Affiliation:
1. Microsoft
2. Microsoft and University of Texas at Austin
3. Microsoft and Amazon Web Services
4. Microsoft and Columbia University
5. Microsoft and Georgia Institute of Technology
6. Microsoft and University of Washington
7. Google, Inc.
8. Microsoft and École Polytechnique Fédérale de Lausanne (EPFL)
Abstract
Datacenter workloads demand high computational capabilities, flexibility, power efficiency, and low cost. It is challenging to improve all of these factors simultaneously. To advance datacenter capabilities beyond what commodity server designs can provide, we have designed and built a composable, reconfigurablefabric to accelerate portions of large-scale software services. Each instantiation of the fabric consists of a 6x8 2-D torus of high-end Stratix V FPGAs embedded into a half-rack of 48 machines. One FPGA is placed into each server, accessible through PCIe, and wired directly to other FPGAs with pairs of 10 Gb SAS cables
In this paper, we describe a medium-scale deployment of this fabric on a bed of 1,632 servers, and measure its efficacy in accelerating the Bing web search engine. We describe the requirements and architecture of the system, detail the critical engineering challenges and solutions needed to make the system robust in the presence of failures, and measure the performance, power, and resilience of the system when ranking candidate documents. Under high load, the largescale reconfigurable fabric improves the ranking throughput of each server by a factor of 95% for a fixed latency distribution--- or, while maintaining equivalent throughput, reduces the tail latency by 29%
Publisher
Association for Computing Machinery (ACM)
Reference29 articles.
1. Leap scratchpads
2. Maxwell - a 64 FPGA Supercomputer;Baxter R.;Engineering Letters,2008
Cited by
78 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. An Introduction to the Compute Express Link (CXL) Interconnect;ACM Computing Surveys;2024-07-08
2. SSR: Spatial Sequential Hybrid Architecture for Latency Throughput Tradeoff in Transformer Acceleration;Proceedings of the 2024 ACM/SIGDA International Symposium on Field Programmable Gate Arrays;2024-04
3. Memory Scraping Attack on Xilinx FPGAs: Private Data Extraction from Terminated Processes;2024 Design, Automation & Test in Europe Conference & Exhibition (DATE);2024-03-25
4. Post-configuration Activation of Hardware Trojans in FPGAs;Journal of Hardware and Systems Security;2024-03-13
5. Data Motion Acceleration: Chaining Cross-Domain Multi Accelerators;2024 IEEE International Symposium on High-Performance Computer Architecture (HPCA);2024-03-02