Sustainable ai: Environmental implications, challenges and opportunities
CJ Wu, R Raghavendra, U Gupta… - Proceedings of …, 2022 - proceedings.mlsys.org
This paper explores the environmental impact of the super-linear growth trends for AI from a
holistic perspective, spanning Data, Algorithms, and System Hardware. We characterize the …
holistic perspective, spanning Data, Algorithms, and System Hardware. We characterize the …
{FIRM}: An intelligent fine-grained resource management framework for {SLO-Oriented} microservices
User-facing latency-sensitive web services include numerous distributed,
intercommunicating microservices that promise to simplify software development and …
intercommunicating microservices that promise to simplify software development and …
Sage: practical and scalable ML-driven performance debugging in microservices
Cloud applications are increasingly shifting from large monolithic services to complex
graphs of loosely-coupled microservices. Despite the advantages of modularity and …
graphs of loosely-coupled microservices. Despite the advantages of modularity and …
ThunderGP: HLS-based graph processing framework on FPGAs
FPGA has been an emerging computing infrastructure in datacenters benefiting from
features of fine-grained parallelism, energy efficiency, and reconfigurability. Meanwhile …
features of fine-grained parallelism, energy efficiency, and reconfigurability. Meanwhile …
Understanding data storage and ingestion for large-scale deep recommendation model training: Industrial product
Datacenter-scale AI training clusters consisting of thousands of domain-specific accelerators
(DSA) are used to train increasingly-complex deep learning models. These clusters rely on a …
(DSA) are used to train increasingly-complex deep learning models. These clusters rely on a …
{AWARE}: Automate workload autoscaling with reinforcement learning in production cloud systems
Workload autoscaling is widely used in public and private cloud systems to maintain stable
service performance and save resources. However, it remains challenging to set the optimal …
service performance and save resources. However, it remains challenging to set the optimal …
Cornflakes: Zero-copy serialization for microsecond-scale networking
D Raghavan, S Ravi, G Yuan, P Thaker… - Proceedings of the 29th …, 2023 - dl.acm.org
Data serialization is critical for many datacenter applications, but the memory copies
required to move application data into packets are costly. Recent zero-copy APIs expose …
required to move application data into packets are costly. Recent zero-copy APIs expose …
A hardware accelerator for protocol buffers
Serialization frameworks are a fundamental component of scale-out systems, but introduce
significant compute overheads. However, they are amenable to acceleration with …
significant compute overheads. However, they are amenable to acceleration with …
A cloud-scale characterization of remote procedure calls
The global scale and challenging requirements of modern cloud applications have led to the
development of complex, widely distributed, service-oriented applications. One enabler of …
development of complex, widely distributed, service-oriented applications. One enabler of …
Deepscaling: microservices autoscaling for stable cpu utilization in large scale cloud systems
Cloud service providers conservatively provision excessive resources to ensure service
level objectives (SLOs) are met. They often set lower CPU utilization targets to ensure …
level objectives (SLOs) are met. They often set lower CPU utilization targets to ensure …