Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Tech Wizard

Dual vs Quad Channel Memory in gaming?

Recommended Posts

Posted · Original PosterOP

Hello,

 

I'm wondering how big the difference is in gaming with Dual Channel Memory vs Quad Channel Memory. Does games even benefit from quad channel, or is it only productivity applications that sees a boost?
Are we talking a few FPS, or 10+? Didn't find any gaming benchmarks for dual vs quad channel :(


Main:  1650 v2   @ 4,6GHz   -   X79 Deluxe                -   GTX 1080 @ 2000MHz   -   24GB DDR3 @ 2400MHz / CL10

Side:   i7-4790K @ 4,5GHz   -   Maximus 7 Hero        -   GTX 1070 @ 2114MHz    -  16GB DDR3 @ 2666MHz / CL12

 

HWBOT

Link to post
Share on other sites

I assume that because faster memory speeds help in CPU bound scenarios, higher memory bandwidth will help as well.


CPU: Intel Core i7-5820K | Motherboard: AsRock X99 Extreme4 | Graphics Card: Gigabyte GTX 1080 G1 Gaming | RAM: 16GB G.Skill Ripjaws4 2133MHz | Storage: 1 x Samsung 840 Series 128GB | 1 x Seagate 1TB | 1 x WD Blue 500GB | PSU: Seasonic M12II Evo 620W | Case: Phanteks Enthoo Pro (White) | Cooling: Arctic Freezer i32

 

 

GTX 1060 vs RX 480 (old)

Link to post
Share on other sites
3 minutes ago, Tech Wizard said:

Didn't find any gaming benchmarks for dual vs quad channel

Yea because it means jack shit. You're better off investing in a better CPU/GPU or larger SSD or just save that money. If it had a significant increase in gaming performance then a lot of people would be investing in it.

Link to post
Share on other sites
Posted · Original PosterOP
Just now, dfg666 said:

Yea because it means jack shit. You're better off investing in a better CPU/GPU or larger SSD or just save that money.

I'm not buying. I just want to learn about the differences, if there are any.


Main:  1650 v2   @ 4,6GHz   -   X79 Deluxe                -   GTX 1080 @ 2000MHz   -   24GB DDR3 @ 2400MHz / CL10

Side:   i7-4790K @ 4,5GHz   -   Maximus 7 Hero        -   GTX 1070 @ 2114MHz    -  16GB DDR3 @ 2666MHz / CL12

 

HWBOT

Link to post
Share on other sites

there is exactly zero difference in games...


Threadripper 1950x

Gigabyte X399 Gaming 7

32GB DDR4 Corsair Vengeance

MSI GTX 1080Ti

Fractal Design Define S

Corsair RM1000i

Link to post
Share on other sites
1 minute ago, dfg666 said:

Yea because it means jack shit. You're better off investing in a better CPU/GPU or larger SSD or just save that money. If it had a significant increase in gaming performance then a lot of people would be investing in it.

If you are on X99 or X79, I am pretty sure that you already have a high-ish end CPU and GPU... Quad channel is not that common because the only platforms that use it are X99 and X79 ;)


CPU: Intel Core i7-5820K | Motherboard: AsRock X99 Extreme4 | Graphics Card: Gigabyte GTX 1080 G1 Gaming | RAM: 16GB G.Skill Ripjaws4 2133MHz | Storage: 1 x Samsung 840 Series 128GB | 1 x Seagate 1TB | 1 x WD Blue 500GB | PSU: Seasonic M12II Evo 620W | Case: Phanteks Enthoo Pro (White) | Cooling: Arctic Freezer i32

 

 

GTX 1060 vs RX 480 (old)

Link to post
Share on other sites

Ram bandwidth in general (not only limited to 2 vs 4 channels) does make a difference, but usually not a big one. They way to look at it is, ram can bottleneck the CPU, in a similar way the CPU can bottleneck the GPU. It will depend on the load at the time and wont affect all scenarios. I'd love to see quad channel more mainstream as 4x slow ram will be cheaper and more compatible than 2x fast ram of the same capacity.


Main rig: Asus Maximus VIII Hero, i7-6700k stock, Noctua D14, G.Skill Ripjaws V 3200 2x8GB, Gigabyte GTX 1650, Corsair HX750i, In Win 303 NVIDIA, Samsung SM951 512GB, WD Blue 1TB, HP LP2475W 1200p wide gamut

Gaming system: Asrock Z370 Pro4, i7-8086k stock, Noctua D15, Corsair Vengeance LPX RGB 3000 2x8GB, Gigabyte RTX 2070, Fractal Edison 550W PSU, Corsair 600C, Optane 900p 280GB, Crucial MX200 1TB, Sandisk 960GB, Acer Predator XB241YU 1440p 144Hz G-sync

Ryzen rig: Asrock B450 ITX, R5 3600, Noctua D9L, G.SKill TridentZ 3000C14 2x8GB, Gigabyte RTX 2070, Corsair CX450M, NZXT Manta, WD Green 240GB SSD, LG OLED55B9PLA

VR rig: Asus Z170I Pro Gaming, i7-6700T stock, Scythe Kozuti, Kingston Hyper-X 2666 2x8GB, Zotac 1070 FE, Corsair CX450M, Silverstone SG13, Samsung PM951 256GB, Crucial BX500 1TB, HTC Vive

Gaming laptop: Asus FX503VD, i5-7300HQ, 2x8GB DDR4, GTX 1050, Sandisk 256GB + 480GB SSD

Total CPU heating: i7-8086k, i3-8350k, i7-7920X, 2x i7-6700k, i7-6700T, i5-6600k, i3-6100, i7-5930k, i7-5820k, i7-5775C, i5-5675C, 2x i7-4590, i5-4570S, 2x i3-4150T, E5-2683v3, 2x E5-2650, E5-2667, R7 3700X, R5 3600, R5 2600, R7 1700

Link to post
Share on other sites
23 minutes ago, Tech Wizard said:

I'm not buying. I just want to learn about the differences, if there are any.

