2080ti performance hit from Ram capacity/speed?

[SpookyCitrus]

I've built a couple 2080ti systems in the last year, however both were for 4K video editing and 3D Animation setups, and on both of the rigs I did 64GB of 3200mhz Ram. I recently purchased a 2080ti for my gaming rig because I got a deal to good to pass up on. I currently have an i7-8700k at 5ghz and 16gb of Corsair Vengeance LPX at 3000mhz. What I'm wondering is when I get the 2080ti would I be fine with keeping the ram I currently have or would it be better to upgrade to say 32GB of 3200mhz Ram? I only use my system for 1440p 144hz gaming, mainly AAA titles on max settings and a few competitive shooters and some video editing mainly just game clips and such nothing higher than 2K resolution 60fps. I haven't tested ram speed and capacity with a 2080ti before I don't really think it will be an issue with what I already have but I'd like to make sure, and the few articles/ forum posts I have seen have been kind of mixed on what's better so I thought I'd ask all the fine people on here and hopefully receive some info instead of giving it this time  thanks everyone!

[Zando_]

Just now, Genwyn said:

I dont think you'll experience any noticeable performance change with a 200mhz ram speed difference. Thats a very very small speed change, and even if it was larger its generally pretty difficult to find a difference between the 3000mhz+ range of ram speeds even in ram intensive synthetics.

 

The 2080ti's graphical performance and capability isnt dependent on the system memory.

^^^ Buying new, 3200Mhz is the better option due to basically no price difference between that and 3000Mhz. If you already have 3000Mhz, just tune it a bit and you should push 3200Mhz CL16 just fine. 

You usually don't need 32GB RAM for gaming, maybe stuff like Star Citizen pushes a lot but there's very few games like that. I still have 32GB though, I just wanted to use quad channel on my X99 board but didn't want sticks lower than 4GB (so if I ever cannibalize my rig I have two 2x8GB kits). 

[SpookyCitrus]

Just now, Genwyn said:

I dont think you'll experience any noticeable performance change with a 200mhz ram speed difference. Thats a very very small speed change, and even if it was larger its generally pretty difficult to find a difference between the 3000mhz+ range of ram speeds even in ram intensive synthetics.

 

The 2080ti's graphical performance and capability isnt dependent on the system memory.

And that's what I thought. I've never seen ram capacity affect a GPU's performance before, I could see if I were to go from 3000mhz to 4000mhz but 200mhz extra is negligible at best. I had a friend tell me this and then saw some toms hardware posts about it and it confused me I don't really have that much experience with 2080ti's besides install and driver setup so I wanted to make sure I wasn't wrong.

[SpookyCitrus]

2 minutes ago, Zando Bob said:

^^^ Buying new, 3200Mhz is the better option due to basically no price difference between that and 3000Mhz. If you already have 3000Mhz, just tune it a bit and you should push 3200Mhz CL16 just fine. 

You usually don't need 32GB RAM for gaming, maybe stuff like Star Citizen pushes a lot but there's very few games like that. I still have 32GB though, I just wanted to use quad channel on my X99 board but didn't want sticks lower than 4GB (so if I ever cannibalize my rig I have two 2x8GB kits). 

I knew the capacity difference between 16 and 32 affecting the 2080ti was bs.  I'm fine with my 16GB and I can just OC to 3200mhz I just wanted to double check. Thanks!

[jones177]

I think it is a tuff one.

I have 2 similar setups. Both have 3200mhz ram. 

Videos like this helped me decide.

 

[SpookyCitrus]

27 minutes ago, jones177 said:

I think it is a tuff one.

I have 2 similar setups. Both have 3200mhz ram. 

Videos like this helped me decide.

-snip-

I'm fairly familiar with ram speeds and what to use for what tasks but the whole "16GB of ram will have lower performance with the 2080ti than 32GB." and "the 2080ti requires 32GB of Ram to run at peak performance." BS some forums had on them and even a buddy of mine said it, made it to where I was pretty skeptical. I've fixed/built a lot of computers and I've never heard of ram capacity affecting the performance of a GPU. Ram speed yes but capacity no.

