Multiqueue-Based FPGA Routing: Relaxed A* Priority Ordering for Improved Parallelism
Bibliographic record
Abstract
Routing is a critical part of the FPGA CAD flow. Its solution quality greatly impacts design speed and power, and its run time significantly contributes to overall compile time. As FPGA design size grows but per-core CPU performance stagnates, parallel FPGA routing algorithms that can leverage multiple compute cores become increasingly valuable. Most prior work in parallel FPGA routing uses coarse-grained approaches that route nonoverlapping nets in parallel; this work targets a complementary fine-grained form of parallelism in which the shortest-path algorithms that complete a single connection are multi-threaded. We speed up several related shortest path algorithms (Dijkstra's, A*, and directed) by utilizing a concurrencyfriendly but weakly ordered data structure, the MultiQueue, and enhance the algorithms to compensate for its imperfect ordering of partial routings. Compared to the VTR 8+ router, these techniques achieve routing time reductions of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$18.7 \times$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$13.2 \times$</tex> on average over the Titan benchmark suite when using Dijkstra's and A* path search, respectively, on a 12 -core CPU. These parallel algorithms achieve wirelength and critical path delay quality comparable to the serial router; they are also deterministic and serially equivalent. When applied to directed search routing, the parallel path search achieves a speed-up of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1.98 \times$</tex> and a slightly higher quality than the serial VTR router, but is nondeterministic. Thanks to queue improvements, our router at 1-thread is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1.7 \times$</tex> faster than VTR's for Dijkstra's and A* search, but comparable in run time for directed search.
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 | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".