본문으로 이동

속성:Applicability

S3 연구 메모리

Text

이 메모가 맞는 조건과 예외를 적습니다.

( | ) (20 | 50 | 100 | 250 | 500) 보기
이 속성을 사용하는 문서 20개를 보여줍니다.
d
dlora dynamically orchestrating requests and adapters for lora llm serving 3827bfa3 +
Multi-tenant LoRA serving with many adapters and skewed arrivals. Limits: Performance depends on adapter popularity, merge/migration cost, model size, and workload dynamics.  +
dram 8d56adf1 +
작은 SRAM만 가진 UFS·모바일 플래시 컨트롤러의 garbage collection 및 유효 페이지 판별.  +
dram architecture for efficient data lifetime management f27c8820 +
Cold/warm boot 또는 deallocated-page disclosure를 막아야 하는 DRAM 설계에 해당한다.  +
dwkv collinear slo confound +
SLO scheduling, ablation design, causal policy comparison, simulator workload generation  +
dwkv doomed chase domino +
deadline scheduling, overload control, best-effort admission, heuristic priority queues  +
dwkv failed run fail closed +
experiment runners, benchmark data quality, failure recovery, reproducibility audits  +
dwkv kv saturation unproven +
GPU serving profiling, KV-cache scheduling, simulator validation, bottleneck attribution  +
dwkv legacy patch not repro +
research code archival, patch-based reproduction, vLLM fork maintenance, provenance review  +
dwkv metric mismatch +
scheduler objective design, multi-objective evaluation, SLO guardrails, paper claim review  +
dwkv prompt seed bug +
LLM serving benchmark, stochastic workload replay, multi-policy paired comparison, seed-repeat 분석  +
dwkv score code contract +
scheduler implementation, ablation testing, policy attribution, code-level reproducibility  +
e
efficient footprint caching for tagless dram caches 9317d60c +
Multi-GB in-package DRAM cache와 TDC 계열 구조에 해당한다.  +
efficient function call tracing with link time binary rewriting for ce devices 0fd13b41 +
임베디드/consumer device의 low-overhead tracing과 proprietary library 성능 분석에 해당한다.  +
efficient hybrid polling for ultra low latency storage devices e753019a +
짧은 service time의 NVMe/Optane/Z-SSD와 CPU 예산이 중요한 latency-sensitive I/O에 높음.  +
efficient memory deduplication for mobile smart devices fa9b72e9 +
자원이 제한된 모바일 시스템의 background memory reclamation/deduplication에 해당한다.  +
efficient memory overcommitment for i o passthrough enabled vms via fine grained page meta data d32f4845 +
SR-IOV·passthrough VM 클라우드. Limits: 게스트·하이퍼바이저·장치 간 협조와 메타데이터 유지가 필요하다.  +
emt an os framework for new memory translation architectures 6d3356f6 +
OS/architecture co-design for alternative page tables and memory translation. Limits: Requires Linux integration and implementation work for each architecture; exact evaluation scope is abstract-limited.  +
enabling high performance and secure userspace nvm file systems with the trio architecture d59d00b6 +
NVM 기반 사용자 공간 파일시스템. Limits: NVM과 특정 handoff·신뢰 모델, 파일시스템별 핵심 상태 설계가 필요하다.  +
energy efficient scheduling of real time tasks on multicore processors a70b387a +
homogeneous multicore real-time scheduler의 energy-aware partitioning/core parking에 해당한다.  +
enhancing network i o performance for a virtualized hadoop cluster d869aae7 +
가상화된 Hadoop 또는 MapReduce cluster와 rack-aware data placement에 해당한다.  +
원본 주소 "https://s3wiki.yonsei.ac.kr/index.php?title=속성:Applicability&oldid=150"