トップページ イベント・広報 R-CCS Cafe R-CCS Cafe 第271回(2024年10月11日)
R-CCS Cafe 第271回(2024年10月11日)
English開催日 | 2024年10月11日(金) |
---|---|
開催時間 | 15:00 - 17:00(15:00 - 15:40 講演者2名による講演、15:50 - 16:10 討論、16:10 - 自由討論とコーヒーブレイク(参加自由)) |
開催都市 | 兵庫県神戸市/オンライン |
場所 | 計算科学研究センター(R-CCS)6階講堂/Zoomによる遠隔セミナー |
使用言語 | 発表・スライド共に英語 |
登壇者 |
講演題目・要旨
1st Speaker: Lingqi Zhang
Title:
Overcoming the Gap Between Compute and Memory Bandwidth in Modern GPUs
Abstract:
The imbalance between compute and memory bandwidth has been a long-standing issue. Despite efforts to address it, the gap between them is still widening. This has led to the categorization of many applications as memory-bound kernels. This dissertation centers on memory-bound kernels, with a particular emphasis on Graphics Processing Units (GPUs), given their rising prevalence in High-Performance Computing (HPC) systems. In this dissertation, we initially focus on the evolution trend of GPU development in the last decades. Examples include cooperative groups (i.e., device-wide barriers), asynchronous copy of shared memory (i.e., hardware prefetching), low(er) latency of operations, and larger volume of on-chip resources (register files and L1 cache). This dissertation seeks to utilize the latest GPU features to optimize memory-bound kernels. Specifically, we propose extending the kernel's lifetime across the time steps and taking advantage of the large volume of on-chip resources (i.e., register files and scratchpad memory) to reduce or eliminate traffic to the device memory. Furthermore, we champion a minimum level of parallelism to maximize the available on-chip resources. Based on the strategies, we propose a general execution model for running memory-bound iterative GPU kernels: PERsistent KernelS (PERKS) and a novel temporal blocking method, EBISU. Evaluations have shown outstanding performance in the latest GPU architectures compared with counterpart state-of-the-art implementations.
2nd Speaker: Md Anowar Hossain
Title:
Cognitive Driver’s Sensitivity for Next-generation Automated Cruising Traffic
Abstract:
A traffic flow model, the so-called Taillight Adapting traffic model, is developed considering the driver's activities impact following the taillight effect of the preceding vehicle, which is switched on and off based on the vehicle's accelerations and decelerations. To quantify a driver's sensitivity to the velocity gap, we introduced the concept of the new driver's sensitivity function, which has twofold formulations. The first formulation follows the HDDS model1, which presumes the sensitivity is dependent on a headway distance. The second formulation presumes that the sensitivity depends on a velocity gap, which might be informed by the taillight of a preceding vehicle. We have calibrated the neutralizing proficiency of a traffic flow field for the HDDS and proposed the Taillight Adapting model by utilizing the linear stability theory. Finally, a series of numerical simulations have been carried out of the improved Taillight Adapting model and conventional FVD model to investigate how the traffic flow field for the Taillight Adapting model behaves differently from the conventional FVD model.
注意事項
- 参加の際はPCマイクの音声・ビデオをオフにされるようお願いいたします。
- 当日の会場環境や通信状態により、やむなく配信を中止・中断する場合がございます。
- プログラムの内容、時間は予告なく変更される場合があります。
- ご使用の機器やネットワークの環境によっては、ご視聴いただけない場合がございます。
- インターネット中継に関する著作権は、主催者及び発表者に帰属します。なお、配信された映像及び音声、若しくはその内容を、理化学研究所の許可無くほかのウェブサイトや著作物等への転載、複製、改変等を行うことを禁じます。
(2024年10月7日)