• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

2 vs 4 ram sticks

Joined
Jun 3, 2012
Messages
2,091 (0.43/day)
Location
Denmark
Processor I5 13600kf
Motherboard Gigabyte Z790 D AX
Cooling Thermalright Assassin X120 R SE Black
Memory G.Skill Ripjaws S5 DDR5-6400 - 32GB - CL32 (6600mhz)
Video Card(s) Asus GeForce® RTX 5060 Ti 16GB Prime
Storage 1x2tb KC3000 & 2tb samsung 970 evo plus, 2 x 2 tb external usb harddrives
Display(s) LG 32GP850, IIyama G2470HSU-B1
Case Deepcool CG580
Audio Device(s) Yamaha R-N800A System audio signature 5 + Audio pro addon sub 1
Power Supply Corsair RM850X White
Mouse Asus Rog Gladius III Wireless Aimpoint
Keyboard Corsair K70 RGB TKL Champion
Software Windows 11 64 bit
I get mixed results watching ram,memory videos 2 vs 4 ram sticks

Foir a long time i have klnown 4 ram,memory stick top be able to give up to 10% more fps (as the highest results, claim)

Generally yes but some games use the cpu alot when having 4 ram stick,why ? Does the memory controller need it ?
2024-06-28 18_21_48-Ryzen 4 vs. 2 sticks RAM - YouTube – Google Chrome.jpg



Again yes

Here i def say no

ryzen 7 5700X3d +Corsair Vengeance LPX DDR4-3600 - 32GB - CL16 - Dual Channel (2 pcs) - Intel XMP+ a rtx 4070 dual evo playing at 2560x1440

2x16 gb extra ram, 4x16gb ram or 4x8gb ram i think a 5700X3d should handle 3600mhz cl 16 ram just fine

What does intel vs ryzen or ryzen vs intel ram,memory do on a ryzen 5000 when it's 3600mhz cl 16

I haven't been able to type in google what gives my results on using ryzen ram vs intel ram with a ryzen 5000 cpu, what it does performance wice
 
Last edited:
dual sided vs single sided sticks makes a difference in some systems
 
I get mixed results watching ram,memory videos 2 vs 4 ram sticks
You will get mixed results because different hardware and software combinations yield different results.

dual sided vs single sided sticks makes a difference in some systems
In "some" systems! In other systems, no difference. Or if any difference, it is totally insignificant.

Point being there are WAY WAY too many variables to come to a "one size fits all" conclusion. These include different motherboard/chipset/bus speeds, different CPUs, different GPUs, different video RAM, different system RAM, disk access, OS, security programs impact, networking, the specific game, and probably a dozens more.

What are you going to do if your motherboard only supports 2 sticks? Panic? Get rid of it and buy a 4 stick board?

And there's the budget too.

Too much emphasis it put on FPS scores where people start pulling their hair out because Joe over there got 150 and you only got 140FPS on benchmark scores. It is simply silly to have the fun of a game ruined simply because one "imagines" they see a difference based only on a score. :( Its the placebo effect in reverse. And, sadly, that really happens.

If all you care about is bragging rights on benchmarks, then fine. But I say, just enjoy the game.
 
2 and 4 ram sticks. are usually dual rank

I just wanna know if it makes a difference instead of having to buy ram, memory and try it for myself, because most video with ryzen say yes but some combinations it doesn't make a difference higher then 1-2 fps (fare from 10% like some videos claim)
 
I just wanna know if it makes a difference instead of having to buy ram, memory and try it for myself
:( You are not understanding my point. If you want to know if it makes a difference on YOUR computer, with YOUR hardware configuration and with YOUR games, you have no choice but to try it yourself on YOUR computer. Nobody else in the entire world has a computer exactly like yours. EVERY computer becomes unique within the first few minutes after being booted the very first time. This as we configure out network access, personalizations, security, our own apps and more.

I will say with everything else being the same, if you have 2 x 32GB or 4 x 16GB, any differences will be so minimal, you would not be able to "see" the difference in a double blind, side-by-side comparison running a variety of tasks - except, maybe, on benchmark scores.
 
The fastest configuration is 2x single-rank sticks, but that's only true if you're manually tuning your RAM to run as fast as possible, since 2x single rank is the easiest for your CPU's memory controller to drive at high frequencies and tight timings. If you're just running a DDR5-6000 kit at the default XMP/EXPO kit timings then the speed is determined by those timings, not the limit of your processor.

Your motherboard topology also plays a huge role in how fast your RAM will go.

If you're just buying a single XMP or EXPO kit and using that preset then don't worry about it. Just remember that those XMP/EXPO timings are only guaranteed if you use a single 2-stick kit. So you can't buy two kits, combine them together for more capacity and expect them to hit their rated speed. You might get lucky, but it's down to the silicon lottery of your CPU's memory controller, and a T-topology board helps if you have 4x DIMMs instead of 2x DIMMs, since T-Topology means traces to both slots in each channel are the same length.

