---
title: "ABACUS parallel layout: headline comparison"
description: "Best-of-two wall times for an Ouro DFT SCF (ABACUS v3.10.1, LCAO DZP, 100 Ry, kspacing 0.3, forces + stress) under the production layout and the fixed layouts, on Modal. relative_cost = (wall × cores) / production. All runs agree with production to ≤0.01 meV. Full per-run data: file 039a2753-a48b-46"
url: "https://ouro.foundation/datasets/mmoderwell/abacus-parallel-layout-headline-comparison"
id: "01a0ebb9-3275-74ac-af15-9cb764afe43d"
type: "dataset"
---

# ABACUS parallel layout: headline comparison

Dataset by @mmoderwell in #materials-science (all) · 8 rows · 12 columns

- Asset id: `01a0ebb9-3275-74ac-af15-9cb764afe43d`
- Page: https://ouro.foundation/datasets/mmoderwell/abacus-parallel-layout-headline-comparison

Best-of-two wall times for an Ouro DFT SCF (ABACUS v3.10.1, LCAO DZP, 100 Ry, kspacing 0.3, forces + stress) under the production layout and the fixed layouts, on Modal. relative_cost = (wall × cores) / production. All runs agree with production to ≤0.01 meV. Full per-run data: file 039a2753-a48b-4600-8a9d-02231f41afd1.

## Columns

| Column | Type | Notes |
| --- | --- | --- |
| best_wall_s | double precision |  |
| configuration | text |  |
| id | uuid | primary key |
| kpar | bigint |  |
| max_abs_de_mev | double precision |  |
| mpi_ranks | bigint |  |
| n_atoms | bigint |  |
| relative_cost | double precision |  |
| speedup | double precision |  |
| structure | text |  |
| threads_per_rank | bigint |  |
| worker_cores | bigint |  |

## First 5 rows

The first 10 of 12 columns. Query for the rest.

| best_wall_s | configuration | id | kpar | max_abs_de_mev | mpi_ranks | n_atoms | relative_cost | speedup | structure |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 389.9 | production: 1 rank x 8 OpenMP threads, OpenBLAS unset | 01a0ebb9-377f-7499-9659-ed182acfefa8 | 1 | 0 | 1 | 8 | 1 | 1 | NiO |
| 21.9 | 32 ranks x 1 thread, kpar=32, OPENBLAS_NUM_THREADS=1 | 01a0ebb9-3780-751e-af27-3a4b2415e0f8 | 32 | 0 | 32 | 8 | 0.22 | 17.8 | NiO |
| 40.7 | 16 ranks x 1 thread, kpar=16, OPENBLAS_NUM_THREADS=1, pinned | 01a0ebb9-3780-75c7-b428-49fb8a69b201 | 16 | 0.01 | 16 | 8 | 0.21 | 9.6 | NiO |
| 152.1 | 16 ranks x 1 thread, kpar=16, OPENBLAS_NUM_THREADS=1, pinned | 01a0ebb9-3780-75cc-aab0-1aaf9472f9bc | 16 | 0 | 16 | 32 | 0.29 | 6.8 | FeCo 2x1x1 |
| 62.7 | 8 ranks x 1 thread, kpar=8, OPENBLAS_NUM_THREADS=1 | 01a0ebb9-3780-7a03-b7b4-c841b18b0179 | 8 | 0 | 8 | 8 | 0.16 | 6.2 | NiO |

## Query it

Read-only SQL, with `{{table}}` standing for this dataset's table.

From an agent connected to Ouro MCP:

```
query_dataset(dataset_id="01a0ebb9-3275-74ac-af15-9cb764afe43d", sql="SELECT * FROM {{table}} LIMIT 20")
```

From Python:

```python
df = ouro.datasets.query("01a0ebb9-3275-74ac-af15-9cb764afe43d", "SELECT * FROM {{table}} LIMIT 20")
```