Software engineers at AMD are going to go through quite a paradigm shift. They will use new artificial intelligence (AI) tools to unlock new levels of efficiency in their daily workflows. By automating repetitive work and improving workflows, AMD hopes to markedly improve productivity among its engineering teams. The company has zeroed in on a big pain point. Software engineers are wasting a lot of their hard-earned time recreating the wheel with activities not associated with writing code.
According to recent assessments, software engineers devote approximately 40 percent of their work hours to developing and testing new code. In stark opposition, they spend only 40 percent of their time on truly impactful tasks. This division illustrates the AI-enabled impact that is yet to come. Engineers spend almost 10 percent of their available hours keeping up with the latest technologies. They spend an additional 20 percent just triaging and debugging issues on the front end. You probably don’t spend 80 percent of your time just looking over and fine-tuning code. You reserve 10 percent for documentation, too.
The Current Landscape of Software Engineering
This troubling data about how much time software engineers spend solving industry-specific challenges is a clear indication of these issues. Engineers are required to spend only 40 percent of their time developing and experimenting with new code. This forces them to be buried in administrative tasks, which unfortunately stifles their capacity for innovation. The remaining 60 percent encompasses a range of activities—learning new skills, troubleshooting issues, refining existing code, and maintaining comprehensive documentation.
The need for lifelong learning has never been more important as technology is changing faster than ever before. For engineers, it’s about 10 percent just staying on the cutting edge. This constant cycle of learning is critical not only for individual development, but for the success and competitiveness of the organization as a whole. Triage and debugging take up a huge percentage of engineers’ time as well, showing a clear demand for better development troubleshooting tools.
About 20 percent of their code efforts go into making sure that the existing code works and runs efficiently. This robust review process is essential in avoiding even the smallest changes that could break software functionality. In addition, there’s the almost invisible documentation part of software engineering that takes up another 10 percent of engineers’ time. While proper documentation helps with sharing knowledge and developing for future, it creates a time burden of its own.
AMD’s Strategic Initiative
AMD recognizes the difficulties its engineering teams face. To address these challenges, the organization is aggressively incorporating AI tools to deliver creative solutions. So it’s critical that we optimize their productivity, just as approximately half of AMD’s workforce consists of software engineers. Their hope is that bringing AI into the fold will massively boost their teams’ productivity. They’re counting on a boost of more than 25 percent.
AI tools allow engineers to focus on more complex, dynamic work. Those steps involve training, authoring code bases, synthesizing code, testing code, and debugging code and issues. By automating repetitive tasks and providing intelligent recommendations, these tools can free up valuable time for engineers to focus on more critical aspects of development.
ETHICAL AMD urges responsible standards of practice in deploying AI tools. They say that these technologies should never violate third-party IP rights. The company has very detailed protocols designed to avoid generating anything that would infringe on intellectual property protections. This pledge further highlights AMD’s long-standing commitment to ethics in the development and deployment of AI technologies, given the pace of innovation in this space.
Team Structure and Collaboration
In laying the groundwork for this strategic push, AMD has organized its engineering resources aptly across all stages of the software lifecycle. As an example, it took about twenty different engineers to support the most recent release, everyone bringing their skills to bear. In the same way, if ten engineers worked on the design phase, then five engineers worked on determining the project’s needs.
The biggest commitment came during development and testing phases, with up to sixty engineers working at each phase. This unique collaborative approach allows for a multi-modal, holistic perspective to be incorporated into each stage of the project lifecycle. By utilizing AI tools during these phases, AMD hopes to foster even greater collaboration and create more efficient workflows.
By adding AI to their workflows, they augment humans’ expertise with technology. Advanced technologies complement the capabilities of skilled professionals, amplifying their strengths. Engineers will have new AI assistants at their disposal to make them drastically more efficient while increasing the rigor of their work.