Difference in terms of memory between TEST_F and TEST_P

I have a project with different levels of testing - UT, MT, Pytests. UTs & MTs are done using gtest library, and for module testing both TEST_F and TEST_P are used. The tests(>1000, needed for coverage) are split into 80 shards. In the last few months more and more parameterized tests have been introduced, and valgrind has started to fail more and more often with signal 9 to the point where now it continuously fails. My question is, can valgrind issues be related to parameterized tests? Are there any difference in terms of ram consumption? Both SetUp() and TearDown() functions are called for each instance of the parameterized test, so at least from that point of view, they seem to work the same as TEST_F(but this may be a wrong assumption)