Essentially, what you've asked is a very complicated question that depends on your CPU, motherboard, RAM type, kit type, BIOS version, and a little bit of luck. The easy answer is "fewer sticks is better"
 
For optimal performance on a Ryzen 5000 system you want two ranks per channel. Meaning you either run 4 single sided modules or two dual sided modules in your system. The reason for this is that rank interleaving improves performance. That works per memory channel and requires two ranks.

This assumes the same memory clock. If you can clock much higher with one single sided module per channel, you might see higher overall performance. But, clock for clock two ranks per channel wins over one rank per channel. 3-4 ranks complicate things.

And as mentioned above: There are many, many variables involved in this. So there is no simple answer unfortunately. With good ram you will run into Infinity Fabric (FCLK) limitations for example.
 
I know 2 ram sticks is best for high speeds but im not going to oc above my memorys stocks speed, atm it's 3600mhz cl 16 or above the new ram,memorys rated speed, which i think would be as it is, 3600mhz cl 16

How about the cpu usage difference and intel (xmp) vs amd (expo) ?
 
At the same clock and timings, dual rank will always be faster than single rank.

That goes for Intel and AMD.. I cannot speak for DDR5 :)
 
At the same clock and timings, dual rank will always be faster than single rank.

That goes for Intel and AMD.. I cannot speak for DDR5 :)

Binning is favorable to Single Rank, but depending on how good binning is you might see plenty of high quality Dual Rank kits. I'd say in general if going with Single Rank aim to populate 4-DIMM slots and if going Dual Rank aim to populate 2-DIMM slots. 4-Dimm slots with Dual Rank is rough on IMC so it's hard to recommend and I'd advise against it unless you've got a specific use case need for the added capacity that actually warrants it and not simply more equals better which is does sometimes. More coffee equals better facts are facts.
 
Interestingly when I converted my AMD machine to 2 dimms I still wasnt able to get it to run at the RAM factory tested speeds, although they are 32 gig dimms (64 gig total).
 
To go against the naysayers here a bit, I had good luck with four sticks of 8 GB single sided Hynix CJR based modules in my old rig with a 5800X in it.
They were running just fine at 3800 MHz, despite being sold as 3733 MHz modules and I manged to tighten up most timings a bit as well.
That said, as has been mentioned, there are no guarantees it'll work painlessly, as it comes down to your other hardware, especially the motherboard.
Was there any real performance benefits? Well, in the odd game maybe, but overall, not really, at least not unless the application needed more RAM.
 
Last edited:
Assuming all else is equal, what usually makes four DIMMs perform a little better is the additional ranks per channel.

DIMMs for consumer PCs are single rank or dual rank. Unless you're buying higher capacities, there's a chance they will be single rank. And if you go with two DIMMs, this means single rank per channel. So single rank is pretty common. Four DIMMs will guarantee dual rank per channel at the minimum, even if the DIMMs are single rank. Thus, the uplift. You can also have quad rank too (four dual rank DIMMs), but the extra gains are even more diminished, and the memory configuration gets harder to stabilize. I'm currently running such a configuration myself, but I didn't opt for it for performance reasons but simply for capacity reasons. Higher capacity DIMMs tend to be dual rank at first (then shift to single rank later), and you tend to need more DIMMs too.

I wouldn't necessarily focus ranks too much, but... I'd be lying if I said I wouldn't prefer to go dual rank if I could. Just remember that everything is a tradeoff. In the past, heavier memory configurations weren't so much heavier that you might have to fail meeting your desired "sweet spot" but with DDR5 and AM5 especially, that's changed a bit.
 
Don't load up 4 DIMMs unless you like spending time in the bios tweaking voltages to get it stable :) Neither AMD or Intel is good with 4 DIMMs unless you run a much lower speed.
I had this issue on TWO DIMMs very much stock clocked when this box was first built.
More sticks = more capacity.
Less sticks = less trouble during setup.
More ranks = more better.
That's all I can really guarantee on this topic.

Four sticks allegedly takes more effort to setup properly but that's not the case with what I had to deal with at all on this X570 board and there may be something more to the SiC639 voltage controllers that isn't quite plain to me but I managed to reduce latency under 70ns on 4x16GB DDR4-3200 DIMMs and cranking them ↗3800, even loosening the timings a bit. This is kind of insane.
1719601584460.png


We have no idea if you're able to do the same but considering that your kit starts at 3600 as the floor, confidence is pretty high no matter the board topology.
 
