Dr. Peng Wang | AI and Cloud Computing | Best Researcher Award

Researcher at Inspur (Jinan) Data Technology Co., Ltd., China

Dr. Peng Wang is a dynamic and ambitious researcher specializing in computer architecture, GPU rendering optimization, and compiler optimization. With a strong academic foundation and a flair for innovative problem-solving, he has carved a niche for himself in the interdisciplinary domains of computer science and statistics. As a Ph.D. candidate at the Chinese Academy of Sciences, Shenzhen Institute of Advanced Technology, he has consistently demonstrated academic excellence and technical prowess. His work is characterized by the integration of theory and practice, notably through the development of tools such as RayBench and RenderBench, which optimize GPU and CPU rendering respectively. Wang’s intellectual contributions extend to multiple high-impact publications and patents, underlining his commitment to advancing computational efficiency and performance benchmarking. Proficient in a wide range of programming languages and frameworks, including LLVM, CUDA, and RISC-V, Dr. Wang combines deep technical skills with a broad interdisciplinary understanding. His expertise is further validated by his active engagement in peer-review processes for reputed journals and international conferences. A fluent English speaker with robust communication abilities, Dr. Wang is poised for a distinguished career in academia or industry, where he aims to continue making transformative contributions to computing and data optimization.

Profile

ORCID

Education

Dr. Peng Wang’s academic journey reflects a strong and deliberate progression toward excellence in computer science and related domains. He is currently concluding his Ph.D. in Computer Architecture at the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, under the mentorship of Professor Yu Zhibin. His research, focusing on GPU rendering and compiler optimization, demonstrates a sophisticated understanding of complex architectural systems. He has successfully defended his thesis and is expected to graduate in June 2024. Prior to his doctoral studies, he earned a Master’s degree in Statistics from North China Electric Power University in Beijing between 2016 and 2019, where he ranked 7th out of 27 students, with a GPA of 3.42 out of 4. His undergraduate studies were equally impressive, as he pursued a double degree in Mathematics and Computer Science in English at the China University of Petroleum from 2011 to 2015. He ranked 13th out of 90 students with a GPA of 3.16 out of 4. This combination of statistical rigor and computational expertise has equipped him with a versatile academic foundation, enabling his research to cross traditional disciplinary boundaries and deliver impactful results in the evolving tech landscape.

Professional Experience

Dr. Wang has accumulated a wealth of professional experience through his involvement in cutting-edge research and benchmark development. During his doctoral studies, he took the lead in creating RayBench, an NVIDIA-centric GPU rendering benchmark suite that has significantly enhanced the understanding of performance characteristics in rendering environments. He also contributed to RenderBench, a CPU rendering benchmark based on microarchitecture-independent characteristics, allowing researchers to analyze rendering workloads without the constraints of specific hardware platforms. Additionally, Dr. Wang’s work on LLVM compiler optimization for RISC-V platforms showcases his ability to manipulate low-level architecture-specific code for improved execution efficiency. He also developed MICPAT, a GPU feature profiling tool aimed at extracting performance data without dependency on specific microarchitectures—currently under review by IEEE Transactions on Computers. These projects highlight his hands-on experience in benchmarking, tool development, and performance evaluation. Furthermore, his involvement in multiple patent filings illustrates his innovative thinking and practical application of theoretical research. His role as a reviewer for top-tier journals and international conferences demonstrates recognition from the wider research community. This blend of research, development, and community service underlines his professional maturity and readiness for high-impact roles in academia or the tech industry.

Research Interest

Dr. Peng Wang’s research interests lie at the intersection of computer architecture, GPU rendering optimization, and compiler techniques. His primary focus is on enhancing computational efficiency and system performance through architectural and compiler-level innovations. He is deeply engaged in the exploration of GPU rendering, where he develops benchmark suites that can evaluate and improve rendering performance across platforms. His work aims to abstract rendering characteristics from specific microarchitectures to build more generalizable tools—this vision is embodied in projects like RayBench and RenderBench. Additionally, he is invested in compiler optimization, particularly in the context of RISC-V architecture, leveraging LLVM to implement vectorization techniques that enhance execution throughput. Dr. Wang is also intrigued by the synergy between computer science and statistics, often applying machine learning methods such as PCA, SVM, and XGBoost to analyze system behaviors and predict performance trends. This cross-disciplinary interest helps him draw meaningful insights from large data sets, thereby improving hardware-software interaction. Ultimately, his research is driven by the ambition to simplify and universalize system performance benchmarking, making it accessible and adaptable to new and evolving computing paradigms. His work supports future-proof design in high-performance computing and embedded system domains.

