1. |
- Colucci, A., et al.
(författare)
-
MLComp : A Methodology for Machine Learning-based Performance Estimation and Adaptive Selection of Pareto-Optimal Compiler Optimization Sequences
- 2021
-
Ingår i: Proceedings -Design, Automation and Test in Europe, DATE. - : Institute of Electrical and Electronics Engineers Inc.. - 9783981926354 ; , s. 108-113
-
Konferensbidrag (refereegranskat)abstract
- Embedded systems have proliferated in various consumer and industrial applications with the evolution of Cyber-Physical Systems and the Internet of Things. These systems are subjected to stringent constraints so that embedded software must be optimized for multiple objectives simultaneously, namely reduced energy consumption, execution time, and code size. Compilers offer optimization phases to improve these metrics. However, proper selection and ordering of them depends on multiple factors and typically requires expert knowledge. State-of-the-art optimizers facilitate different platforms and applications case by case, and they are limited by optimizing one metric at a time, as well as requiring a time-consuming adaptation for different targets through dynamic profiling. To address these problems, we propose the novel MLComp methodology, in which optimization phases are sequenced by a Reinforcement Learning-based policy. Training of the policy is supported by Machine Learning-based analytical models for quick performance estimation, thereby drastically reducing the time spent for dynamic profiling. In our framework, different Machine Learning models are automatically tested to choose the best-fitting one. The trained Performance Estimator model is leveraged to efficiently devise Reinforcement Learning-based multi-objective policies for creating quasi-optimal phase sequences. Compared to state-of-the-art estimation models, our Performance Estimator model achieves lower relative error (< 2%) with up to 50 × faster training time over multiple platforms and application domains. Our Phase Selection Policy improves execution time and energy consumption of a given code by up to 12% and 6%, respectively. The Performance Estimator and the Phase Selection Policy can be trained efficiently for any target platform and application domain. © 2021 EDAA.
|
|
2. |
- Deckers, Tobias, et al.
(författare)
-
Impact of processing gas composition on process stability and properties of PBF-LB/M processed alloy 718
- 2024
-
Ingår i: Journal of Manufacturing Processes. - 1526-6125. ; 120, s. 712-718
-
Tidskriftsartikel (refereegranskat)abstract
- The almost unlimited design freedom of the laser-based powder bed fusion of metals (PBF-LB/M) makes this technology very attractive for industry. However, as a developing technology, it still faces some challenges when it comes to productivity and robustness, to name some. Whereas numerous studies covered the impact of laser-based parameters on material properties and robustness, the effect of the processing gas received limited attention. The objective of this study was to evaluate the effect of processing gas composition, containing helium (He) and hydrogen (H2), compared to conventionally used argon (Ar), during PBF-LB/M processing of virgin alloy 718 powder, on printing behavior and part properties. The four gases studied were Ar, Ar +30%He, Ar +30%He +2%H2, and Ar +70%He. Optical Tomography (OT) was used to monitor process stability, which unveiled a significant decrease in process-by products (spatters) between 51 % and 89 % using He and H2-containing gases. It was also found that the process gas decreased the bulk porosity from an average value of 0.08 % when processed with Ar to 0.04 % when using Ar + 70%He. The oxygen pickup by the spatter particles was reduced from 630 ppm (Ar) to 331 ppm (Ar +70%He). EBSD analysis revealed that there were no evident changes in microstructure with the processing gas. The samples processed also had similar tensile properties with yield and ultimate tensile strength of 1180 MPa and 1395 MPa, respectively. However, there was a slight increase in ductility from 16.5 % to 17.2 %, when processed with pure Ar and Ar + 70%He, respectively. This study shows that utilizing standard Ar processing atmosphere with He addition leads to a more stable process with reduced spatter, porosity and a marginal increase in ductility for Alloy 718.
|
|