Performance
The performance will be tested on an Intel Haswell platform which includes a Core i7 4690K CPU and a ASUS Maximus VII Gene motherboard.
The benchmarks used in the performance stage are popular programs available to download on the manufacturer’s sites so that every reader can easily compare results on the home/office computer.
AIDA64 Cache and Memory Benchmark
Starting from AIDA64 we already see that performance of HyperX Savage is quite good. It’s not the best in our comparison table but results are not much lower than the fastest memory series of the competition. We have to remember that Kingston has also higher memory series than Savage.
Numbers are really respectable considering that 2400 Patriot memory is running at much tighter timings.
MaxxMem Preview v1.99
MaxxMem Preview has similar set of tests like AIDA64 but based on a single thread. In this test HyperX Savage looks even better. Maximum write bandwidth is clearly limited by processor and memory controller’s speed in single threaded operations.
HyperPi 32M
HyperPi so the overclockers benchmark. This benchmark loves fast memory running at tight timings. Since HyperX Savage has quite relaxed timings then is clearly losing to Patriot memory. On the other hand we can see huge difference between Avexir 1600 memory which is running at still one of the most popular settings on the market. Here we see that memory at higher clock actually counts.
Cinebench R15
Cinebench R15 so the latest version of popular rendering benchmark. We can’t see here big differences in clock but HyperX Savage 2400 memory achieved best result what clearly means it’s well balanced for daily work.
PCMark 8
PCMark 8 is one more example that HyperX Savage does its work right. Results in this benchmark are about as good as on the more expensive kits from the competition. PCMark is testing whole PC so it’s one of the best benchmarks showing how theoretically small hardware changes may affect daily work.
The HyperX Savage 2400 CL11 memory is performing really good comparing to other memory kits.
All of those who wish for higher performance can try overclocking. Let us check how our memory performs at a higher clock than declared by Kingston.
Overclocking
Overclocking is never guaranteed so the presented results may vary from results on other memory kits. I am not recommending overclocking if you do not know what are you doing. High voltages may damage hardware and it will not be covered by warranty.
Overclocking has been performed on the Devil’s Canyon i7 4790K CPU and ASUS Maximus VII Gene motherboard with the latest official BIOS/UEFI.
Presented result has been made on standard voltage of 1.65V so no additional voltage adjustment was required.
As we see overclocking potential of HyperX Savage is really high. I could not stabilize a clock of 2933 at reasonable timings so I left it at 2800. Also worth to mention is that 2666 clock was possible without any changes in timings or voltages above what we can see in XMP1 profile. All that from a standard, retail memory kit.
Even though the reviewed memory is already the highest version of HyperX Savage, it still has a lot of headroom which is a really good sign especially for all overclockers.
For all of those who are interested, inside of the HyperX Savage 2400 CL11 we can find Hynix IC based on BFR series. It is a new bin of BFR PBA which as we see is scalling well at reasonable voltages.
Removing the heatsink can damage memory and it will not be covered by warranty. I did it so that you will not have to try. The heatsinks included with HyperX Savage are good enough to keep memory cool even after overclocking so any cooling changes are not required.
