FEATURED Kingston HyperX Impact 2x8GB DDR4-2400 and 2x4GB DDR4-2133 SODIMM

[Woomack]

I'm usually making threads/reviews about single memory kits where I'm presenting how memory is overclocking and performing. In this case it's hard to talk about overclocking as I wish to present laptop DDR4 memory which is nearly impossible to overclock because of platform, not because of memory itself.
I just received 2x8GB DDR4-2400 HyperX Impact memory kit which I wish to compare to the one I already have for a while - 2x4GB DDR4-2133 also HyperX Impact.

Because of my test platform limitations I can't perform tests at DDR4-2400 settings so both memory kits will run at DDR4-2133 13-13-13 what is already much tighter setting than JEDEC standard for DDR4-2133 memory. Laptops are usually equipped with DDR4-2133 CL15-15-15 memory and often only single memory stick.
Laptop used for tests is ASUS ROG GL552VW based on Skylake i5 6300HQ so real 4 cores but no HT. It also has HM170 chipset which is limiting factor for higher memory kits. Intel locked it at DDR4-2133 specs. Even though Kingston used SPD and XMP profiles at DDR4-2400 then laptop is forcing DDR4-2133 settings, at least at tighter timings ( what are near the timings of DDR4-2133 kit ).












Screenshots below are including basic SPD/XMP profiles. In this case it's the same even though there are programmed both profiles.

First is always 2x4GB 2133 kit.

 

[Woomack]

Additional tests

3DMark Fire Strike








3DMark Fire Strike Extreme








3DMark Fire Strike Ultra








ASUS RealBench








x256 Benchmark - 4K








HEVC Decode Benchmark

 

[ssjwizard]

Interesting. Look forward to seeing the results.

 

[Johan45]

Make that lappy squeal Bart

 

[Woomack]

Results added to first posts.

It's hard to say which memory is better. Looks like in general tests 2x8GB will be faster.
In 3DMark there is already noticable difference in Extreme/Ultra presets. On the other hand barely anyone will decide to play in 4K resolution on any laptop
Cinebench R15 is also faster on 2x8GB while some other tests perform slightly better on 2x4GB. I assume it's because of secondary timings which are always more relaxed on higher capacity memory kits.
Realbench is total mistake. Every couple of runs results are +/- 10k so it's useless in general performance comparison. On both kits after couple of runs I can make the same results. Every like 10 runs there are 5-10K higher scores.