Intersect360 Research has released a comprehensive new report concerning the challenges facing the combined fields of high-performance computing (HPC) and artificial intelligence (AI). Titled “Issues Facing the HPC-AI Industry: Insights from the Advisory Committees of the HPC-AI Leadership Organization (HALO),” this report – which was based on extensive interviews conducted with members of the HALO Advisory Committee – details the industry’s current landscape and future directions of HPC and AI technologies.
While the full report is 22 pages long and complete with figures and graphs, certain information and opinions stand out.
To begin, there is currently a lack of people who are trained and ready to handle new HPC-AI environments. Intersect360 Research interviews found that most respondents expressed concerns about a shortage of personnel, particularly in the computational sciences.
This problem stems from and is exacerbated by educational programs that do not emphasize computational sciences. Both computational/engineering sciences as well as the design and operation of large-scale computational infrastructures suffer from this shortage of knowledgeable professionals.
Additionally, interviewees remarked that while GPU-heavy deployments often used in AI are capable of delivering a large number of theoretical peak FLOPS, many traditional HPC applications do not scale well on large numbers of GPUs. This finding is consistent with broader Intersect360 Research studies, which found that 12% of HPC-AI users report “moderate” or “major” performance issues with their HPC-AI systems, while only 47% report “no problems.”
Accuracy and the reproducibility of results were also both major points of contention within the report. The variety of chip technologies – including CPUs, GPUs, NPUs, FPGAs, and custom chips – is introducing complexity in HPC-AI environments that can lead to an inability to achieve consistent results and performance across different architectures. Diverse chip technologies and software stacks also create issues with independent verification of research, thereby requiring a large investment of resources to replicate results across different HPC-AI environments.
Just as varied as current chip technologies are the legal restrictions concerning data sharing across different countries. The report mentions that the EU and China currently have the strictest regulations, but even the U.S. has HIPAA requirements that restrict patient data exchange. Global collaboration is therefore difficult with such varying restrictions.
Similarly, countries are beginning to invest in the idea of “HPC nationalism.” This term refers to the tendency of countries to develop their own HPC technologies, infrastructure, and expertise to achieve technological independence or superiority. Again, international cooperation is impacted when countries have vested interests in developing their own processor chips, systems, and software stacks.
This idea of HPC nationalism reflects concerns that AI/LLMs may homogenize customs and languages across different cultural or ethnic groups. In an age of rapid globalization, what impact will AI have on cultural preservation and the maintenance of linguistic diversity?
The report also mentions the increasing importance of sustainability. HPC-AI facilities necessarily consume large amounts of power, specifically with GPU-heavy architectures. This trend creates environmental impacts and increases energy costs, and companies are looking for ways to cut back. Organizations may need to consider outsourcing to external facilities like cloud providers, hyperscale companies, or co-location centers to avoid rising energy costs. This situation, therefore, will largely change how HPC-AI infrastructure is managed at a global level.
Of course, this is only a glimpse into the full report, which is now available online for Intersect360 Research Clients and HALO members. End users interested in this work can apply join HALO at no cost at https://hpcaileadership.org. This report will also be discussed alongside an upcoming update to the Intersect360 Research HPC-AI market forecast in a webinar for clients and HALO members available in multiple time zones on November 4-5, 2024. Information about the webinar is available at https://www.intersect360.com/webinar/.