diff --git a/pom.xml b/pom.xml
index 5762dc2289e..0f69d5ae01c 100644
--- a/pom.xml
+++ b/pom.xml
@@ -1588,5 +1588,19 @@
 			<artifactId>fastdoubleparser</artifactId>
 			<version>0.9.0</version>
 		</dependency>
+
+		<dependency>
+			<groupId>org.openjdk.jmh</groupId>
+			<artifactId>jmh-core</artifactId>
+			<version>1.37</version>
+			<scope>test</scope>
+		</dependency>
+		<dependency>
+			<groupId>org.openjdk.jmh</groupId>
+			<artifactId>jmh-generator-annprocess</artifactId>
+			<version>1.37</version>
+			<scope>test</scope>
+		</dependency>
+
 	</dependencies>
 </project>
diff --git a/src/main/java/org/apache/sysds/runtime/matrix/data/LibMatrixMult.java b/src/main/java/org/apache/sysds/runtime/matrix/data/LibMatrixMult.java
index 9763e4ea57d..282a1db9c58 100644
--- a/src/main/java/org/apache/sysds/runtime/matrix/data/LibMatrixMult.java
+++ b/src/main/java/org/apache/sysds/runtime/matrix/data/LibMatrixMult.java
@@ -2051,7 +2051,7 @@ private static void matrixMultSparseSparseVM(SparseBlock a, SparseBlock b, Dense
 				int blen = b.size(aix[k]);
 				int[] bix = b.indexes(aix[k]);
 				double[] bvals = b.values(aix[k]);
-				vectMultiplyAdd(avals[k], bvals, cvals, bix, bpos, 0, blen);
+				vectMultiplyAddScatter(avals[k], bvals, cvals, bix, bpos, 0, blen);
 			}
 	}
 	
@@ -2069,7 +2069,7 @@ private static long matrixMultSparseSparseSparseMM(SparseBlock a, SparseBlock b,
 			for(int k = apos; k < apos+alen; k++) {
 				int aixk = aix[k];
 				if( b.isEmpty(aixk) ) continue;
-				vectMultiplyAdd(avals[k], b.values(aixk), tmp,
+				vectMultiplyAddScatter(avals[k], b.values(aixk), tmp,
 					b.indexes(aixk), b.pos(aixk), 0, b.size(aixk));
 				hitNonEmpty = true;
 			}
@@ -2100,7 +2100,7 @@ private static void matrixMultSparseSparseMMSmallRHS(SparseBlock a, SparseBlock
 			for(int k = apos; k < apos+alen; k++) {
 				int aixk = aix[k];
 				if( b.isEmpty(aixk) ) continue;
-				vectMultiplyAdd(avals[k], b.values(aixk), cvals,
+				vectMultiplyAddScatter(avals[k], b.values(aixk), cvals,
 					b.indexes(aixk), b.pos(aixk), cix, b.size(aixk));
 			}
 		}
@@ -2132,7 +2132,7 @@ private static void matrixMultSparseSparseMM(SparseBlock a, SparseBlock b, Dense
 					int k = curk[i-bi] + apos;
 					for(; k < apos+alen && aix[k]<bkmin; k++) {
 						if( b.isEmpty(aix[k]) ) continue;
-						vectMultiplyAdd(avals[k], b.values(aix[k]), cvals,
+						vectMultiplyAddScatter(avals[k], b.values(aix[k]), cvals,
 							b.indexes(aix[k]), b.pos(aix[k]), cix, b.size(aix[k]));
 					}
 					curk[i-bi] = k - apos;
diff --git a/src/test/java/org/apache/sysds/test/component/matrix/BenchmarkMatMult.java b/src/test/java/org/apache/sysds/test/component/matrix/BenchmarkMatMult.java
new file mode 100644
index 00000000000..a175a50639e
--- /dev/null
+++ b/src/test/java/org/apache/sysds/test/component/matrix/BenchmarkMatMult.java
@@ -0,0 +1,83 @@
+package org.apache.sysds.test.component.matrix;
+
+import java.util.concurrent.TimeUnit;
+import org.apache.sysds.runtime.matrix.data.LibMatrixMult;
+import org.apache.sysds.runtime.matrix.data.MatrixBlock;
+import org.apache.sysds.test.TestUtils;
+import org.openjdk.jmh.annotations.*;
+import org.openjdk.jmh.infra.Blackhole;
+import org.openjdk.jmh.results.format.ResultFormatType;
+import org.openjdk.jmh.runner.Runner;
+import org.openjdk.jmh.runner.RunnerException;
+import org.openjdk.jmh.runner.options.Options;
+import org.openjdk.jmh.runner.options.OptionsBuilder;
+
+@State(Scope.Thread)
+@BenchmarkMode(Mode.AverageTime)
+@OutputTimeUnit(TimeUnit.MICROSECONDS)
+@Warmup(iterations = 5, time = 1, timeUnit = TimeUnit.SECONDS)
+@Measurement(iterations = 10, time = 1, timeUnit = TimeUnit.SECONDS)
+@Fork(1)
+public class BenchmarkMatMult {
+
+	private MatrixBlock left;
+	private MatrixBlock right;
+	private MatrixBlock result;
+
+	@Param({"1024", "2048", "4096", "8192"})
+	public int m;
+
+	@Param({"1"})
+	public int cd;
+
+	@Param({"0.5", "0.75", "1.0"})
+	public double sparsityLeft;
+
+	@Param({"0.001", "0.01", "0.1", "0.2"})
+	public double sparsityRight;
+
+	@Param({"1024", "2048", "4096", "8192"})
+	public int n;
+
+	@Setup(Level.Trial)
+	public void setup() {
+		left = TestUtils.ceil(TestUtils.generateTestMatrixBlock(m, cd, -10, 10, sparsityLeft, 13));
+		if (!left.isInSparseFormat()) {
+			left.sparseToDense();
+			left.denseToSparse(true);
+		}
+
+		MatrixBlock rightOriginal = TestUtils.ceil(TestUtils.generateTestMatrixBlock(cd, n, -10, 10, sparsityRight, 14));
+		right = new MatrixBlock();
+		right.copy(rightOriginal, false);
+		if (!right.isInSparseFormat()) {
+			right.sparseToDense();
+			right.denseToSparse(true);
+		}
+
+		result = new MatrixBlock(m, n, true);
+		result.allocateBlock();
+	}
+
+	@Setup(Level.Iteration)
+	public void clearResult() {
+		result.reset(m, n, true);
+	}
+
+	@Benchmark
+	public void benchmarkSparseSparse(Blackhole bh) {
+		LibMatrixMult.matrixMult(left, right, result);
+		bh.consume(result);
+	}
+
+	public static void main(String[] args) throws RunnerException {
+		String tag = "vector-api-mult-sparse-sparse-" + System.currentTimeMillis();
+		Options opt = new OptionsBuilder()
+			.include(BenchmarkMatMult.class.getName())
+			.jvmArgs("--add-modules=jdk.incubator.vector")
+			.result(tag + ".csv")
+			.resultFormat(ResultFormatType.CSV)
+			.build();
+		new Runner(opt).run();
+	}
+}