/
githubmirror
/
spark
Обзор
Документация
Войти
/
githubmirror
/
spark
Код
Запросы
0
Пакеты
0
Релизы
0
Аналитика
Безопасность
master
sql/core/benchmarks/DataSourceReadBenchmark-jdk21-results.txt
436 строк
43 KB
Dongjoon Hyun
[SPARK-58415][INFRA][TEST] Regenerate benchmark results
29 июл 2026, 10:51
29 июл 2026, 10:51
6b793ce
Код
Авторство
О чём код?
================================================================================================ SQL Single Numeric Column Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 9314 9381 94 1.7 592.2 1.0X SQL Json 7085 7172 122 2.2 450.5 1.3X SQL Json with UnsafeRow 8077 8120 61 1.9 513.5 1.2X SQL Parquet Vectorized: DataPageV1 82 101 15 190.9 5.2 113.1X SQL Parquet Vectorized: DataPageV2 91 103 12 172.2 5.8 102.0X SQL Parquet MR: DataPageV1 1819 1827 10 8.6 115.7 5.1X SQL Parquet MR: DataPageV2 1755 1778 32 9.0 111.6 5.3X SQL ORC Vectorized 146 155 9 107.6 9.3 63.7X SQL ORC MR 1580 1585 7 10.0 100.5 5.9X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 68 70 2 230.4 4.3 1.0X ParquetReader Vectorized: DataPageV2 84 86 2 186.6 5.4 0.8X ParquetReader Vectorized -> Row: DataPageV1 30 31 2 527.4 1.9 2.3X ParquetReader Vectorized -> Row: DataPageV2 46 48 2 338.9 3.0 1.5X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 9321 9331 14 1.7 592.6 1.0X SQL Json 7569 7617 68 2.1 481.2 1.2X SQL Json with UnsafeRow 8627 8635 11 1.8 548.5 1.1X SQL Parquet Vectorized: DataPageV1 106 118 10 148.8 6.7 88.2X SQL Parquet Vectorized: DataPageV2 108 119 9 146.2 6.8 86.7X SQL Parquet MR: DataPageV1 1938 1946 12 8.1 123.2 4.8X SQL Parquet MR: DataPageV2 1960 1969 12 8.0 124.6 4.8X SQL ORC Vectorized 102 108 8 153.7 6.5 91.1X SQL ORC MR 1604 1605 1 9.8 102.0 5.8X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 77 78 1 204.6 4.9 1.0X ParquetReader Vectorized: DataPageV2 77 78 1 204.9 4.9 1.0X ParquetReader Vectorized -> Row: DataPageV1 46 47 1 343.4 2.9 1.7X ParquetReader Vectorized -> Row: DataPageV2 46 47 1 343.6 2.9 1.7X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 9395 9408 19 1.7 597.3 1.0X SQL Json 7929 7942 18 2.0 504.1 1.2X SQL Json with UnsafeRow 8803 8814 15 1.8 559.7 1.1X SQL Parquet Vectorized: DataPageV1 127 138 8 124.1 8.1 74.2X SQL Parquet Vectorized: DataPageV2 163 175 8 96.7 10.3 57.8X SQL Parquet MR: DataPageV1 2099 2113 20 7.5 133.5 4.5X SQL Parquet MR: DataPageV2 1978 1979 2 8.0 125.8 4.7X SQL ORC Vectorized 142 151 7 110.5 9.0 66.0X SQL ORC MR 1715 1726 17 9.2 109.0 5.5X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 165 168 2 95.5 10.5 1.0X ParquetReader Vectorized: DataPageV2 209 214 6 75.3 13.3 0.8X ParquetReader Vectorized -> Row: DataPageV1 148 154 3 106.4 9.4 1.1X ParquetReader Vectorized -> Row: DataPageV2 187 189 2 84.1 11.9 0.9X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 10855 10906 72 1.4 690.2 1.0X SQL Json 8421 8433 16 1.9 535.4 1.3X SQL Json with UnsafeRow 9333 9334 1 1.7 593.4 1.2X SQL Parquet Vectorized: DataPageV1 116 128 9 135.7 7.4 93.7X SQL Parquet Vectorized: DataPageV2 214 221 5 73.4 13.6 50.6X SQL Parquet MR: DataPageV1 2109 2114 8 7.5 134.1 5.1X SQL Parquet MR: DataPageV2 2036 2070 49 7.7 129.4 5.3X SQL ORC Vectorized 179 195 6 88.0 11.4 60.7X SQL ORC MR 1791 1799 11 8.8 113.9 6.1X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 142 144 3 110.7 9.0 1.0X ParquetReader Vectorized: DataPageV2 211 212 1 74.6 13.4 0.7X ParquetReader Vectorized -> Row: DataPageV1 142 145 2 110.9 9.0 1.0X ParquetReader Vectorized -> Row: DataPageV2 208 211 1 75.5 13.2 0.7X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 10588 10627 55 1.5 673.2 1.0X SQL Json 8373 8387 20 1.9 532.4 1.3X SQL Json with UnsafeRow 9245 9255 15 1.7 587.8 1.1X SQL Parquet Vectorized: DataPageV1 284 296 7 55.4 18.0 37.3X SQL Parquet Vectorized: DataPageV2 192 197 3 81.9 12.2 55.1X SQL Parquet MR: DataPageV1 2608 2645 52 6.0 165.8 4.1X SQL Parquet MR: DataPageV2 2113 2140 37 7.4 134.4 5.0X SQL ORC Vectorized 192 198 4 81.8 12.2 55.1X SQL ORC MR 1813 1818 7 8.7 115.3 5.8X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 337 337 1 46.7 21.4 1.0X ParquetReader Vectorized: DataPageV2 228 232 2 68.8 14.5 1.5X ParquetReader Vectorized -> Row: DataPageV1 348 369 13 45.2 22.1 1.0X ParquetReader Vectorized -> Row: DataPageV2 245 253 11 64.1 15.6 1.4X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 10938 10964 37 1.4 695.4 1.0X SQL Json 10254 10275 30 1.5 651.9 1.1X SQL Json with UnsafeRow 10922 10928 7 1.4 694.4 1.0X SQL Parquet Vectorized: DataPageV1 90 105 7 174.9 5.7 121.6X SQL Parquet Vectorized: DataPageV2 87 104 6 180.9 5.5 125.8X SQL Parquet MR: DataPageV1 2142 2165 32 7.3 136.2 5.1X SQL Parquet MR: DataPageV2 2054 2064 14 7.7 130.6 5.3X SQL ORC Vectorized 243 252 10 64.8 15.4 45.1X SQL ORC MR 1834 1855 30 8.6 116.6 6.0X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 142 144 1 110.8 9.0 1.0X ParquetReader Vectorized: DataPageV2 145 147 1 108.5 9.2 1.0X ParquetReader Vectorized -> Row: DataPageV1 149 160 10 105.8 9.5 1.0X ParquetReader Vectorized -> Row: DataPageV2 143 146 1 109.7 9.1 1.0X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 11283 11313 42 1.4 717.4 1.0X SQL Json 10437 10438 2 1.5 663.5 1.1X SQL Json with UnsafeRow 11390 11393 4 1.4 724.2 1.0X SQL Parquet Vectorized: DataPageV1 288 294 5 54.5 18.3 39.1X SQL Parquet Vectorized: DataPageV2 287 293 5 54.9 18.2 39.4X SQL Parquet MR: DataPageV1 2531 2534 4 6.2 160.9 4.5X SQL Parquet MR: DataPageV2 2455 2458 4 6.4 156.1 4.6X SQL ORC Vectorized 491 496 4 32.0 31.2 23.0X SQL ORC MR 2102 2168 93 7.5 133.6 5.4X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Parquet Reader Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------- ParquetReader Vectorized: DataPageV1 351 362 10 44.8 22.3 1.0X ParquetReader Vectorized: DataPageV2 354 373 10 44.5 22.5 1.0X ParquetReader Vectorized -> Row: DataPageV1 341 342 1 46.1 21.7 1.0X ParquetReader Vectorized -> Row: DataPageV2 337 353 11 46.7 21.4 1.0X ================================================================================================ SQL Single Numeric Column Scan in Struct ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single TINYINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2202 2210 12 7.1 140.0 1.0X SQL ORC Vectorized (Nested Column Disabled) 2231 2232 2 7.0 141.9 1.0X SQL ORC Vectorized (Nested Column Enabled) 145 151 8 108.4 9.2 15.2X SQL Parquet MR: DataPageV1 2374 2399 35 6.6 151.0 0.9X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 2695 2731 51 5.8 171.3 0.8X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 105 110 6 150.1 6.7 21.0X SQL Parquet MR: DataPageV2 2348 2361 18 6.7 149.3 0.9X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2662 2675 19 5.9 169.3 0.8X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 104 113 10 151.4 6.6 21.2X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single SMALLINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2239 2240 2 7.0 142.3 1.0X SQL ORC Vectorized (Nested Column Disabled) 2193 2218 36 7.2 139.4 1.0X SQL ORC Vectorized (Nested Column Enabled) 269 301 31 58.4 17.1 8.3X SQL Parquet MR: DataPageV1 2502 2520 25 6.3 159.1 0.9X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 2836 2865 40 5.5 180.3 0.8X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 127 133 3 124.1 8.1 17.7X SQL Parquet MR: DataPageV2 2474 2475 1 6.4 157.3 0.9X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2753 2763 14 5.7 175.1 0.8X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 148 157 8 106.5 9.4 15.2X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single INT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2288 2329 58 6.9 145.5 1.0X SQL ORC Vectorized (Nested Column Disabled) 2281 2299 25 6.9 145.0 1.0X SQL ORC Vectorized (Nested Column Enabled) 306 318 10 51.4 19.5 7.5X SQL Parquet MR: DataPageV1 2527 2533 8 6.2 160.7 0.9X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 2942 2976 48 5.3 187.1 0.8X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 107 114 8 147.1 6.8 21.4X SQL Parquet MR: DataPageV2 2455 2467 17 6.4 156.1 0.9X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2857 2908 73 5.5 181.6 0.8X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 179 188 8 88.0 11.4 12.8X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single BIGINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2317 2338 29 6.8 147.3 1.0X SQL ORC Vectorized (Nested Column Disabled) 2298 2317 28 6.8 146.1 1.0X SQL ORC Vectorized (Nested Column Enabled) 298 317 17 52.7 19.0 7.8X SQL Parquet MR: DataPageV1 2953 2960 10 5.3 187.7 0.8X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3546 3582 52 4.4 225.4 0.7X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 295 300 8 53.4 18.7 7.9X SQL Parquet MR: DataPageV2 2410 2416 8 6.5 153.2 1.0X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2838 2874 52 5.5 180.4 0.8X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 186 193 9 84.4 11.9 12.4X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single FLOAT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2357 2367 14 6.7 149.8 1.0X SQL ORC Vectorized (Nested Column Disabled) 2356 2389 47 6.7 149.8 1.0X SQL ORC Vectorized (Nested Column Enabled) 371 398 23 42.3 23.6 6.3X SQL Parquet MR: DataPageV1 2485 2548 90 6.3 158.0 0.9X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 2902 2913 16 5.4 184.5 0.8X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 91 101 8 173.5 5.8 26.0X SQL Parquet MR: DataPageV2 2434 2440 8 6.5 154.8 1.0X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2822 2841 27 5.6 179.4 0.8X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 106 112 4 148.4 6.7 22.2X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Single DOUBLE Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 2544 2559 21 6.2 161.8 1.0X SQL ORC Vectorized (Nested Column Disabled) 2488 2532 62 6.3 158.2 1.0X SQL ORC Vectorized (Nested Column Enabled) 608 636 34 25.9 38.6 4.2X SQL Parquet MR: DataPageV1 2916 2920 6 5.4 185.4 0.9X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3526 3552 36 4.5 224.2 0.7X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 285 316 17 55.3 18.1 8.9X SQL Parquet MR: DataPageV2 2843 2845 3 5.5 180.8 0.9X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3452 3477 35 4.6 219.5 0.7X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 290 299 6 54.2 18.4 8.8X ================================================================================================ SQL Nested Column Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor SQL Nested Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------------------- SQL ORC MR 13510 13769 231 0.1 12884.2 1.0X SQL ORC Vectorized (Nested Column Disabled) 13456 13716 166 0.1 12832.9 1.0X SQL ORC Vectorized (Nested Column Enabled) 7076 7092 10 0.1 6748.4 1.9X SQL Parquet MR: DataPageV1 8655 8718 59 0.1 8254.4 1.6X SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 8969 9040 51 0.1 8553.1 1.5X SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 5455 5522 62 0.2 5202.7 2.5X SQL Parquet MR: DataPageV2 9576 9647 51 0.1 9132.4 1.4X SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 9926 10023 66 0.1 9466.0 1.4X SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 5198 5278 61 0.2 4957.3 2.6X ================================================================================================ Int and String Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Int and String Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 10416 10581 233 1.0 993.4 1.0X SQL Json 9767 9785 26 1.1 931.4 1.1X SQL Parquet Vectorized: DataPageV1 1939 1952 18 5.4 185.0 5.4X SQL Parquet Vectorized: DataPageV2 2058 2084 36 5.1 196.3 5.1X SQL Parquet MR: DataPageV1 4046 4076 42 2.6 385.9 2.6X SQL Parquet MR: DataPageV2 3943 3948 7 2.7 376.1 2.6X SQL ORC Vectorized 1721 1773 74 6.1 164.1 6.1X SQL ORC MR 3368 3388 28 3.1 321.2 3.1X ================================================================================================ Repeated String Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Repeated String: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 5421 5441 29 1.9 517.0 1.0X SQL Json 5934 5946 16 1.8 565.9 0.9X SQL Parquet Vectorized: DataPageV1 493 496 4 21.3 47.0 11.0X SQL Parquet Vectorized: DataPageV2 496 506 10 21.2 47.3 10.9X SQL Parquet MR: DataPageV1 1695 1700 8 6.2 161.6 3.2X SQL Parquet MR: DataPageV2 1639 1652 19 6.4 156.3 3.3X SQL ORC Vectorized 411 416 4 25.5 39.2 13.2X SQL ORC MR 1687 1689 3 6.2 160.9 3.2X ================================================================================================ Partitioned Table Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Partitioned Table: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative --------------------------------------------------------------------------------------------------------------------------------- Data column - CSV 10857 10945 124 1.4 690.3 1.0X Data column - Json 8252 8294 60 1.9 524.6 1.3X Data column - Parquet Vectorized: DataPageV1 113 130 13 139.6 7.2 96.3X Data column - Parquet Vectorized: DataPageV2 183 206 24 86.0 11.6 59.3X Data column - Parquet MR: DataPageV1 2344 2347 4 6.7 149.0 4.6X Data column - Parquet MR: DataPageV2 2249 2280 43 7.0 143.0 4.8X Data column - ORC Vectorized 194 215 26 81.2 12.3 56.1X Data column - ORC MR 2123 2133 14 7.4 135.0 5.1X Partition column - CSV 3614 3630 22 4.4 229.8 3.0X Partition column - Json 7132 7133 1 2.2 453.4 1.5X Partition column - Parquet Vectorized: DataPageV1 39 49 10 399.0 2.5 275.4X Partition column - Parquet Vectorized: DataPageV2 38 46 7 408.5 2.4 282.0X Partition column - Parquet MR: DataPageV1 1303 1309 8 12.1 82.9 8.3X Partition column - Parquet MR: DataPageV2 1299 1301 4 12.1 82.6 8.4X Partition column - ORC Vectorized 35 44 8 450.9 2.2 311.2X Partition column - ORC MR 1306 1317 15 12.0 83.0 8.3X Both columns - CSV 10857 10905 68 1.4 690.3 1.0X Both columns - Json 8823 8829 8 1.8 561.0 1.2X Both columns - Parquet Vectorized: DataPageV1 157 165 8 100.0 10.0 69.1X Both columns - Parquet Vectorized: DataPageV2 226 235 6 69.7 14.3 48.1X Both columns - Parquet MR: DataPageV1 2404 2407 4 6.5 152.9 4.5X Both columns - Parquet MR: DataPageV2 2349 2354 6 6.7 149.4 4.6X Both columns - ORC Vectorized 215 228 11 73.1 13.7 50.4X Both columns - ORC MR 2126 2149 32 7.4 135.2 5.1X ================================================================================================ String with Nulls Scan ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor String with Nulls Scan (0.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 6861 6864 5 1.5 654.3 1.0X SQL Json 8259 8262 4 1.3 787.7 0.8X SQL Parquet Vectorized: DataPageV1 1221 1229 13 8.6 116.4 5.6X SQL Parquet Vectorized: DataPageV2 1356 1357 1 7.7 129.4 5.1X SQL Parquet MR: DataPageV1 3673 3673 1 2.9 350.3 1.9X SQL Parquet MR: DataPageV2 3528 3540 18 3.0 336.4 1.9X ParquetReader Vectorized: DataPageV1 852 864 13 12.3 81.3 8.1X ParquetReader Vectorized: DataPageV2 939 974 35 11.2 89.6 7.3X SQL ORC Vectorized 704 731 38 14.9 67.2 9.7X SQL ORC MR 2830 2843 17 3.7 269.9 2.4X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor String with Nulls Scan (50.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 5556 5561 7 1.9 529.9 1.0X SQL Json 6911 6925 19 1.5 659.1 0.8X SQL Parquet Vectorized: DataPageV1 829 842 12 12.6 79.1 6.7X SQL Parquet Vectorized: DataPageV2 964 977 16 10.9 92.0 5.8X SQL Parquet MR: DataPageV1 2673 2676 3 3.9 255.0 2.1X SQL Parquet MR: DataPageV2 2716 2720 5 3.9 259.0 2.0X ParquetReader Vectorized: DataPageV1 728 730 2 14.4 69.4 7.6X ParquetReader Vectorized: DataPageV2 871 874 3 12.0 83.1 6.4X SQL ORC Vectorized 1010 1011 1 10.4 96.3 5.5X SQL ORC MR 2841 2849 11 3.7 270.9 2.0X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor String with Nulls Scan (95.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 3675 3677 2 2.9 350.5 1.0X SQL Json 4584 4587 4 2.3 437.2 0.8X SQL Parquet Vectorized: DataPageV1 178 180 3 58.9 17.0 20.7X SQL Parquet Vectorized: DataPageV2 197 200 2 53.2 18.8 18.7X SQL Parquet MR: DataPageV1 1706 1706 1 6.1 162.7 2.2X SQL Parquet MR: DataPageV2 1661 1683 31 6.3 158.4 2.2X ParquetReader Vectorized: DataPageV1 192 194 2 54.5 18.4 19.1X ParquetReader Vectorized: DataPageV2 210 212 1 49.8 20.1 17.5X SQL ORC Vectorized 333 345 11 31.5 31.8 11.0X SQL ORC MR 1614 1623 13 6.5 153.9 2.3X ================================================================================================ Single Column Scan From Wide Columns ================================================================================================ OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Single Column Scan from 10 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 1163 1168 7 0.9 1109.3 1.0X SQL Json 1702 1704 3 0.6 1623.2 0.7X SQL Parquet Vectorized: DataPageV1 30 38 7 34.7 28.8 38.5X SQL Parquet Vectorized: DataPageV2 36 40 7 29.1 34.4 32.2X SQL Parquet MR: DataPageV1 165 174 9 6.4 156.9 7.1X SQL Parquet MR: DataPageV2 165 179 13 6.4 157.1 7.1X SQL ORC Vectorized 33 39 7 31.6 31.6 35.1X SQL ORC MR 143 150 8 7.3 136.4 8.1X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Single Column Scan from 50 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 2576 2576 0 0.4 2457.0 1.0X SQL Json 6191 6193 3 0.2 5904.3 0.4X SQL Parquet Vectorized: DataPageV1 32 38 8 33.2 30.1 81.5X SQL Parquet Vectorized: DataPageV2 36 44 8 29.0 34.5 71.2X SQL Parquet MR: DataPageV1 171 179 8 6.1 163.1 15.1X SQL Parquet MR: DataPageV2 184 188 4 5.7 175.7 14.0X SQL ORC Vectorized 36 40 6 29.2 34.2 71.9X SQL ORC MR 145 152 6 7.2 138.3 17.8X OpenJDK 64-Bit Server VM 21.0.12+8-LTS on Linux 6.17.0-1020-azure AMD EPYC 9V74 80-Core Processor Single Column Scan from 100 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------ SQL CSV 4417 4427 15 0.2 4212.0 1.0X SQL Json 10550 10562 17 0.1 10061.7 0.4X SQL Parquet Vectorized: DataPageV1 41 53 10 25.6 39.0 107.9X SQL Parquet Vectorized: DataPageV2 42 47 6 25.2 39.7 106.1X SQL Parquet MR: DataPageV1 182 199 12 5.8 173.3 24.3X SQL Parquet MR: DataPageV2 187 195 3 5.6 178.4 23.6X SQL ORC Vectorized 42 52 9 25.0 40.1 105.1X SQL ORC MR 150 156 6 7.0 142.7 29.5X