Marcos Kawazoe Aguilera
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Marcos Kawazoe Aguilera has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
- Rachid Guerraoui
EPFL
3 shared publications - Antoine Murat
École Polytechnique Fédérale de Lausanne EPFL) Switzerland
2 shared publications - Athanasios Xygkis
Oracle Labs, Zürich, CH
2 shared publications - Igor Zablotchi
Mysten Labs, Zürich, CH
2 shared publications - Naama Ben-David
Technion, Haifa, Israel
2 shared publications Sam Toueg
University of Toronto
2 shared publications
Latest Publications
- Open Access
- 28 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 229-246
Disk-Based LSMs: An Unexpectedly Good Index for Partly Coherent CXL Memory
- 28 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 229-246
Open AccessDisk-Based LSMs: An Unexpectedly Good Index for Partly Coherent CXL Memory
- Open Access
- 6 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 325-328
Brief Announcement: Tiered Memory Computation
- 6 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 325-328
Open AccessBrief Announcement: Tiered Memory Computation
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1413-1429
Performance Predictability in Heterogeneous Memory
Heterogeneous memory combining DRAM and CXL exhibits variable performance, yet existing metrics correlate weakly with actual slowdown. We present CAMP, a principled framework for predicting CXL-induced slowdown. Our key insight is that a DRAM run (plus a CXL run for bandwidth-bound workloads) exposes the causal microarchitectural pressure points where CXL latency translates into additional processor stall cycles. CAMP captures these signals using 12 performance counters to analytically decompose slowdown into three orthogonal components: demand reads, cache/prefetching, and stores. CAMP also introduces a closed-form model for software-based weighted interleaving that predicts performance across DRAM--CXL ratios. Across 265 workloads on NUMA and three CXL devices, CAMP achieves 91--97% prediction accuracy within 10% absolute error. We demonstrate that these models enable practical system policies, including ''Best-shot'' interleaving and colocated workload placement, improving performance by up to 21% and 23% over existing tiering and colocation approaches.
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1413-1429
Open AccessPerformance Predictability in Heterogeneous Memory
Heterogeneous memory combining DRAM and CXL exhibits variable performance, yet existing metrics correlate weakly with actual slowdown. We present CAMP, a principled framework for predicting CXL-induced slowdown. Our key insight is that a DRAM run (plus a CXL run for bandwidth-bound workloads) exposes the causal microarchitectural pressure points where CXL latency translates into additional processor stall cycles. CAMP captures these signals using 12 performance counters to analytically decompose slowdown into three orthogonal components: demand reads, cache/prefetching, and stores. CAMP also introduces a closed-form model for software-based weighted interleaving that predicts performance across DRAM--CXL ratios. Across 265 workloads on NUMA and three CXL devices, CAMP achieves 91--97% prediction accuracy within 10% absolute error. We demonstrate that these models enable practical system policies, including ''Best-shot'' interleaving and colocated workload placement, improving performance by up to 21% and 23% over existing tiering and colocation approaches.
- Open Access
- 17 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 131-138
Lost in Translation: The Search for Meaning in Network-Attached AI Accelerator Disaggregation
- 17 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 131-138
Open AccessLost in Translation: The Search for Meaning in Network-Attached AI Accelerator Disaggregation
- Open Access
- 13 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 1
Keynote: Disaggregated Memory and the Revival of Memory Research
- 13 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 1
Open AccessKeynote: Disaggregated Memory and the Revival of Memory Research
- Open Access
- 14 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 80-87
Real Life Is Uncertain. Consensus Should Be Too!
- 14 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 80-87
Open AccessReal Life Is Uncertain. Consensus Should Be Too!
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 24-45
SWARM: Replicating Shared Disaggregated-Memory Data in No Time
Memory disaggregation is an emerging data center architecture that improves resource utilization and scalability. Replication is key to ensure the fault tolerance of applications, but replicating shared data in disaggregated memory is hard. We propose SWARM (Swift WAit-free Replication in disaggregated Memory), the first replication scheme for in-disaggregated-memory shared objects to provide (1) single-roundtrip reads and writes in the common case, (2) strong consistency (linearizability), and (3) strong liveness (wait-freedom). SWARM makes two independent contributions. The first is Safe-Guess, a novel wait-free replication protocol with single-roundtrip operations. The second is In-n-Out, a novel technique to provide conditional atomic update and atomic retrieval of large buffers in disaggregated memory in one roundtrip. Using SWARM, we build SWARM-KV, a low-latency, strongly consistent and highly available disaggregated key-value store. We evaluate SWARM-KV and find that it has marginal latency overhead compared to an unreplicated key-value store, and that it offers much lower latency and better availability than FUSEE, a state-of-the-art replicated disaggregated key-value store.
