ASIC Clouds: Specializing the Datacenter

  title={ASIC Clouds: Specializing the Datacenter},
  author={Ikuo Magaki and M. Khazraee and L. V. Gutierrez and M. Taylor},
Master of None Acceleration: A Comparison of Accelerator Architectures for Analytical Query Processing
This paper compares a previously proposed heterogeneous hardware accelerator for analytical query processing to a homogeneous systolic array alternative, finding that the heterogeneous and homogeneous accelerators are equivalent for large designs, while for small designs the homogeneous is better. Expand
QZFS: QAT Accelerated Compression in File System for Application Agnostic and Cost Efficient Data Storage
The comprehensive evaluation validates that QZFS can achieve up to 5x write throughput improvement for FIO micro-benchmark and more than 6x costefficiency enhancement for genomic data post-processing over the software-implemented alternative. Expand
Sustainability of bitcoin and blockchains
Bitcoin is an electronic currency that has become increasingly popular since its introduction in 2008. Transactions in the bitcoin system are stored in a public transaction ledger (‘the blockchain’),Expand
Scaling Datacenter Accelerators with Compute-Reuse Architectures
  • Adi Fuchs, D. Wentzlaff
  • Computer Science
  • 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA)
  • 2018
The COmpute-REuse Accelerators (COREx) architecture that shifts computations from the scalability-hindered transistor-based logic towards the continuing-to-scale storage domain, and achieves an average speedup of 6.4x and average savings of 50% in energy and 63% inEnergy-delay product. Expand
Moonwalk : NRE Optimization in ASIC Clouds or , accelerators will use old silicon
Cloud services are becoming increasingly globalized and data-center workloads are expanding exponentially. GPU and FPGA-based clouds have illustrated improvements in power and performance byExpand
Cloud-Based FPGA Custom Computing Machines for Streaming Applications
A novel platform for launching and using field-programmable gate arrays (FPFA) custom computing machines (CCMs) in clouds and data centers is proposed, which has relatively low overhead in terms of FPGA resources while providing the highest level of abstraction and virtualization. Expand
FlexSaaS: A Reconfigurable Accelerator for Web Search Selection
The design for FlexSaaS (Flexible Selection as a Service), an FPGA-based accelerator for web search selection that contains a reconfigurable number of matching processors that can handle various possible query plans, and includes a universal memory accessor that hides the complex memory hierarchy and reduces host data access latency is presented. Expand
Kelp: QoS for Accelerated Machine Learning Systems
Kelp, a software runtime that isolates high priority accelerated ML tasks from memory resource interference, is designed and implemented and evaluated with both production and artificial aggressor workloads, and its effectiveness is evaluated. Expand
Parallelism Analysis of Prominent Desktop Applications: An 18- Year Perspective
The analyses show that the harnessed parallelism has improved and emerging workloads show good utilization of hardware resources, and the effectiveness of software in utilizing the underlying hardware has improved, but still has scope for optimizations. Expand
The Accelerator Wall: Limits of Chip Specialization
  • Adi Fuchs, D. Wentzlaff
  • Computer Science
  • 2019 IEEE International Symposium on High Performance Computer Architecture (HPCA)
  • 2019
This work characterizes how current accelerators depend on CMOS scaling, based on a physical modeling tool that is constructed using datasheets of thousands of chips, and builds a model which projects forward to see what future gains can and cannot be enabled from chip specialization. Expand