Automated topology synthesis for analog integrated circuits
Bibliographic record
Abstract
Currently, except for circuit topology synthesis, all the other phases in the analog integrated circuit design procedure are equipped with electronic design automation (EDA) commercial tools to greatly facilitate the human laborious work and significantly improve the design productivity, even though they are still not as mature as digital EDA counterparts. This dissertation focuses on developing a circuit topology synthesis EDA tool for analog integrated circuits. In order to make the developed EDA tool commercializable, there are many challenges that have to be solved, including trustworthy solutions, innovative solutions, wide applicability, sound generalization capability, and affordable computation effort. This thesis proposes a graph-based generation method to automatically synthesize analog integrated circuits, which has partially solved some challenges. But one serious problem of this method is its unaffordable computation effort due to the time-consuming sizing process for a huge number of generated circuit structures. To address this problem, we propose a novel performance modeling method that can boost the sizing efficiency by more than 30 times with ignorable model building overhead, which is especially suitable for the circuit synthesis work that involves generating various circuit structures. With the assistance of the emerging machine learning advancement, EDA tools can be more efficient and effective. We have employed the deep reinforcement learning technique in this dissertation to synthesize analog integrated circuit structures. Its technical merits make it be able to address those pending challenges much better than the graph-based generation method. But it still suffers from a shortcoming, that is, the learning process has to be performed from scratch once the technology or design specification changes. In order to overcome this shortcoming, the transfer learning technique is applied to transfer the learned knowledge from a learning process to another in order to largely save the learning effort. The experimental results exhibit strong efficacy and great applicability of our proposed methods.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.004 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".