You shouldn't really encounter much instability unless pushing 4 DIMM's, Dual Rank, and/or higher capacity kit's and at higher frequencies with tighter latency. The CPU binning, MB quality, and memory kit qualities all play some role as well. As for which is which is most important definitively probably MB quality is most important followed by the other two given it sits in between the two connecting them, but idk speculation.
 
To go against the naysayers here a bit, I had good luck with four sticks of 8 GB single sided Hynix CJR based modules in my old rig with a 5800X in it.
They were running just fine at 3800 MHz, despite being sold as 3733 MHz modules and I manged to tighten up most timings a bit as well.
That said, as has been mentioned, there are no guarantees it'll work painlessly, as it comes down to your other hardware, especially the motherboard.
Was there any real performance benefits? Well, in the odd game maybe, but overall, not really, at least not unless the application needed more RAM.
CJR was good stuff though. Not quite as legendary as Samsung B-Die but when you saw CJR show up in Thaiphoon you knew you had a good chance of a fast setup.
 
32 gb extra 100 euros or a little more lets say 50 euros for my ram if i buy a 64 gb kit 130 euros or mabye 100 or just a little more for 4x8gb

G skill ripjaws V are super cheap
 
To go against the naysayers here a bit, I had good luck with four sticks of 8 GB single sided Hynix CJR based modules in my old rig with a 5800X in it.
They were running just fine at 3800 MHz,
currently running four 8GB sticks as well @ 3800 mhz
 
2x16 gb extra ram ... 5700X3d should handle 3600mhz cl 16 ram just fine
Your MB does support "2DPC 2R max speed 3600 MHZ" so go ahead if you feel like you do need that extra 2x16GB memory.
 
this is an older test on techspot but I think some of what they found still holds true


Installing four memory modules in a dual-channel system can improve performance in certain scenarios. However it should be made clear that this won't automatically translate into a 5-7% performance boost in games. For that to occur you need to be more CPU limited than GPU limited and that's rarely the case when using good graphics quality settings or hardware configurations.


For example, we were testing with a GeForce RTX 2080 Ti at 1080p. Increasing the resolution to 1440p will likely result in neutralizing any benefit we just saw with the two extra modules as games become predominantly GPU bound. The same would be true at 1080p using a lesser graphics card.

It's also worth noting that four modules can also limit memory performance in the sense that you might not be able to achieve the same frequency and timings that you could with just two modules. In other words, a straight comparison becomes quite difficult to make as there are a number of variables that need to be considered.


The Ryzen 9 3900X is limited to DDR4-2933 when using four modules, whereas AMD's official spec says the CPU will handle DDR4-3200 with just two modules. Depending on how good the silicon is, the integrated memory controller might do better than that. Our chip handles right up to DDR4-3600 CL16 memory with four modules, but whatever the limit, the point is, it will go higher with just two modules.


Dual-ranked modules introduce a new set of limits, though none were explored in this article. AMD lists the official dual-ranked DIMM support as DDR4-3200 for two modules, but just DDR4-2666 when using four modules. It would be interesting to see how two and four dual-ranked modules compare in these dual-channel systems, but that's a test for another day.
 
this is an older test on techspot but I think some of what they found still holds true


Installing four memory modules in a dual-channel system can improve performance in certain scenarios. However it should be made clear that this won't automatically translate into a 5-7% performance boost in games. For that to occur you need to be more CPU limited than GPU limited and that's rarely the case when using good graphics quality settings or hardware configurations.


For example, we were testing with a GeForce RTX 2080 Ti at 1080p. Increasing the resolution to 1440p will likely result in neutralizing any benefit we just saw with the two extra modules as games become predominantly GPU bound. The same would be true at 1080p using a lesser graphics card.

It's also worth noting that four modules can also limit memory performance in the sense that you might not be able to achieve the same frequency and timings that you could with just two modules. In other words, a straight comparison becomes quite difficult to make as there are a number of variables that need to be considered.


The Ryzen 9 3900X is limited to DDR4-2933 when using four modules, whereas AMD's official spec says the CPU will handle DDR4-3200 with just two modules. Depending on how good the silicon is, the integrated memory controller might do better than that. Our chip handles right up to DDR4-3600 CL16 memory with four modules, but whatever the limit, the point is, it will go higher with just two modules.


Dual-ranked modules introduce a new set of limits, though none were explored in this article. AMD lists the official dual-ranked DIMM support as DDR4-3200 for two modules, but just DDR4-2666 when using four modules. It would be interesting to see how two and four dual-ranked modules compare in these dual-channel systems, but that's a test for another day.
I just don't have a ryzen 3000 cpu

I have a fairly new ryzen 5700X3d (when it was released,built)
 
Another older article about it, but I haven't seen anyone re-test on the Ryzen 5000-series.
I guess the assumptions is that it'll be very similar.
 
Back
Top