{"id":64,"date":"2022-12-28T11:27:54","date_gmt":"2022-12-28T02:27:54","guid":{"rendered":"https:\/\/thunderexample5.wordpress.com\/?page_id=64"},"modified":"2026-03-09T12:31:45","modified_gmt":"2026-03-09T12:31:45","slug":"software-and-systems","status":"publish","type":"page","link":"https:\/\/thunder.snu.ac.kr\/?page_id=64","title":{"rendered":"Software and Systems"},"content":{"rendered":"\n

Systems <\/em><\/strong>– Chundoong<\/a>
Datasets <\/em><\/strong>– UDC-SIT<\/a> \/ UDC-VIT<\/a>
Software <\/em>(GitHub<\/a>)<\/strong> – SnuQS<\/a> \/ SnuRHAC<\/a> \/ SOFF<\/a> \/ HUM<\/a> \/ SNU NPB 2019 Suite<\/a> \/ SnuCL<\/a> \/ SNU-SAMSUNG OpenCL framework<\/a> \/ SFMalloc<\/a> \/ FaCSim<\/a> \/ COMIC<\/a><\/p>\n\n\n\n

\n\n\n\n
\n
\n

Systems<\/strong><\/h3>\n\n\n\n

Chundoong<\/a><\/h4>\n\n\n\n

A Low-Cost Energy-Efficient Heterogeneous Supercomputer.<\/p>\n\n\n\n

<\/p>\n\n\n\n

Datasets<\/strong><\/h3>\n\n\n\n

UDC-SIT<\/a><\/h4>\n\n\n\n

A Real-World Still Image Dataset for Under-Display Cameras.<\/p>\n\n\n\n

UDC-VIT<\/a><\/h4>\n\n\n\n

A Real-World Video Dataset for Under-Display Cameras.<\/p>\n<\/div>\n\n\n\n

\n

Software<\/strong> (GitHub<\/a>)<\/strong><\/h3>\n\n\n\n

SnuQS<\/a><\/h4>\n\n\n\n

A Quantum circuit simulator using storage devices.<\/p>\n\n\n\n

SnuRHAC<\/a><\/h4>\n\n\n\n

A Runtime for Heterogeneous Accelerator Clusters with CUDA Unified Memory.<\/p>\n\n\n\n

SOFF<\/a><\/h4>\n\n\n\n

An OpenCL framework for FPGAs.<\/p>\n\n\n\n

HUM<\/a><\/h4>\n\n\n\n

A CUDA framework that hides host-to-device memory copy time by exploiting CUDA Unified Memory and fault mechanisms.<\/p>\n\n\n\n

SNU NPB 2019 Suite<\/a><\/h4>\n\n\n\n

The NASA Parallel Benchmark suite ported in C, OpenMP C, and OpenCL for both a single compute device and multiple compute devices.<\/p>\n\n\n\n

SnuCL<\/a><\/h4>\n\n\n\n

An OpenCL framework for a heterogeneous cluster that consists of nodes with multiple heterogeneous devices (e.g., multicore CPUs, AMD GPUs, NVIDIA GPUs, and Intel Xeon Phi coprocessors).<\/p>\n\n\n\n

SNU-SAMSUNG OpenCL framework<\/a><\/h4>\n\n\n\n

An OpenCL framework for ARM CPUs, IBM Cell BE processors, and TI DSPs.<\/p>\n\n\n\n

SFMalloc<\/a><\/h4>\n\n\n\n

A lock-free and mostly synchronization-free dynamic memory allocator for manycores.<\/p>\n\n\n\n

FaCSim<\/a><\/h4>\n\n\n\n

Fast and cycle-accurate ARM architecture simulator.<\/p>\n\n\n\n

COMIC<\/a><\/h4>\n\n\n\n

Coherent shared memory interface (runtime) for IBM Cell BE processors.<\/p>\n<\/div>\n<\/div>\n\n\n\n

<\/p>\n\n\n\n

\n\n\n\n\n\n\n\n

Systems <\/em><\/strong>– Chundoong<\/a>
Datasets <\/em><\/strong>– UDC-SIT<\/a> \/ UDC-VIT<\/a>
Software <\/em>(GitHub<\/a>)<\/strong> – SnuQS<\/a> \/ SnuRHAC<\/a> \/ SOFF<\/a> \/ HUM<\/a> \/ SNU NPB 2019 Suite<\/a> \/ SnuCL<\/a> \/ SNU-SAMSUNG OpenCL framework<\/a> \/ SFMalloc<\/a> \/ FaCSim<\/a> \/ COMIC<\/a><\/p>\n\n\n\n

\n\n\n\n

Chundoong<\/strong> A Low-Cost Energy-Efficient Heterogeneous Supercomputer<\/em><\/small><\/h3>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Chundoong is a heterogeneous CPU\/GPU supercomputer designed and built by the THUNDER Research Group at Seoul National University and ManyCoreSoft<\/a> in October 2012. It is used to evaluate parallel programming models (SnuCL<\/a>), software techniques, and applications developed in Seoul National University.<\/p>\n\n\n\n

After porting the double precision LINPACK (HPL) to Chundoong (written in MPI + OpenCL) and optimizing it with our software techniques for multiple GPUs, we have achieved 106.8 TFLOPS (1.907 TFLOPS per node). Chundoong is ranked 277th in the TOP500<\/a> list and 32nd in the Green500<\/a> list of November 2012. Its per-node performance is #1 among 412 clusters in TOP500<\/a> of November 2012.<\/p>\n\n\n\n

The design of Chundoong is focused on achieving low cost and low power consumption. As a result, Chundoong costed only US$ 0.67 million (159 MFLOPS\/dollar). Chundoong is referred as the 7th power efficient architecture<\/a> in the TOP500<\/a> list of November 2012.<\/p>\n\n\n\n

\n
\n
\"\"<\/figure>\n<\/div>\n\n\n\n
\n
\"\"<\/figure>\n<\/div>\n<\/div>\n\n\n\n

<\/p>\n\n\n\n

System Overview<\/h4>\n\n\n\n

Hardware<\/h5>\n\n\n\n

Chundoong is built using only commodity hardware. The detailed description of the Chundoong cluster is as follows:<\/p>\n\n\n\n