{
  "id": 14720835,
  "title": "When faster supercomputers aren’t enough",
  "url": "https://urgent.news/2026/10/15/when-faster-supercomputers-arent-enough",
  "topic": "science",
  "section": "Science",
  "published": "2026-10-15T15:00:00.000Z",
  "source": {
    "name": "The Register Science",
    "slug": "the-register-science",
    "url": "https://www.theregister.com/hpc/2026/10/15/partner-content-when-faster-supercomputers-arent-enough/5302562"
  },
  "original_language": "en",
  "account": "When a million veterans contributed their genetic data to the Department of Veterans Affairs' \"Million Veteran Program,\" they did so without expecting anything in return. Ravi Madduri, a senior scientist at Argonne National Laboratory, is committed to delivering the benefits of this data to the veterans who donated it. Madduri, who also works at the University of Chicago, believes that the data should be used to benefit anyone who can gain insights from it.\n\nMadduri has dedicated his career to the intersection of supercomputing and biomedicine. Over the past two decades, he has been working on building the computing power necessary to turn the veterans' data into results that impact various health conditions such as post-traumatic stress disorder, heart disease, and cancer. While the machines have kept up with their end, there are still challenges that cannot be solved by faster machines.\n\nUnderstanding these obstacles is crucial for advancing scientific research. One of the main challenges lies in the vast amount of data involved. Two human genomes differ at millions of specific positions, which creates a massive dataset for analysis. Running a genome-wide association study requires solving a seemingly simple question across an enormous volume of data.\n\nHowever, the real challenge lies in the coding and computational architecture of the algorithms used for analysis. Madduri's team discovered that the original code was written for ordinary processors, which became a bottleneck as the dataset grew. The team found that by rewriting the code to take advantage of graphics processors (GPUs), they could significantly speed up the analysis. This rewrite transformed an analysis that would have taken 481,000 CPU-core hours into something more manageable.\n\nThe study, published in Science and Bioinformatics, analyzed the genetic data of 635,969 veterans and identified 13,672 regions of the genome linked to one or more conditions. About 1,600 of these regions were previously unknown, largely due to the inclusion of veterans from diverse ancestral backgrounds. This finding challenges the traditional approach of genetic studies, which often focuses on people of European descent. The newly discovered associations are only visible when the separate analyses for different ancestral groups are combined.\n\nMadduri emphasizes that this is still a preclinical phase of research, meant for discovery purposes. The ultimate goal is to create a statistical map of where to look for potential genetic markers. However, the results themselves do not pave the way for direct medical interventions.",
  "summary": "PARTNER CONTENT: A million reasons why advanced computing matters for genomics",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 2,
    "also_reported_by": [
      {
        "outlet": "The Register",
        "title": "When faster supercomputers aren’t enough",
        "url": "https://urgent.news/2026/10/15/when-faster-supercomputers-arent-enough-14723115",
        "published": "2026-10-15T15:00:00.000Z"
      }
    ]
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}