[jones177]

2 minutes ago, SpookyCitrus said:

I'm fairly familiar with ram speeds and what to use for what tasks but the whole "16GB of ram will have lower performance with the 2080ti than 32GB." and "the 2080ti requires 32GB of Ram to run at peak performance." BS some forums had on them and even a buddy of mine said it, made it to where I was pretty skeptical. I've fixed/built a lot of computers and I've never heard of ram capacity affecting the performance of a GPU. Ram speed yes but capacity no.

The performance of the CPU helps the performance of the GPU and the ram helps the performance of the CPU but as you can see in the video, not by much.

 

I bought 2 3200mhz 16gb sticks a few weeks ago.

Since I have 2 systems each with a i7 8086ks at 5ghz and 2080 tis I will put 32 in one and leave 16 in the other for a while to see the difference for myself. 

 

 

[SpookyCitrus]

31 minutes ago, jones177 said:

The performance of the CPU helps the performance of the GPU and the ram helps the performance of the CPU but as you can see in the video, not by much.

 

I bought 2 3200mhz 16gb sticks a few weeks ago.

Since I have 2 systems each with a i7 8086ks at 5ghz and 2080 tis I will put 32 in one and leave 16 in the other for a while to see the difference for myself. 

I plan on OCing my kit to 3200mhz when I get home. If you can test if 32 really does make a difference over 16 with GPU performance(from what I know it shouldn't affect it) that would be great, I'm really interested to see if it does.

[jones177]

1 minute ago, SpookyCitrus said:

I plan on OCing my kit to 3200mhz when I get home. If you can test if 32 really does make a difference over 16 with GPU performance(from what I know it shouldn't affect it) that would be great, I'm really interested to see if it does.

I will start tomorrow.

I have been putting it off since I have to remove a Noctua Nh-d15 to get to the ram. Next time it will be an Aio for me.

[SpookyCitrus]

5 minutes ago, jones177 said:

I will start tomorrow.

I have been putting it off since I have to remove a Noctua Nh-d15 to get to the ram. Next time it will be an Aio for me.

The Nh-d15 is a nice cooler but just like the Dark Rock Pro 4 they make it pretty much impossible to get to the ram if you need to without taking them off, that's one of the main reasons I've done Aio's for most of my builds.

Message me or post back on here after you've done some testing, I'm curious on how it will go.

[Tristerin]

The amount of ram isn't going to impact the performance of the 2080ti.  The CPU's ability to build a window each second to maximize the GPUs potential is the issue here.  As far as I know, Intel is not nearly as impacted by higher frequency ram compared to Ryzen.  For Ryzen Ive run my own tests from 2133mhz kit, a 3200mhz kit, and a 3600mhz kits that I own, and with Ryzen...it matters a whole lot (for gaming, which is my use case)

[jones177]

13 hours ago, SpookyCitrus said:

The Nh-d15 is a nice cooler but just like the Dark Rock Pro 4 they make it pretty much impossible to get to the ram if you need to without taking them off, that's one of the main reasons I've done Aio's for most of my builds.

Message me or post back on here after you've done some testing, I'm curious on how it will go.

Ok.

 

All 3 games 3840 X 1600 at ultra except low anti aliasing and no DoF. 

-------------------------------------------------------------

16gb

Heaven 1440p 
3188
126.5fps average

 

32gb

Heaven 1440p
3194
126.9 average

---------------------------------------------------------

16gb

Time Spy
14406
Graphic score 16197
G test 1 = 103.84fps
G test 2 = 94.23fps
CPU score 9000
CPU test 30.24fps

 

32gb

Time Spy
14595
Graphic score 16333
G test 1 = 104.67
G test 2 =  95.07
CPU score 9105
CPU test 31fps

----------------------------------------------------------

16gb

Shadow of the Tomb Raider with RTX 
64 fps

 

32gb

Shadow of the Tomb Raider with RTX 
64 fps

--------------------------------------------------------

16gb

Metro Exodus with RTX
50fps

 

32gb

Metro Exodus wit RTX
50fps

--------------------------------------------------------

16gb

Assassin's Creed Odyssey
Ultra 63fps 

 

32gb

Assassin's Creed Odyssey
Ultra 66fps 

--------------------------------------------------------------

Window likes the upgrade more than the games. 

 

 

 

 

 

 

 

 

 

[SpookyCitrus]

3 hours ago, jones177 said:

-snip-

This is pretty much what I was expecting, I was positive the amount of ram wouldn't affect the performance in games. Thank you for testing it.