It's incremental. Up to 3% maybe.

Link to post
Share on other sites
42 minutes ago, Tech Wizard said:

Hello,

 

I'm wondering how big the difference is in gaming with Dual Channel Memory vs Quad Channel Memory. Does games even benefit from quad channel, or is it only productivity applications that sees a boost?
Are we talking a few FPS, or 10+? Didn't find any gaming benchmarks for dual vs quad channel :(

 

Guys, there is almost no difference between dual or quad channel... BUT there is a huge difference in performance when using dual channel over single channel. There are many videos out there showing this to be true

 

Link to post
Share on other sites
1 minute ago, TheDane said:

Guys, there is almost no difference between dual or quad channel... BUT there is a huge difference in performance when using dual channel over single channel. There are many videos out there showing this to be true

 

can you show said videos?

Link to post
Share on other sites
5 minutes ago, TheDane said:

that video showed there was pretty much no difference between the single and dual channel the in the video editing related stuff it was less then a 10 second change and in the game benchmarks there was less than a 2 fps difference 

Link to post
Share on other sites
5 minutes ago, Alaradia said:

that video showed there was pretty much no difference between the single and dual channel the in the video editing related stuff it was less then a 10 second change and in the game benchmarks there was less than a 2 fps difference 

 
 
 

Little difference is still a difference.... and saving 10 seconds everytime you encode or process something will mean something after some time, and since a dual stick kit is about the same price as a single stick of the same capacity than it is just stupid not to get dual over single channel, but getting quad over dual is really only to get more GB of ram in your system

Link to post
Share on other sites

most people don't encode also most of the time by going dual channel your sacrificing a lane you could put a 16 gig stick in later unless you're already buying the biggest stick that can go in your dimm already

Link to post
Share on other sites
2 hours ago, Tech Wizard said:

Are we talking a few FPS, or 10+? Didn't find any gaming benchmarks for dual vs quad channel :(

Apparently nobody benchmarks it because we already know quad-channel memory doesn't do anything for gaming, but without any benchmarks how do we know that? That seems like a bit of circular logic to me. I've personally seen a few benchmarks of this, but they're mostly from back in 2011 when quad-channel was a new feature.

 

We know that memory frequency and latency can translate to meaningful gains in certain games, even if such games are outliers. The number of channels isn't exactly the same thing, but lots of games have very unique performance characteristics and I find it very hard to believe there is not a single game out there where you can measure some difference between dual- and quad-channel memory.

Link to post
Share on other sites
On 5/11/2017 at 3:28 PM, typographie said:

Apparently nobody benchmarks it because we already know quad-channel memory doesn't do anything for gaming, but without any benchmarks how do we know that? That seems like a bit of circular logic to me. I've personally seen a few benchmarks of this, but they're mostly from back in 2011 when quad-channel was a new feature.

 

We know that memory frequency and latency can translate to meaningful gains in certain games, even if such games are outliers. The number of channels isn't exactly the same thing, but lots of games have very unique performance characteristics and I find it very hard to believe there is not a single game out there where you can measure some difference between dual- and quad-channel memory.

Sory for necroing, but I registered specifically to agree with you here. You're absolutely right, and I've been looking for benchmarks the same way you have (I assume you haven't found any yet as I haven't either).


In general a lot of the benchmarks are inconclusive because they're not done right. What is the use looking at scenario's that are not memory-bottlenecked and then declaring it doesn't matter? That tells you nothing. Reviewers should look at scenes where you'd expect a difference. And then, they also have to focus on the right numbers. Average fps completely obscures stutter, but it ruins your game experience. The ARMA benchmark you linked is actually very good in this respect.

 

At least the idea that RAM speed is irrelevant has died. There's some discussion on latency vs frequency, but consensus seems to be that both matter with a slight edge to frequency (super low timings eg C12 3200 is about equal to C16 3600 in fps). I discovered on reddit that single rank vs dual rank is also a thing, and that dual rank sticks supposedly are faster everything else being equal (but have more trouble clocking higher, at least on ryzen). The best ram that is still moderately affordable is something like samsung B-die G SKill TridentZ 3200 C14.

 

Gamers Nexus did a video on single vs dual channel and found that you'll get about ~ 5-10% upgrading from single to dual channel (but that's avg fps, so actual impact may be larger). PCworld did an article about dual channel to quad channel and saw no significant results (but again only looked at avg fps).


I personally do not expect that it makes ZERO difference going to quad channel. But whether it is beneficial or harmfull likely comes down to latency vs bandwith. Quad channel gives you insane theoretical bandwith, but often the amount of data the cpu needs will probably not be so large. For smaller amounts of data, latency may be where most time is lost. And quad channel has slightly more latency, so it's hard to predict.

Arma would be a very interesting benchmark in this regard. It's known (in ARMA 3) that Broadwell C (5775C) with it's L4 cache (only 128MB but that's a lot for cache) improves AVERAGE fps by up to 25%. That's an insane increase in performance. I would like to see this YAAB benchmark with the 5775C, that would be very informative.

But in conclusion -  you're 100% spot on. We do NOT have the required data to say for certain that quad channel does not help in any games. I would love if someone dit a YAAB benchmark like the article you linked for (fast) quad channel ram vs dual channel ram. And maybe also for single vs dual channel. But these are the kind of articles that may have too much of a niche following to really be worth writing for professional benchmarkers. You have to really love the tech itself to be motivated to do it.
 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×