{
  "annotation_id": "d4042904-dc9b-4132-b669-370e02ed2167",
  "date_created": "2026-02-17T05:53:08.515000Z",
  "date_modified": "2026-02-17T05:53:08.515000Z",
  "file_hash": "1e0d0bf832683d95dcd382353a3ebade3d8bfb9110710e628dd18feccd686008",
  "private": false,
  "record": {
    "abstract": "Recent computational advances enable protein design pipelines to run end-to-end on GPUs, yet their heterogeneous computational behaviors remain undercharacterized at the system level. We implement and profile a representative pipeline at both component and full-pipeline granularities across varying inputs and hyperparameters. Our characterization identifies generally low GPU utilization and high sensitivity to sequence length and sampling strategies. We outline future research directions based on these insights and release an open-source pipeline and profiling scripts to facilitate further studies.",
    "arxiv_id": "2601.06885",
    "authors": [
      "Jinwoo Hwang",
      "Yeongmin Hwang",
      "Tadiwos Meaza",
      "Hyeonbin Bae",
      "Jongse Park"
    ],
    "categories": [
      "cs.ET",
      "cs.SY",
      "eess.SY"
    ],
    "doi": "10.1109/LCA.2025.3646250",
    "journal_ref": "IEEE Computer Architecture Letters, vol. 25, no. 1, pp. 9-12, 2026",
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "title": "Understanding the Performance Behaviors of End-to-End Protein Design Pipelines on GPUs",
    "url": "https://arxiv.org/abs/2601.06885",
    "version": "v1"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "8e1c6e81-63e8-4849-b285-54be15197a46",
    "id": "arXiv Dataset",
    "type": "Model",
    "variant": "snapshot-2026-01-17",
    "version": "0.1.0"
  },
  "user_id": 1000002
}