@misc{indiciae53f47ba8352b,
  title = {When the KV Cache Is Slower Than the GPU: Scaling Remote KV Tiers for Context-Parallel Prefill},
  author = {Muhammad Ibrahim},
  year = {2026},
  url = {https://arxiv.org/abs/2610.21500},
  note = {Source identifier: 2610.21500}
}