A Revolution in Performance using a GPU Based Simulator
Bibliographic record
Abstract
Abstract Geological models are preferably built with very fine grids in a 3D dimensional geometry to capture reservoir complexity and heterogeneity. However, simulating such detailed models with sizes that can range to the 100’s of millions of cells is a huge challenge for current commercial CPU (Central Processing Unit) simulators as the results cannot be achieved within an acceptable time frame and simulation may last days or weeks. These long run times cause a delay in achieving a robust history matched or production forecast model as it constrains the ability to effectively characterize the uncertainty range of subsurface parameters and potential development solutions. Restricting the number of alternate scenarios in this manner diminishes the ability to arrive at solutions which may lead to loss of business opportunity and economic value. The alternatives have been used by oil and gas industry so far to reduce simulation run time, are either upscaling fine grid models to a coarser grid, consequently reducing the number of cells or increasing the number of nodes on a high-performance cluster so that simulation results can be achieved more rapidly by parallel computing using CPU based simulators. The former solution is limited by reservoir complexity and the latter by the scaling limits of the software. However, recent advances in high performance technical computing using Graphic Processing Units (GPU) has generated significant interest in the performance characteristics of a GPU-based simulator. This paper demonstrates performance of new generation of GPU based simulator as compared to CPU based simulators. A simulation run time of 9-11 hours on a 4-core CPU based simulator was reduced to 25-40 min on 1 GPU based simulator with tight bounds on results.
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 | 0.000 | 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".