Overclocking a Geforce GTX260 with nVidia nTune
The Sparkle GTX260 Core 216 videocard comes from the
factory with its nVidia GT200 GPU clocked at 576 MHz and 896MB of GDDR3 memory
running at a cool 999 MHz. PCSTATS has generally had great luck overclocking nVidia videocards, so we were moderately excited about
pushing Sparkle's Geforce GTX260 to infinity and beyond.
Overclocking was done through nVidia's nTune utility
which unlocks GPU and memory clock speed. PCSTATS started with the GPU first,
pausing to test the results with a quick 3D benchmark before continuing.
Starting with the
GeForce GTX260 GPU the clock speed was increased in 25MHz increments from
575MHz, in no time at all it hit 607MHz, then 629... 667... 697... and finally
749MHz which caused the system to lock up and crash. After dialing the GPU clock
speed back a few degrees the Sparkle GTX260 eventually found it's feet at 682MHz
Next came the 896MB of GDDR3 memory which was clocked by
default at 1000 MHz. The memory was overclocked in 50 MHz increments from 112MHz
to 1156... 1167... 1188.. 1210... to 1254MHz. Unfortunately while the card would
run, tearing was visible in 3DMark06 so we had to dial back the memory speed to
1221MHz. A nice overclock of 221MHz, but far from a world record.
Together the nVidia GTX260 GPU and GDDR3 memory ran
comfortably overclocked at 682MHz GPU / 1221MHz memory without crashing or any
obvious signs of artifacts in the 3D benchmarks.
The details of how the Sparkle GTX260 Core 216 videocard
(model: SXX260896D35-VP) test system was configured for benchmarking; the
specific hardware, software drivers, operating system and benchmark versions are
indicated below. In the second column are the general specs for the reference
platforms this nVidia GeForce GTX260 896MB based videocard is to be compared
against. Please take a moment to look over PCSTATS test system configurations
before moving on to the individual benchmark results on the next page.
|PCSTATS Test System Configurations|
Benchmark results are organized by GPU manufacturer
first (AMD/ATI or nVidia), then by GPU generation, and then by GPU class (high
end, mainstream, value). This approach provides a clearer view of how
performance can differ from generation to generation, and class to previous
generation. The product being tested is marked with the red colour bar.
||Use the power of Google to search all of PCSTATS and the PCSTATS Forums. Tell us what you think of this new feature - FEEDBACK?|