Modern LLM serving workloads are increasingly heterogeneous, involving a growing portfolio of models with vastly different compute and memory requirements. Existing approaches to model serving-ranging from static GPU partitioning to dynamic reconfiguration and GPU multiplexing—fail to effectively support heterogeneity.

Ensuring order-fairness in distributed data management systems deployed in untrustworthy environments is crucial to prevent adversarial manipulation of transaction ordering, particularly in unpredictable markets where transaction order directly influences financial outcomes. While Byzantine Fault-Tolerant (BFT) consensus protocols guarantee safety and liveness, they inherently lack mechanisms to enforce order-fairness, exposing distributed systems to attacks such as frontrunning and sandwiching. Previous attempts to integrate order-fairness have often introduced substantial performance overhead, largely due to limitations of the underlying consensus protocols. This paper presents DAG of DAGs (DoD), a high-performance order-fairness protocol designed on top of DAG-based BFT consensus protocols. By leveraging the high throughput and resilience of DAG-based protocols, DoD addresses the performance limitations of existing order-fairness solutions. DoD's novel DAG of DAGs architecture enables seamless integration of order fairness with BFT consensus protocols. Through concurrent block proposals and a wave-based leader election mechanism, DoD significantly improves resilience against adversarial manipulation. A prototype implementation and experimental evaluation demonstrate that DoD effectively provides order fairness with minimal performance overhead.

Recent proposals for the disaggregation of compute, memory, storage, and accelerators in data centers promise substantial operational benefits. Unfortunately, for resources like memory, this comes at the cost of performance overhead due to the potential insertion of network latency into every load and store operation. This effect is particularly felt by data-intensive systems due to the size of their working sets, the frequency at which they need to access memory, and the relatively low computation per access. This performance impairment offsets the elasticity benefit of disaggregated memory. This paper presents TELEPORT, a compute pushdown framework for data-intensive systems that run on disaggregated architectures; compared to prior work on compute pushdown, TELEPORT is unique in its efficiency and flexibility. We have developed optimization prin- ciples for several popular systems including a columnar in-memory DBMS, a graph processing system, and a MapReduce system. The evaluation results show that using TELEPORT to push down simple operators improves the performance of these systems on state-of-the-art disaggregated OSes by an order of magnitude, thus fully exploiting the elasticity of disaggregated data centers.