Stop wasting money on fast Ram!! 7200MHz vs 4800MHz...

People always ask me if fast ram is worth it for gaming... well today we test the fastest set of DDR5 I have vs the stock DDR5 non XMP settings in 3 resolutions to see if there is any performance gain!
 
Sponsored Links:

○○○○○○ Items featured in this video available at Amazon ○○○○○○

••• Follow me on your favorite Social Media! •••

[dgant]

definitely do the comparison with ddr4 and ddr5. most people are still on ddr4

[EliteJr]

I wish we could have seen 1% lows between the three as well.

[musicianJKEE]

1% and 0.1% lows are absent in ram tests. Thanks Jay :)

[justin_ashburn]

Would have loved to see this for AMD systems as well, X3D and non-X3D to see how much the cache can negate the impact of ram speeds (if any).

[enzicoxe]

really? no 1% lows? im baffled you left out THE most imporatnt reason to go with faster memory.

[nikolakatanic]

I remember 3 years back when I purchased my PC and doing researches about this. I went with DDR4 on 3200Mhz which was the sweet spot between cost and benefit. All above were too much more investing in small percentage in performance..

[keblin]

Please do it for DDR4 comparing to DDR5!

[clifflenoir]

Maybe for the next video you could keep one set of fast ram with good timings and just change the Frequency in the BIOS. That way we would know if the differences were down to frequency or timings. You could then compare 3 sticks at the same speed (e.g. 6000MHz) with different timings (CL30, CL 34 and CL 40) and see what difference THAT makes. Thanks for the video

[mesicek]

Jay doing what he does best. Being clueless about stuff he knows feck all about.

[xforlyfe]

I have never thought of RAM as a part of my system that significantly increases FPS. I've always thought of it in terms of gameplay smoothness i.e. 1% lows and 0.1% lows. THAT'S where higher clocked RAM shows it's value.

[Llun]

Yes Jay... do a DDR4 speed test too and tell me what mouse are you using in the video? I like those headlights...

[Piotr_Majchrzak]

Definitely DDR4 tests are very welcome and this time please include 0, 1% and 1% lows, I suspect them to show us what the deal withe the RAM really is.

[Pulverrostmannen]

I pretty much always buy higher clocked ram because I typically lower the clock speed and overclock the timings with the extra headroom I get instead, I had very good results with this

[taft]

The more and more I learn about PC hardware and gaming, the less and less I trust Jay...

[Manterok]

He says this...but I bet his PC has 7200mhz of ram speed lol

[oldmanian]

Yes. I should save my $$ for a monitor that has a motor in it. Lol 😂

[kasimirdenhertog]

As you said, this is about gaming performance. I recently switched from running DDR5-5600 to DDR5-8000 and for productivity/general tasks, it does make a noticeable and measurable difference. This is especially down to latency. Also interesting is that the gain I got from frequency (running XMP profile) was doubled when I tuned the timings. So RAM tuning (watch Buildzoid’s videos for this) is actually worth-wile.

[Tomcat_kanal]

Man, this was already tested by Hardware Unboxed like a year ago. And their result was that Intel processors are actually more sensitive to the RAM frequency but not much sensitive to the timings BUT Ryzen 7000 processors are not that sensitive to the frequency, as they cannot benefit much from DDR5 faster than 6000MHz but they benefit from low timings and not only from the low Primary timings but especially from the Secondary timings when tuned as low as possible that makes a huge difference in performance!!

[darkesthour]

Jay, you shouldn’t be using AI OC in this scenario. You absolutely cannot guarantee that the OC would be the same as you swap from one ram kit to another. That’s a variable in your benchmarks that needs to be avoided by principle, big or small.

[Profetorum]

I don't really understand what's going on here. If the idea is to show what's the average use case of people turning on XMP and ignore everything, then sure. Otherwise... weird results. Motherboard is not specified, we don't know if all the configurations have been tested for stability (XMP is not granted to work)