Improving cache locality for thread-level speculation | IEEE Conference Publication | IEEE Xplore