AMDAmd, a global industry leader in computing technology, recently dropped some amazing news. The company is dedicated to advancing software engineering practices through the adoption of artificial intelligence (AI). This new strategic direction represents a shift in priorities towards increasing productivity. It helps internal workflows at the company, too—particularly when creating the Adrenalin graphics-software package. Software engineers are spending 80% or more of their time doing manual work such as coding, testing and debugging. This effort comes just in time as they are confronted with these new demands.
Software engineers at AMD now spend about 40 percent of their time just writing new code and testing it. On top of that, they commit 10 percent of their time to learning new technologies and another 20 percent to triaging and debugging other people’s problems. It focuses their engineers’ time—almost 20 percent of it—on revising their own code to make it better. They spend upwards of 10 percent trying to document it. These statistics only begin to peel back the layers of this complex and dangerous software engineering dynamic. They further point out that AI tools can greatly reduce the time-consuming processes.
The Structure of Software Engineering at AMD
At AMD, software engineers make up almost 50 percent of the workforce. This serves to underscore their foundational function in guiding product innovation. Even more so than Nvidia’s Geforce drivers, the typical release cycle for a new Adrenalin graphics-software package averages out to about six months’ time. After this period, the stewardship phase can last at least three to six months. Over the course of this timeline, many teams work together to make sure these software solutions are developed, tested, and ultimately supported in the real world.
The development phase is represented by their core team of about 60 engineers. That same team then leads the creation of the project. They do extensive Quality Assurance testing to ensure the quality of the final product is consistently high. At the same time, the support phase keeps 20 engineers busy fixing issues that pop up after release, ensuring a steady flow of maintenance and improvements. In addition, a smaller group of engineers is involved in the design phase (10 engineers) and the definition phase (5 engineers), allowing for a comprehensive approach to software development.
AI’s Impact on Efficiency and Productivity
Integrating AI into AMD’s software engineering processes has the potential to deliver big efficiencies. For example, using AI at the functional testing phase has reduced new manual test creation by 15 percent. This allows engineers to spend time working on more impactful activities, as opposed to redundant testing processes. AI increases the range of scenarios that can be tested by 20-fold. This improvement allows a more comprehensive look at software’s security posture prior to being released.
Increased overall team Productivity AMD anticipates an increase in overall team productivity by greater than 25 percent. This breakthrough will happen only after they embed AI tools deeply into their work processes. This tsunami of efficiency super-charged AMD’s engineering teams. Beyond this economic argument, it creates more efficient, higher quality and faster software delivery to customers. By automating repetitive processes through the use of AI, AMD can continue to find itself on the bleeding edge of technology that changes every minute.
Future Prospects for AI in Software Engineering
For AMD, being able to integrate AI capabilities stands to dramatically enhance its productivity and operations. The business is passionate about embracing smart new solutions that increase levels of teamwork and efficiency. AI has the potential to do so much more than help tests run more efficiently. This opens the door for improved code optimization, more efficient debugging workflows, and increased documentation efforts.
The ongoing developments within AMD’s software engineering department serve as a testament to the company’s forward-thinking approach. AMD has pivoted to address AI trends through a current operational lens. This decision is an important step toward bigger changes in the field of software development and safety practices.