I am a music gamer who is obsessed in database systems and performance.
I am conducting research on graph database systems in the language of
linear algebra.
I believe that the language of linear algebra can be one of the sustainable bridges
from high-performance computing to graph database systems.
One of my goals is to create a frontier data system group in Thailand.
PhD Candidate (3rd Year) @ Carnegie Mellon University and CMKL University
(Dual Degree - Electrical and Computer Engineering)
Teaching Assistant
CMU 18-746 Storage Systems (F22)
CMKL 18-613 Foundation of Computer Systems (S24)
Selected Coursework
CMU 15-721 Advanced Database Systems (S23)
CMU 18-746 Storage Systems (F21)
CMU 15-745 Optimizing Compilers for Modern Architectures (S23)
CMU 18-740 Modern Computer Architecture and Design (F22)
CMU 18-660 Optimization (S22)
CMU 18-447 Introduction to Computer Architecture (S22)
CMKL 18-613 Foundation of Computer Systems (S21)
CMKL 18-652 Foundations of Software Engineering (S21)
Previously be an Undergraduate Student @ KMITL
(Computer Science)
I founded CS-INFINITE, a group of students who initiates activities for CS-KMITL students.
Examples of those activities are Code Arcade and Programming Bootcamp.
Living in Bangkok, Thailand
Originally, lived in Nakhon Pathom, Thailand
Research
Code Generation for Graph Queries through Linear Algebra
Ongoing
Fast Fixed-Point Decimals
Ongoing
Exploiting Fusion Opportunities in Linear Algebraic Graph Query Engines
Accepted
Queries in a graph database are often converted into a sequence of graph operations by a graph query engine.
In recent years, it has been recognized that the query engine benefits from using high-performance graph libraries
via the GraphBLAS interface to implement time-consuming operations such as graph traversal. However, using
GraphBLAS requires explicitly casting data into linear algebra objects and decomposing the query into multiple
operations, some of which are expressible by the GraphBLAS. The combination of these two requirements translates
into increased memory footprints and additional execution times. In this paper, we show that fusing different
stages of the query engines into GraphBLAS calls can reduce the size of the intermediate data generated during the
query. Furthermore, by relaxing the semi-ring constraints imposed by GraphBLAS, more aggressive fusions of the
stages can be performed. We show a speedup of up to 1235.89x (8.82x on geometric average) relative to an
open-source graph query engine using GraphBLAS (i.e. RedisGraph) for processing undirected subgraph enumeration
queries.
Opportunities for Linear Algebraic Graph Databases
Extended Abstract
In recent years, there has been renewed interest in casting graph algorithms in the language of linear algebra. By
replacing the computations with appropriate operations over different semi-rings, different graph algorithms can
be cast as a sequence of linear algebra operations. In this work, we study the use of the linear algebraic
approach to graph algorithms within the context of graph database systems. Specifically, we identify the issues
with using existing linear algebraic graph libraries, such as SuiteSparse, which conform to the GraphBLAS
specifications. We also highlight gaps between the GraphBLAS specification and computations that are required by
the graph query algorithms utilized in graph databases. We show that overcoming these challenges in using a linear
algebraic approach within a graph database system can lead to significant performance improvements to an
open-source graph database system.
A PostgreSQL foreign data wrapper for accessing columnar file formats (similar to Apache Parquet).
The foreign data wrapper includes query optimizations such as projection and predicate pushdown.
Compiling Neural Network with Dynamic Resource Adaptation
An approach to improve elasticity of FlexGen for supporting multi-tenancy environments.
Webpage
Hybrid Cloud/Local File System
I implemented a hybrid cloud/local (Amazon S3 and local SSDs) file system to minimize cloud storage costs.
The file system can deduplicate identical file contents to reduce storage capacity.
The file system can take a snapshot and recovery using taken snapshots.
The file system was optimized through software-level write-back cache.
The file system was implemented in C++ via FUSE.
ounglang
ounglang is
a interpretative, turing-incomplete, and esoteric programming language for seals written in C.
PEARLS (Programming Evaluation and Rapid Learning Systems)
PEARLS is a
learning management system for CS-KMITL.
I designed and implemented a relational database schema for PostgreSQL.
PEARLS Lab
PEARLS Lab
is a programming competition platform for Code Arcade.
I designed and implemented web services and a relational database schema for PostgreSQL.