{"title":"Introduction to the Special Section on Security in FPGA-accelerated Cloud and Datacenters","authors":"C. Bobda, R. Tessier, Ken Eguro, R. Kastner","doi":"10.1145/3352060","DOIUrl":null,"url":null,"abstract":"The rapid advance of cloud computing platforms has made these computing resources a vital infrastructure for many application developers. For a modest fee, scalable and diverse compute components are available to application developers on demand, eliminating the need for large computer hardware investments by organizations. Although cloud computing platforms have included microprocessors and graphics processing units (GPUs) for many years, the availability of field-programmable gate arrays (FPGAs) in these platforms has only become prevalent in the past few years. Amazon, Baidu, and Maxeler now expose FPGAs to application developers in their cloud infrastructures. The integration of FPGAs in Microsoft Catapult to accelerate various tasks, including Bing, has led to a 2× performance speed-up versus processor-only implementation with only a 30% increase in energy. Intel recently announced in-package FPGA integration in Xeon multi-core processors. The use of FPGAs in the cloud raises a series of important security issues regarding their use, since the devices have traditionally been used by a single user in a closed environment. FPGA use by many cloud users over time, and potentially by multiple independent users at the same time, opens up a number of attack vectors in which FPGA devices could be damaged, computation could be attacked leading to incorrect results, computation results could be snooped, or covert communication channels in the FPGA could be developed. Research identifying these issues and developing countermeasures is still at an early stage, although interest is increasing. To highlight early work in these areas, we initiated a call for a special issue of TRETS with the topic of security in FPGA-accelerated cloud and datacenters. After an initial evaluation of submissions for the special issue for quality and relevance, eight papers were selected for review. At least three reviewers who are experts in security and FPGAs evaluated each manuscript during multiple rounds of review. Ultimately, four high-quality papers related to the security of FPGAs in the cloud were selected for inclusion in the special issue. The following summaries provide high-level views of these manuscripts and a brief analysis of their contributions:","PeriodicalId":162787,"journal":{"name":"ACM Transactions on Reconfigurable Technology and Systems (TRETS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"ACM Transactions on Reconfigurable Technology and Systems (TRETS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3352060","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The rapid advance of cloud computing platforms has made these computing resources a vital infrastructure for many application developers. For a modest fee, scalable and diverse compute components are available to application developers on demand, eliminating the need for large computer hardware investments by organizations. Although cloud computing platforms have included microprocessors and graphics processing units (GPUs) for many years, the availability of field-programmable gate arrays (FPGAs) in these platforms has only become prevalent in the past few years. Amazon, Baidu, and Maxeler now expose FPGAs to application developers in their cloud infrastructures. The integration of FPGAs in Microsoft Catapult to accelerate various tasks, including Bing, has led to a 2× performance speed-up versus processor-only implementation with only a 30% increase in energy. Intel recently announced in-package FPGA integration in Xeon multi-core processors. The use of FPGAs in the cloud raises a series of important security issues regarding their use, since the devices have traditionally been used by a single user in a closed environment. FPGA use by many cloud users over time, and potentially by multiple independent users at the same time, opens up a number of attack vectors in which FPGA devices could be damaged, computation could be attacked leading to incorrect results, computation results could be snooped, or covert communication channels in the FPGA could be developed. Research identifying these issues and developing countermeasures is still at an early stage, although interest is increasing. To highlight early work in these areas, we initiated a call for a special issue of TRETS with the topic of security in FPGA-accelerated cloud and datacenters. After an initial evaluation of submissions for the special issue for quality and relevance, eight papers were selected for review. At least three reviewers who are experts in security and FPGAs evaluated each manuscript during multiple rounds of review. Ultimately, four high-quality papers related to the security of FPGAs in the cloud were selected for inclusion in the special issue. The following summaries provide high-level views of these manuscripts and a brief analysis of their contributions: