[GitHub] [spark] LuciferYang commented on a diff in pull request #43065: [SPARK-45287][TESTS] Add Java 21 benchmark result and update Java 17 ones
LuciferYang commented on code in PR #43065: URL: https://github.com/apache/spark/pull/43065#discussion_r1334916319 ## sql/core/benchmarks/SortBenchmark-results.txt: ## @@ -2,15 +2,15 @@ radix sort -OpenJDK 64-Bit Server VM 1.8.0_362-b09 on Linux 5.15.0-1031-azure -Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure +Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz radix sort 2500: Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -reference TimSort key prefix array12374 12403 41 2.0 495.0 1.0X -reference Arrays.sort 3377 3381 5 7.4 135.1 3.7X -radix sort one byte 209212 2119.5 8.4 59.2X -radix sort two bytes398403 3 62.8 15.9 31.1X -radix sort eight bytes 1538 1538 0 16.3 61.5 8.0X -radix sort key prefix array1953 1998 64 12.8 78.1 6.3X +reference TimSort key prefix array14141 14208 96 1.8 565.6 1.0X Review Comment: ditto ## sql/core/benchmarks/ColumnarBatchBenchmark-results.txt: ## @@ -2,58 +2,58 @@ Int Read/Write -OpenJDK 64-Bit Server VM 1.8.0_362-b09 on Linux 5.15.0-1031-azure -Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure +Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz Int Read/Write: Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -Java Array 257266 8 1273.6 0.8 1.0X -ByteBuffer Unsafe 480490 5682.0 1.5 0.5X -ByteBuffer API 1994 1996 2164.4 6.1 0.1X -DirectByteBuffer756762 7433.6 2.3 0.3X -Unsafe Buffer 255263 4 1283.1 0.8 1.0X -Column(on heap) 266272 6 1231.5 0.8 1.0X -Column(off heap)526529 2623.1 1.6 0.5X -Column(off heap direct) 258265 7 1270.3 0.8 1.0X -UnsafeRow (on heap) 556560 6589.0 1.7 0.5X -UnsafeRow (off heap)599606 5546.9 1.8 0.4X -Column On Heap Append 478488 6686.0 1.5 0.5X +Java Array 254261 5 1290.1 0.8 1.0X +ByteBuffer Unsafe 420427 8780.2 1.3 0.6X +ByteBuffer API 801822 28409.0 2.4 0.3X +DirectByteBuffer661668 7495.8 2.0 0.4X +Unsafe Buffer 253266 10 1296.0 0.8 1.0X +Column(on heap) 254261 4 1292.2 0.8 1.0X +Column(off heap)255261 5 1287.3 0.8 1.0X +Column(off heap direct) 253258 6 1297.3 0.8 1.0X +UnsafeRow (on heap) 722729 9454.1 2.2 0.4X +UnsafeRow (off heap)532543 13616.3 1.6 0.5X +Column On Heap Append 516
[GitHub] [spark] LuciferYang commented on a diff in pull request #43065: [SPARK-45287][TESTS] Add Java 21 benchmark result and update Java 17 ones
LuciferYang commented on code in PR #43065: URL: https://github.com/apache/spark/pull/43065#discussion_r1334909906 ## sql/catalyst/benchmarks/GenericArrayDataBenchmark-results.txt: ## @@ -1,10 +1,10 @@ -OpenJDK 64-Bit Server VM 1.8.0_362-b09 on Linux 5.15.0-1031-azure -Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure +Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz constructor: Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -arrayOfAny4 4 0 2491.5 0.4 1.0X -arrayOfAnyAsObject 256257 1 39.1 25.6 0.0X -arrayOfAnyAsSeq 18 18 0551.9 1.8 0.2X -arrayOfInt 536537 1 18.7 53.6 0.0X -arrayOfIntAsObject 788794 10 12.7 78.8 0.0X +arrayOfAny7 7 0 1495.4 0.7 1.0X +arrayOfAnyAsObject7 7 0 1495.3 0.7 1.0X +arrayOfAnyAsSeq 201202 1 49.8 20.1 0.0X Review Comment: The results of `arrayOfAnyAsSeq` have undergone significant changes and need attention (it may be a known issue, but I can't remember the details). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] [spark] LuciferYang commented on a diff in pull request #43065: [SPARK-45287][TESTS] Add Java 21 benchmark result and update Java 17 ones
