SAGE：面向AI数据移动的语义感知地理错误恢复
SAGE: Semantic-Aware Geographic Error Recovery for AI Data Movement
- City University of Hong Kong（香港城市大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
SAGE通过语义感知的地理错误恢复架构，区分灾难性与非灾难性位错误，优化重放决策，降低AI数据移动延迟30.1%并提升语义质量。
AI中文摘要：
AI互连通常统一保护并重放数据包，然而数值位错误在后果上差异显著：低阶尾数翻转可能类似于量化噪声，而高重要性指数翻转则可能产生灾难性异常值或非有限值。我们提出SAGE，一种语义感知的地理错误恢复架构，它将检测到的故障是否值得重放与从何处重放解耦。对于类似BF16的数据，一个工作负载校准的契约将灾难性的Class-H故障与有界的Class-M和精度Class-L损伤区分开来。它首先应用Class-H静默投递约束，然后根据质量归一化的终端延迟$\Psi_{\rm del}$对允许的策略进行排序。独立地，一个源本地区域表根据故障地理调整检查点间隔，缩短噪声区域中的恢复段。检测到的Class-H故障可能触发受保护的否定确认和全微片重放；Class-M和Class-L结果不触发默认网络重放。我们在gem5 Garnet中实现了SAGE的端点和重放协议，并在ASAP7中合成了其全流水线检查器。在一个稳定的合成操作点，十种子竞争保真直接Garnet活动表明，相对于固定的34跳恢复，SAGE将$\Psi_{\rm del}$降低了30.1%，结合了28.0%更低的平均延迟和改善的投递语义质量。在相同提供负载下的更高误码率合成压力下，SAGE保持有界队列，而固定基线累积积压。应用衍生的DeiT-S通信轨迹也显示出在评估的非零误码率下更低的平均和p99.5延迟。在合格的操作包络内，CRC32解码器试验得出每个原始Class-H静默投递的95%上限为$3.18\times10^{-7}$。
英文摘要：
AI interconnects typically protect and replay packets uniformly, yet numerical bit faults differ sharply in consequence: a low-order mantissa flip may resemble quantization noise, while a high-significance exponent flip can produce a catastrophic outlier or non-finite value. We present SAGE, a semantic-aware geographic error-recovery architecture that decouples whether a detected fault merits replay from where replay restarts. For BF16-like data, a workload-calibrated contract separates catastrophic Class-H faults from bounded Class-M and precision Class-L damage. It first applies a Class-H silent-delivery constraint, then ranks admissible policies by quality-normalized terminal latency, $Ψ_{\rm del}$. Independently, a source-local region table adapts checkpoint intervals to fault geography, shortening recovery segments in noisy regions. Detected Class-H failures may trigger protected negative acknowledgments and full-flit replay; Class-M and Class-L outcomes do not trigger default network replay. We implement SAGE's endpoint and replay protocol in gem5 Garnet and synthesize its fully pipelined checker in ASAP7. At a stable synthetic operating point, a ten-seed contention-faithful direct-Garnet campaign shows that SAGE reduces $Ψ_{\rm del}$ by 30.1% relative to fixed 34-hop recovery, combining 28.0% lower mean latency with improved delivered semantic quality. Under higher-BER synthetic stress at the same offered load, SAGE maintains bounded queues while the fixed baseline accumulates backlog. Application-derived DeiT-S communication traces also show lower mean and p99.5 latency at the evaluated nonzero BERs. Within the qualified operating envelope, CRC32 decoder trials yield a simultaneous 95% per-original Class-H silent-delivery upper bound of $3.18\times10^{-7}$.