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 24-45
Open AccessSWARM: Replicating Shared Disaggregated-Memory Data in No Time
Memory disaggregation is an emerging data center architecture that improves resource utilization and scalability. Replication is key to ensure the fault tolerance of applications, but replicating shared data in disaggregated memory is hard. We propose SWARM (Swift WAit-free Replication in disaggregated Memory), the first replication scheme for in-disaggregated-memory shared objects to provide (1) single-roundtrip reads and writes in the common case, (2) strong consistency (linearizability), and (3) strong liveness (wait-freedom). SWARM makes two independent contributions. The first is Safe-Guess, a novel wait-free replication protocol with single-roundtrip operations. The second is In-n-Out, a novel technique to provide conditional atomic update and atomic retrieval of large buffers in disaggregated memory in one roundtrip. Using SWARM, we build SWARM-KV, a low-latency, strongly consistent and highly available disaggregated key-value store. We evaluate SWARM-KV and find that it has marginal latency overhead compared to an unreplicated key-value store, and that it offers much lower latency and better availability than FUSEE, a state-of-the-art replicated disaggregated key-value store.
- Open Access
- 28 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 25-32
Logical Memory Pools
We propose logical memory pools, a memory disaggregation architecture for the emerging Compute Express Link (CXL) technology in datacenters. The key idea is to create a memory pool by carving out parts of the local memory in each server, rather than using a physical memory pool that is separate from servers. Logical pools provide significant benefits over physical pools, namely, lower cost, support for near-memory computing without extra hardware, and flexibility on designating whether memory is part of the memory pool or not. We demonstrate that logical pools can execute workloads that are unfeasible in physical pools, and that its faster access leads to better performance. Realizing logical memory pools poses five major challenges, which we believe can be overcome. Given the benefits of logical pools, we believe the CXL community should refocus efforts on logical, rather than physical memory pools.
- 28 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 25-32
Open AccessLogical Memory Pools
We propose logical memory pools, a memory disaggregation architecture for the emerging Compute Express Link (CXL) technology in datacenters. The key idea is to create a memory pool by carving out parts of the local memory in each server, rather than using a physical memory pool that is separate from servers. Logical pools provide significant benefits over physical pools, namely, lower cost, support for near-memory computing without extra hardware, and flexibility on designating whether memory is part of the memory pool or not. We demonstrate that logical pools can execute workloads that are unfeasible in physical pools, and that its faster access leads to better performance. Realizing logical memory pools poses five major challenges, which we believe can be overcome. Given the benefits of logical pools, we believe the CXL community should refocus efforts on logical, rather than physical memory pools.
- Open Access
- 23 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 37-43
Limited Access: The Truth Behind Far Memory
Memory capacity in data centers is becoming a scarce resource. To address this issue, emerging runtimes enable applications to supplement their local memory with additional tiers of compressed, non-volatile, or far memory, often accessed via OS-supported paging. In these systems, minimizing page faults is crucial for good performance. Yet, there is little common understanding of which parts of application code are responsible for triggering page faults. In this paper, we analyze page-fault behavior across a suite of 26 applications and find that the vast majority of page faults are triggered by a very small number of lines of application code. In the light of this and related observations, we discuss the feasibility of several ways to reduce page faults.
- 23 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 37-43
Open AccessLimited Access: The Truth Behind Far Memory
Memory capacity in data centers is becoming a scarce resource. To address this issue, emerging runtimes enable applications to supplement their local memory with additional tiers of compressed, non-volatile, or far memory, often accessed via OS-supported paging. In these systems, minimizing page faults is crucial for good performance. Yet, there is little common understanding of which parts of application code are responsible for triggering page faults. In this paper, we analyze page-fault behavior across a suite of 26 applications and find that the vast majority of page faults are triggered by a very small number of lines of application code. In the light of this and related observations, we discuss the feasibility of several ways to reduce page faults.
- Open Access
- 22 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 196-205
Unleashing True Utility Computing with Quicksand
Today's clouds are inefficient: their utilization of resources like CPUs, GPUs, memory, and storage is low. This inefficiency occurs because applications consume resources at variable rates and ratios, while clouds offer resources at fixed rates and ratios. This mismatch of offering and consumption styles prevents fully realizing the utility computing vision. We advocate for fungible applications, that is, applications that can distribute, scale, and migrate their consumption of different resources independently while fitting their availability across different servers (e.g., memory at one server, CPU at another). Our goal is to make use of resources even if they are transiently available on a server for only a few milliseconds. We are developing a framework called Quicksand for building such applications and unleashing the utility computing vision. Initial results using Quicksand to implement a DNN training pipeline are promising: Quicksand saturates resources that are imbalanced across machines or rapidly shift in quantity.
- 22 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 196-205
Open AccessUnleashing True Utility Computing with Quicksand
Today's clouds are inefficient: their utilization of resources like CPUs, GPUs, memory, and storage is low. This inefficiency occurs because applications consume resources at variable rates and ratios, while clouds offer resources at fixed rates and ratios. This mismatch of offering and consumption styles prevents fully realizing the utility computing vision. We advocate for fungible applications, that is, applications that can distribute, scale, and migrate their consumption of different resources independently while fitting their availability across different servers (e.g., memory at one server, CPU at another). Our goal is to make use of resources even if they are transiently available on a server for only a few milliseconds. We are developing a framework called Quicksand for building such applications and unleashing the utility computing vision. Initial results using Quicksand to implement a DNN training pipeline are promising: Quicksand saturates resources that are imbalanced across machines or rapidly shift in quantity.