The built-in function "sum" of Python 2.7 does run faster. But it's not true in PyPy.

SumTest.py

import itertools

# tstime(https://bitbucket.org/gsauthof/tstime) is used to measure the performance.
# OS: Archlinux 3.6.10-1
# Python Version: 2.7.3
# PyPy Version: 2.0.0-beta1

c = 0
for (i, j) in itertools.product(range(10000), repeat=2):
    c += i * j
print c
# Python2 Result:
# 2499500025000000
#
# Exit status: 0
#
# pid: 513 (python2) started: Thu Dec 20 17:50:00 2012
#         real  44.859 s, user  44.787 s, sys   0.010s
#         rss     3556 kb, vm     9424 kb
#
# PyPy Result:
# 2499500025000000
# 
# Exit status: 0    
# 
# pid: 517 (pypy) started: Thu Dec 20 18:33:55 2012
#         real  16.566 s, user  16.463 s, sys   0.030s
#         rss    19424 kb, vm    66876 kb


print sum([i * j for (i, j) in itertools.product(range(10000), repeat=2)])
# Python2 Result:
# 2499500025000000 
#
# Exit status: 0  
#
# pid: 555 (python2) started: Thu Dec 20 17:51:58 2012
#         real  40.591 s, user  39.310 s, sys   1.223s
#         rss  1575516 kb, vm  1610280 kb
# 
# PyPy Result:
# 2499500025000000
#
# Exit status: 0    
#
# pid: 499 (pypy) started: Thu Dec 20 18:33:13 2012
#         real  23.189 s, user  18.393 s, sys   1.180s
#         rss   812232 kb, vm   859600 kb

def f():
    for (i, j) in itertools.product(range(10000), repeat=2):
        yield i * j
print sum(f())
# Python2 Result:
# 2499500025000000
#
# Exit status: 0
#
# pid: 538 (python2) started: Thu Dec 20 17:51:19 2012
#         real  32.456 s, user  32.393 s, sys   0.003s
#         rss     3624 kb, vm     9424 kb
# 
# PyPy Result:
# 2499500025000000
#
# Exit status: 0    
# 
# pid: 529 (pypy) started: Thu Dec 20 18:34:15 2012
#         real  20.083 s, user  20.027 s, sys   0.027s
#         rss    19432 kb, vm    67008 kb