Research Skills

Dr. Wang possesses a rich repertoire of research skills that reflect his interdisciplinary training and hands-on experience. He is adept at multiple programming languages including C, C++, and Verilog, which serve as the backbone for his development in hardware-software interaction and compiler design. His technical proficiency is complemented by his command of machine learning algorithms like SVM, Random Forests, XGBoost, and CatBoost, which he effectively applies to analyze system performance and classify architectural features. In terms of platforms and tools, he is highly experienced in LLVM (for compiler optimization), CUDA (for GPU programming), and instruction set architectures such as RISC-V, ARM, and x86. His fluency in English (CET-6 score: 608) enables him to communicate complex technical ideas clearly in international settings. Dr. Wang is also skilled in developing benchmark suites—he has successfully designed RayBench and RenderBench to evaluate GPU and CPU rendering performance respectively. His ability to synthesize performance metrics from both software and hardware perspectives allows him to provide nuanced insights into system behavior. Moreover, his patent filings and research publications underscore his capacity for innovation. Altogether, his diverse skill set positions him as a highly competent researcher capable of tackling contemporary challenges in computer architecture and system optimization.

Awards and Honors

His publications in reputed journals such as Electronics and IEEE Access reflect the scholarly validation of his work in GPU rendering and compiler optimization. The acceptance of his benchmark suites, RayBench and RenderBench, into the academic community illustrates both the novelty and applicability of his tools. His role as a reviewer for high-impact journals like Neural Computing and Applications and Current Drug Metabolism and participation in international conferences such as MLIS and CSAE underscore his active involvement in the global research landscape. In addition to these scholarly accolades, Dr. Wang has filed multiple patents related to GPU optimization and cloud gaming—an indication of his forward-thinking approach and industrial relevance. These achievements, combined with strong academic standings throughout his educational career, serve as cumulative honors validating his potential. Though he may not list traditional awards explicitly, his innovative contributions, peer recognition, and publication success collectively reflect a career adorned with academic distinction and promise for future honors.

Publications

Dr. Wang has an impressive publication record that underscores his contributions to computer architecture and GPU optimization. His works have been published in peer-reviewed journals such as Electronics and IEEE Access, with topics ranging from benchmark suite development to compiler vectorization strategies. Notable publications include “RenderBench: The CPU Rendering Benchmark Suite Based on Microarchitecture-Independent Characteristics” and “RayBench: An Advanced NVIDIA-Centric GPU Rendering Benchmark Suite for Optimal Performance Analysis,” both of which have appeared in Electronics and reflect his focus on performance benchmarking tools. His collaborative work with Professor Yu Zhibin has resulted in several impactful papers that have advanced the field’s understanding of microarchitecture-independent profiling. Another significant publication is “LLVM RISC-V RV32X Graphics Extension Support and Characteristics Analysis of Graphics Programs,” which highlights his deep understanding of compiler behavior and its application to emerging processor architectures. In addition to journal articles, he has filed several patents concerning GPU optimization and cloud rendering solutions, extending his influence to practical technological applications. His publication portfolio, which continues to grow, reflects a consistent and coherent research agenda aimed at improving the performance and portability of modern computing systems through rigorous evaluation and innovative design.

Conclusion

Dr. Peng Wang stands out as a promising researcher whose academic and professional trajectory reflects a strong commitment to innovation, technical mastery, and scholarly excellence. His expertise spans key areas in computer architecture, including GPU rendering, benchmark suite development, and compiler optimization for modern hardware platforms. His contributions have not only been validated through high-impact publications and patents but also through his active engagement in the academic community as a reviewer and conference participant. Armed with a strong educational background, practical skills in system-level programming and data analysis, and an ability to bridge theory with real-world applications, Dr. Wang is well-prepared to tackle the complex challenges of modern computing. His work has already laid the groundwork for more robust, efficient, and architecture-independent performance benchmarking tools. Moving forward, he is poised to make even more significant contributions, whether in academia, industry, or collaborative research environments. Dr. Wang embodies the qualities of a future leader in computing systems research—innovative, disciplined, and forward-thinking.