Super Pi 1M / 32M, wPrime
Meanwhile Super
Pi even in its latest version has become pretty old. Never the less it is able
to show scaling to the clock speed nicely. Following we will also run Super Pi
with different memory frequencies as well as different latencies. This shows the
influence of the memory subsystem.
1* : DDR3 1600Mhz, 8-8-8-24 / DDR2 1066Mhz, 5-5-5-15
2* : DDR3 1333Mhz, 8-8-8-24 / DDR2 800Mhz, 5-5-5-15
3* : DDR3 1333Mhz, 7-7-7-21 / DDR2 800Mhz, 4-4-4-12
Super Pi 1M
SuperPi Mod 1.5 XS 1M |
1* / 2* / 3* |
Percent |
ASUS Crosshair IV Extreme |
21.341s / 21.419s / 21.466s |
99.08 % |
ASUS M4A78-HTPC (DDR2) |
21.606s / 21.621s / 21.591s |
100.00 % |
DFI LanParty DK 790FXB-M3H5 |
21.590s / 21.699s / 21.637s |
100.16 % |
ECS A890GXM-A2 |
21.653s / 21.700s / 21.653s |
102.34 % |
|
lower is better |
wPrime 32M
wPrime is based on a
multithreaded algorithm to calculate the square root to up to 32 billion digits.
Therefore it is possible to put 100 percent load on every core or thread and to
see differences between architectures.
WPrime v1.55 32M |
1* / 2* / 3* |
Percent |
ECS A890GXM-A2 |
8.143s / 8.190s / 8.190s |
96.54 % |
ASUS Crosshair IV Extreme |
8.345s / 8.035s / 8.362s |
98.32 % |
ASUS M4A78-HTPC (DDR2) |
8.237s / 8.580s / 8.362s |
100.00 % |
DFI LanParty DK 790FXB-M3H5 |
8.656s / 8.221s / 8.858s |
102.28 % |
|
lower is better |
Discuss this article in the forums