Skip to content

Latest commit

 

History

History
81 lines (62 loc) · 6.01 KB

PAST.md

File metadata and controls

81 lines (62 loc) · 6.01 KB

Past Architecture Papers

Contributed by Jingwen Leng

CNN Accelerator 

Title Key Words
Diannao: A small-footprint high-throughput accelerator for ubiquitous machine-learning First DNN accelerator paper
Eyeriss:A spatial architecture for energy-efficient dataflow for convolutional neural networks Data reuse, more principles for DNN accelerator
EIE: efficient inference engine on compressed deep neural network First DNN compression paper
Minerva: Enabling low-power, highly-accurate deep neural network accelerators Low voltage
Dadiannao: A machine-learning supercomputer
Cnvlutin: Ineffectual-neuron-free deep neural network computing
Cambricon: An instruction set architecture for neural networks

 

General Specialized Accelerator

Title Key Words
Convolution engine: balancing efficiency & flexibility in specialized computing 2D convolvution
TABLA: A Unified Template-based Framework for Accelerating Statistical Machine Learning 
Pushing the Limits of Accelerator Efficiency While Retaining Programmability
Dynamically Specialized Datapaths for Energy Efficient Computing
Efficient execution of memory access phases using dataflow specialization
Analyzing Behavior Specialized Acceleration
Exploring the potential of heterogeneous von neumann/dataflow execution models
Bundled Execution of Recurring Traces for Energy-Efficient General Purpose Processing
A General Constraint-centric Scheduling Framework for Spatiale Architectures
Architectural Specialization for Inter-Iteration Loop Dependence Patterns
Stream-Dataflow Acceleration
GPU-CC: a Reconfigurable GPU Architecture with Communicating Cores
Single-Graph Multiple Flows: Energy Efficient Design Alternative for GPGPUs
Control Flow Coalescing on a Hybrid Dataflow von Neumann GPGPU

 

CPU Microarchitecture

Title Key Words
SIMULTANEOUS MULTITHREADING: A Platform for Next-Generation Processors SMT
Understanding sources of inefficiency in general-purpose chips Energy efficiency
Dark Silicon and the End of Multicore Scaling Multicore CPU
Composite Cores: Pushing Heterogeneity into a Core Heterogeneous cores
Dynamic Branch Prediction with Perceptrons Branch Prediction
Conservation Cores: Reducing the Energy of Mature Computations CPU+ASIC
MorphCore: An Energy-Efficient Microarchitecture forHigh Performance ILP and High Throughput TLP Heterogeneous cores
High-Performance and Energy-Efficient Mobile Web Browsing on Big/Little Systems mobile computing

 

GPU Microarchitecture

Title Key Words
Analyzing CUDA workloads using a detailed GPU simulator GPGPU-Sim
GPUWattch: enabling energy optimizations in GPGPUs GPU Power
CAPRI: Prediction of Compaction-Adequacy for Handling Control-Divergence in GPGPU Architectures Control Divergence in GPU
A Locality-Aware Memory Hierarchy for Energy-Efficient GPU Architectures Memory Divergence in GPU

 

FPGA

Title Key Words
A reconfig-urable fabric for accelerating large-scale datacenter services Datacenter

 

Datacenter 

Title Key Words
Bubble-up: Increasing utilization in modern warehouse scale computers via sensible co-locations First co-location paper
Heracles: Improving Resource Efficiency at Scale