LuciferYang commented on code in PR #43065: URL: https://github.com/apache/spark/pull/43065#discussion_r1334909865 ## sql/catalyst/benchmarks/GenericArrayDataBenchmark-results.txt: ## @@ -1,10 +1,10 @@ -OpenJDK 64-Bit Server VM 1.8.0_362-b09 on Linux 5.15.0-1031-azure -Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure +Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz constructor: Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -arrayOfAny4 4 0 2491.5 0.4 1.0X -arrayOfAnyAsObject 256257 1 39.1 25.6 0.0X -arrayOfAnyAsSeq 18 18 0551.9 1.8 0.2X -arrayOfInt 536537 1 18.7 53.6 0.0X -arrayOfIntAsObject 788794 10 12.7 78.8 0.0X +arrayOfAny7 7 0 1495.4 0.7 1.0X +arrayOfAnyAsObject7 7 0 1495.3 0.7 1.0X Review Comment: The results of `arrayOfAnyAsSeq` have undergone significant changes and need attention (it may be a known issue, but I can't remember the details). ## sql/catalyst/benchmarks/GenericArrayDataBenchmark-results.txt: ## @@ -1,10 +1,10 @@ -OpenJDK 64-Bit Server VM 1.8.0_362-b09 on Linux 5.15.0-1031-azure -Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure +Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz constructor: Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -arrayOfAny4 4 0 2491.5 0.4 1.0X -arrayOfAnyAsObject 256257 1 39.1 25.6 0.0X -arrayOfAnyAsSeq 18 18 0551.9 1.8 0.2X -arrayOfInt 536537 1 18.7 53.6 0.0X -arrayOfIntAsObject 788794 10 12.7 78.8 0.0X +arrayOfAny7 7 0 1495.4 0.7 1.0X +arrayOfAnyAsObject7 7 0 1495.3 0.7 1.0X Review Comment: The results of `arrayOfAnyAsSeq` have undergone significant changes and need attention (it may be a known issue, but I can't remember the details). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
[GitHub] [spark] LuciferYang commented on a diff in pull request #43065: [SPARK-45287][TESTS] Add Java 21 benchmark result and update Java 17 ones
LuciferYang commented on code in PR #43065: URL: https://github.com/apache/spark/pull/43065#discussion_r1334908931 ## core/benchmarks/ZStandardBenchmark-results.txt: ## @@ -2,26 +2,26 @@ Benchmark ZStandardCompressionCodec -OpenJDK 64-Bit Server VM 1.8.0_372-b07 on Linux 5.15.0-1041-azure +OpenJDK 64-Bit Server VM 17.0.8+7-LTS on Linux 5.15.0-1046-azure Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz Benchmark ZStandardCompressionCodec:Best Time(ms) Avg Time(ms) Stdev(ms)Rate(M/s) Per Row(ns) Relative -- -Compression 1 times at level 1 without buffer pool293 327 85 0.0 29283.2 1.0X -Compression 1 times at level 2 without buffer pool322 324 2 0.0 32184.8 0.9X -Compression 1 times at level 3 without buffer pool453 456 2 0.0 45285.1 0.6X -Compression 1 times at level 1 with buffer pool 171 173 1 0.1 17065.2 1.7X -Compression 1 times at level 2 with buffer pool 208 209 1 0.0 20786.5 1.4X -Compression 1 times at level 3 with buffer pool 334 335 2 0.0 33350.3 0.9X +Compression 1 times at level 1 without buffer pool 2800 2801 2 0.0 279995.2 1.0X Review Comment: From what I remember, the results of this microbenchmark are always unstable. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org - To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org