It takes the top spot from Titan, a Cray XK7 system installed at the U.S. Department of Energy's (DOE) Oak Ridge National Laboratory, which drops to number two in the latest list.
The Milky Way 2 recorded 33.86 petaflops on the Linpack benchmark, compared to the Titan's score of 17.59 petaflops.
The two top machines rely on very different architectures. Milky Way 2 has 16,000 nodes, each with two Intel Xeon IvyBridge processors and three Xeon Phi processors, for a combined total of 3,120,000 computing cores.
Meanwhile the Titan relies on 261,632 NVIDIA Tesla K20x GPU accelerator cores. Titan is one of the most energy efficient systems on the list, consuming a total of 8.21 MW and delivering 2,143 Mflops per Watt.
Milky Way-2 will be deployed at the National Supercomputer Center in Guangzho, China by the end of the year.
Sequoia, an IBM BlueGene/Q system installed at DOE's Lawrence Livermore National Laboratory, also dropped one position and is now at number three in the list. Sequoia was first delivered in 2011 and has achieved 17.17 petaflops on the Linpack benchmark using 1,572,864 cores.
Sequoia is also one of the most energy efficient systems on the list, consuming a total of 7.84 MW and delivering 2,031.6 Mflops per Watt.
Fujitsu's "K computer" installed at the RIKEN Advanced Institute for Computational Science (AICS) in Kobe, Japan, is the number four system, with a performance of 10.51 petaflops on the Linpack benchmark using 705,024 SPARC64 processing cores.
A second BlueGene/Q system, Mira, installed at Argonne National Laboratory is at number five with 8.59 petaflops on the Linpack benchmark using 786,432 cores.
Rounding out the Top 10 are the upgraded Stampede at the Texas Advanced Computing Center of the University of Texas, Austin; JUQUEEN at the Forschungszentrum Juelich in Germany - the most powerful system in Europe; an IBM BlueGene/Q system at Lawrence Livermore National Laboratory; SuperMUC, an IBM iDataplex system installed at Leibniz Rechenzentrum in Germany; and Tianhe-1A at the National Supercomputing Center in Tianjin, China.
A total of 54 systems on the list are using accelerator/co-processor technology, down from 62 in the last list in November 2012. Thirty-nine of these use NVIDIA chips, three use ATI Radeon, and 11 systems use Intel Xeon Phi.
Total combined performance of all 500 systems has grown to 223 petaflops, compared to 162 petaflops six months ago and 123 petaflops one year ago.
The results of the list were revealed at the International Supercomputing Conference in Leipzig in Germany yesterday.