Benchmark Testing
SiSoft Sandra 2015 SP2b
At 2133MHz stock speeds, the Vengeance LPX DDR4-3200 modules matched performance with the other tested modules. The bandwidth increased by almost 50% with the memory running at its rated 3200MHz speed. While the Dominator modules performed with a higher bandwidth, they required a non-stock base clock setting while the Vengeance LPX modules ran at their rated speed easily at the stock base clock speed.
AIDA64 Extreme Edition
The Vengeance LPX DDR-3200 kit performance scaling mimics that seen in the Sandra memory benchmarking tests with the kit's performance increasing by almost 50% with the modules set to run at their 3200MHz rated speed.
7-Zip
7-Zip is an open-source file compression and decompression software. File compression and decompression are very system and memory intensive, acting as a good indicator for system performance. The 7-Zip benchmark test was run with a 192MB dictionary size, 8 CPU threads active, and for 5 passes per test. The included benchmark test was repeated three times with the highest repeatable MIPS (Millions of Instructions Per Second) scores recorded.
The performance results better demonstrate the real-world performance differences that you are likely to see when using the Vengeance LPX modules under review. The module performance scales in a similar fashion to that seen in the previous tests when the modules are run at their rated speeds. However, the Vengeance LPX DDR4-3200 modules performance increased by almost 15% instead of the 50% increase seen in the memory-specific benchmarks.
Seems bizarre to me where you
Seems bizarre to me where you basically need to use water cooling for your CPU or a stock heat sink (which do not exist for the socket 2011 processors) in order to use the air flow kit for the RAM.
We’ve gone from “water cooling your RAM isn’t necessary” to this? If I have a custom loop in my PC I’m probably going to look into water cooling the RAM instead of something like this.
It’s ugly, really.
I don’t see the need to have
I don’t see the need to have any type of cooling on RAM, even more so with the 1.2v stock voltage of DDR4. Even heat-spreaders are more for product differentiation and marketing than any real necessity.
Watercooling your RAM is just adding restriction to the loop and added cost for zero real-world benefit. All imho of course.
Thanks for the review.
Can
Thanks for the review.
Can you add some benchmarks for multitasking? Sort of the way some reviewers test SSDs with multiple things happening at once. I think there could be some real gains there with faster ram that we just aren’t seeing with these single task benchmarks.
I run my computers in the ground with multiple VMs in flight compiling code, for instance. How much better would the increase in speed help scenarios like that with the CPU running overdrive?
“running at there rated 1.35V
“running at there rated 1.35V voltage”
*their
fixed, thanks…
fixed, thanks…