Overclocking and Performance
The Overclock
Overclocking this memory kit was relatively straight forward. While increasing to 2666MHz was easy, trying to achieve a higher stable result was hard. 2800MHz was achieved but failed at the first stress test. The timings were slightly increased to 17-16-16 to remain system stability.
Cinebench
While Cinebench focuses more on the CPU performance, memory tweaks do play a small role in the increase and decrease of performance. In this test, in particular, the overclock presented a worse result, however, the result is well within the margin of error for the CPU.
AIDA64
Aida64 is on of the more thorough and simpler to read tests that we perform. The overclock gave some better figures here, although it still lags behind the Kingston 2666MHz 32GB set.
Memory latency is key for the execution of any data stored on the RAM. While a 4 nanosecond drop isn’t considered much, that is around a 5% decrease in latency speed
SiSoft Sandra
SiSoft Sandra is slightly harder to read, but the performance figures show a truer representation of the overall performance of the RAM itself. Between stock and overclocked, there is a gain of around 4GB/s, but it still falls short of the Kingston memory.
WPrime
WPrime is fairly sensitive to memory frequency changes, however, due to the high performance of the CPU; tangible differences are hard to see. In this test, similar to Cinebench, the performance was impacted negatively; however, this is well within the margin of error for the CPU score.
3DMark Fire Strike
There is very little impact to gaming performance when selecting your memory, although a difference could be witnessed between a 2133MHz and 3300+MHz kit. The Physics test within 3DMark Firestrike is the most sensitive to memory changes, but as displayed below, a jump of 266MHz didn’t give a massive performance boost.
