Confidential High-Performance Computing in the Public Cloud
نویسندگان
چکیده
High-performance computing (HPC) in the public cloud democratizes supercomputing power that most users cannot afford to purchase and maintain. Researchers have studied its viability, performance, usability. However, HPC has a unique feature—users export data computation somewhat untrusted platforms. Users will either fully trust providers protect from all kinds of attacks or keep sensitive assets in-house instead. With recent deployment trusted execution environment (TEE) cloud, confidential for is becoming practical addressing users’ privacy concerns. This article discusses threat models, challenges, possible solutions, significant gaps, focusing on TEE-based computing. We hope this discussion improve understanding new topic promote research directions.
منابع مشابه
Performance Analysis of Public Cloud Computing Providers
The objective of this paper is to perform a comprehensive performance comparison of public cloud services for computing and to analyze the correlation between their prices and performance. Eight representative public cloud providers were divided into two groups using market share: small cloud providers and large cloud providers. Results revealed that these offered computing services vary widely...
متن کاملThe impact of Cloud Computing in the banking industry resources
Today, one of the biggest problems that gripped the banking sphere, the high cost of implementing advanced technologies and the efficient use of the hardware. Cloud computing is the use of shared services on the Internet provides a large role in developing the banking system, without the need for operating expenses including staffing, equipment, hardware and software Reducing the cost of implem...
متن کاملThe State of High Performance Computing in the Cloud
HPC applications have been gaining lot of attention in the cloud computing world. Most of these applications are scientific applications that require large CPU capabilities and are also data intensive requiring large data storage. Traditionally they have always required large number of computers inter-connected in a network such as clusters or supercomputers. These clusters are difficult to set...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Internet Computing
سال: 2023
ISSN: ['1089-7801', '1941-0131']
DOI: https://doi.org/10.1109/mic.2022.3226757