to the same processors that would be available if you actually had the Intel hardware sitting on your desk. 

By using DevCloud and OpenVINO in combination, you can test your optimised models. You can develop your application and codebase utilising OpenVINO, then try it on different levels of Intel hardware using DevCloud. 

So what benefits does all this unique capability bring to your development work and your business? There are four key benefits from DevCloud: 

  1. Gain access to complete code walkthroughs and tutorials. With code samples to try, even before you bring your own model to DevCloud, developers have everything they need to get testing quickly.

  2. Prototype on the latest hardware and software to future-proof your solution. Ensure performance, throughput and latency requirements are met, and eliminate hardware-choice paralysis due to the plethora of in-market platforms. 

  3. Develop faster, deploy quicker. Test and benchmark AI models across multiple platforms remotely in a real-life environment on a dedicated machine. You can see test performance anywhere in the world using DevCloud – all you need is an internet connection. 

  4. Try before you buy, so you can: prevent overspending by finding the right hardware to meet your business' metrics; and reduce development time and costs by being able to scale your AI solution across a wide range of use cases. 

In summary

Intel DevCloud is the perfect complementary solution for developers who've decided to run their AI applications on Intel hardware. DevCloud makes it easy for you to test, choose and validate your hardware solution.

To find out more about AI & DevCloud, register for an upcoming webinar