Is 3200 mhz considered pretty fast RAM in 2021?
Nope. Even value RAM already 3200mhz but comes with higher latency (CAS Latency).
What is the fastest speed available these days?
5000mhz or higher. They don’t come cheap. 3600-4000mhz prices are more reasonable.
I got a 4x8gb ddr4 4000/cl18 kit for $189
Im on 4200mhz and my coworker is 4600mhz so I will say no and that its pretty much the bare minimum.
The answer to this question is going to depend heavily on what processor and motherboard Iran you’re on. It’s all about if the processor you’re on and motherboard can handle the faster Ram. So you need to start by telling us that
3200 is where the value is. Anything else and you’ll be paying a lot more, especially since you can overclock the 3200.
I would highly advise against buying the bare minimum AKA 3200mhz.
The truth of the matter is that if you were on a ryzen processor 3600 CL16 is the Sweet Spot. Going beyond that on ryzen you won’t notice much of a difference in frames per second in gaming.
However that is completely untrue with Intel. Intel doesn’t care much about latency timings as it does pure raw Ram speed. Intel processors drink up Ram speed like water in the desert. So yes you will notice a drastic difference between the bare minimum 3200mhz and 4200mhz ( the Intel sweet spot) on Intel
For gaming it doesn,t mather ram speed get capped pretty fast.
I can’t post link sadly but you can make various research and the result is always the same. Tt barely does anything pass 3200hz for almost any gaming build.
It really depends on a lot of things; most tests that do this still show gpu limited scenarios or games that typically are not CPU limited.
Throw it on a cpu limited game like most mmos, and memory speed can be pretty significant.
If you play a bunch of Doom Eternal, memory speed might not matter too much. But games like WoW well see improvement.
This is untrue man and I really wish people would stop spreading misinformation you probably got this info based someone elses misinformation that was passed on to you.
Ram speed absolutely does matter in gaming and so do CL timings. Gamer Nexus made a video on this and they got almost 15-20 more frames per second just by adjusting their speed and CL timings. On AMD processors older than the 5000 series CL timings are HUGE! Intel not so much.
As I said above this is going to be highly dependent on what motherboard and CPU you have. If you have a crappy cheap motherboard then yes what you said above is most likely true. However if you are on AMD and you have a high-end B550 or X570 board (such as the MSI or Asus gaming motherboards) or you have a high-end Z490 board on Intel? You will absolutely notice a difference by upping your RAM speed/CL timings.
For example I am on the Z490 Maximus Hero. The difference between 3600mhz and 4200mhz which I am currently on is night and day.
Lastly another thing that can improve your RAM speed with all of this being said especially on AMD is using a 4 ram stick kit over a 2.
In FF14 on my 5.1ghz core/4.8ghz cache 10900KF going from 4x8 3200/cl16 to 4x8 4000/cl18, in the busiest hub (Limsa Lominsa Aetherite crystal) which is totally CPU limited (~30-40% GPU usage), I went from around 85fps to 100. That’s pretty significant IMO.
Type ram speed for gaming in google next time b4 you spread your misinformation.
I trust tomshardware tech more then you.
The problem isn’t the information it’s how you’ve interpreted it.
“RAM speeds don’t matter” - this is true sometimes. GPU limited scenarios? Yeah, this is true.
The more CPU limited you are, and the more memory sensitive the application you’re using is, the more memory speed matters. In the case of average framerate when you are at or close to GPU limitation, you’re right in that it won’t matter too much.
The issue is we are on a WoW forum and we’re focused on this game. And in games like WoW and FF14, where often your GPU is half asleep due to game limitations, it is frequently the CPU that is the bottleneck. In these cases, memory speed matters significantly.
Beyond my own anecdotal experiences, here’s some other outlets who review this information and have drawn similar conclusions:
Here’s another showing memory performance (single vs. dual rank configuration) making a significant difference when CPU limited:
There is also server and network bottleneck as well. You’ll get it from world boss just after weekly reset with 100+ players.
It would also help if you understood the information you were reading. Also when Tom’s Hardware ran their test they were on a standard run-of-the-mill motherboard.
Tom’s Hardware does not use higher-end Hardware when doing their testing. GamerNexus does which is why they often get different results. This is why many of us have stated what motherboard you are on is crucial in whether or not it’s going to matter.
If you understood the technology you were using you would understand that the motherboard is the skeleton and central nervous system of your computer.
You go with a crappy run-of-the-mill skeleton and you’re not going to be able to run like Jesse Owens or may not even be able to run at all. Heck if you go with a crappy enough skeleton you’re going to be in a wheelchair your entire life I hope this analogy helps you understand.
This is why when you build a computer the main place you don’t want to try and cut corners on and you want to get the best component you possibly can on is your motherboard
The motherboard only really matters in terms of if it can support more memory modules without failing (some just can’t handle 4 DIMMS outside of JDEC spec) and if it can support higher memory speeds - if at the same memory speed and timings on different motherboards there’s no real difference in performance.
It just so happens that higher end motherboards tend to be more successful at doing those two things.
There’s other factors like thermals, PCB layers, and voltage efficiency, but if you’re judging clock for clock it should not matter.
Thank you for better explaining what I was trying to say I’m not always the best with words
The problem with a lot of reviews is they’re testing these things where they won’t really matter.
They almost always use “1080p/ultra” as their reference, when even with a 2080ti/3090 often in new games they can still hit GPU limitation.
You also get canned benchmark tests, which aren’t really helpful because especially in multiplayer games, they remove that dynamic element where CPU horsepower and throughput matters.