We evaluated performance of libgrape-lite with LDBC Graph Analytics Benchmark. In addition to the ease of programming, we find that libgrape-lite achieves high performance comparably to the state-of-the-art systems. The experiments were conducted on 4 instances of r6.8xlarge on AlibabaCloud ECS, each with 32 threads, over LDBC XL-size datasets. Instances are imaged with Aliyun Linux (a CentOS-variant).
We compared libgrape-lite with PowerGraph(commit a038f97 ) GeminiGraph(commit 170e7d3 ) and Plato(commit 21009d6). Each system is built with GCC(v4.8.5) and MPICH(v3.1). To make the comparisons fair, libgrape_lite was built with HUGE_PAGES and jemalloc disabled.
We made minor changes on their code and datasets:
- Turned on
-O3optimization for all three systems. - Added timing stubs for Plato.
- Replaced the random-pick logic with the deterministic logic from LDBC for the CDLP in Plato.
- Changed the weight type of SSSP from
floattodoublein GeminiGraph. - Changed the load strategy from
load_directedtoload_undirected_from_directedfor PageRank in GeminiGraph - Reformat the datasets to adapt Plato and GeminiGraph's formats (e.g., 0-based continuous vertex ids)
The results are reported below. The numbers in the table represent the evaluation time in seconds. The best results are marked in bold.
| Algorithm | Dataset | PowerGraph | GeminiGraph | Plato | libgrape-lite |
|---|---|---|---|---|---|
| SSSP | datagen-9_0-fb | 6.48 | 0.74 | N/A | 0.52 |
| datagen-9_1-fb | 7.14 | 0.99 | N/A | 0.62 | |
| datagen-9_2-zf | 59.48 | 3.50 | N/A | 1.85 | |
| WCC | datagen-9_0-fb | 16.99 | 1.05 | 2.85 | 0.52 |
| datagen-9_1-fb | 30.33 | 1.24 | 3.80 | 0.60 | |
| datagen-9_2-zf | 231.79 | 6.58 | 30.79 | 1.77 | |
| graph500-26 | 17.36 | 1.98 | 4.96 | 0.90 | |
| com-friendster | 68.79 | 4.77 | 9.39 | 2.40 | |
| BFS | datagen-9_0-fb | 4.31 | 0.38 | 0.73 | 0.23 |
| datagen-9_1-fb | 5.23 | 0.39 | 0.84 | 0.26 | |
| datagen-9_2-zf | 41.36 | 2.59 | 14.59 | 1.64 | |
| graph500-26 | 5.42 | 0.73 | 2.19 | 0.43 | |
| com-friendster | 13.14 | 1.28 | 3.55 | 1.12 | |
| PageRank | datagen-9_0-fb | 26.90 | X | X | 1.61 |
| datagen-9_1-fb | 34.01 | X | X | 1.99 | |
| datagen-9_2-zf | 152.11 | X | X | 4.00 | |
| graph500-26 | 34.89 | X | X | 2.95 | |
| com-friendster | 66.61 | X | X | 5.91 | |
| CDLP | datagen-9_0-fb | 1535.09 | N/A | 16.67 | 8.49 |
| datagen-9_1-fb | 2725.54 | N/A | 21.60 | 10.78 | |
| datagen-9_2-zf | > 3600 | N/A | 46.05 | 17.17 | |
| graph500-26 | 401.21 | N/A | 13.44 | 8.50 | |
| com-friendster | > 3600 | N/A | 37.81 | 20.17 | |
| LCC | datagen-9_0-fb | 453.07 | N/A | N/A | 16.53 |
| datagen-9_1-fb | 613.64 | N/A | N/A | 20.83 | |
| datagen-9_2-zf | 299.42 | N/A | N/A | 19.81 | |
| graph500-26 | 1838.60 | N/A | N/A | 232.50 | |
| com-friendster | 847.97 | N/A | N/A | 68.52 |
We used “default” code provided by the competitor systems when it is available.
- N/A indicate that the system didn't provide the application. And
- X indicates the results produced are not consistent with the verified results provided by LDBC.
The inconsistences of PageRank come from different settings on convergence conditions. To give a comprehensive comparison, we made our best efforts to revise our application(pagerank_local.h), making them output the same results as competitor systems. The performance results are shown as below.
| Algorithm | Dataset | GeminiGraph | Plato | libgrape-lite |
|---|---|---|---|---|
| PageRank | datagen-9_0-fb | 2.60 | 4.81 | 1.69 |
| datagen-9_1-fb | 3.24 | 5.17 | 1.88 | |
| datagen-9_2-zf | 7.18 | 41.43 | 2.99 | |
| graph500-26 | 6.61 | 14.67 | 2.88 | |
| com-friendster | 9.90 | 18.65 | 5.95 |
We will release a public image containing the script, together with all the systems and datasets on AlibabaCloud and AWS soon.