[SOLVED] LOW PERFOMANCE RTX 3080 BLACK

[Goiabaman]

Hello, I would like to share and if possible receive some help. I recently bought an RTX 3080 BLACK, but, compared to several benchmarks on the internet, since the same model and similar models, I have been observing that my card has a much lower performance than in these benchmarks. For example, COD WARZONE, I have an average of 80 to 90 frames; Red Dead Redemption a minimum of 29fps, an average of 86 and a maximum of 112.

Components of my PC:
CPU: RYZEN 5 3600X
VGA: RTX 3080 BLACK
MOTHERBOARD: ROG Strix B450-F Gaming, AMD AM4, ATX, DDR4
XPG Core Reactor, 750W, 80 Plus Gold Modular
2x8 DDR4 2400MHZ HyperX Fury
SSD Adata XPG SX6000, 1TB, M.2 NVMe

 

[rgd1101]

check gpu temp?
ddu the old gpu driver before install nvidia driver?

 

[Goiabaman]

The temperature varies a lot, in the warzone it is around 70 degrees, but in Red Dead Redemption 2 it is 78 to 80 degrees, the region where I live is very hot, an average of 28 degrees. Yes, it is the latest version of NVIDIA drivers. In fact, I forgot to add Warzone, I play in 1080p and RD2 I play in 2560 × 1080.

 

[dorsai]

Your GPU temps are normal.

What benchmark results are you looking at ? Did you mirror their settings exactly ? I ask because if you have settings like anti-aliasing maxed and they didn't...or if they ran DLSS and you did not...results will vary wildly depending on game.

Also since you're playing at 1080p your cpu can really become a deciding factor in how high your FPS will be. Most online reviews run very high end CPUs in their reviews. Your 3600x is a solid gaming cpu but it won't match a 5900x or i9-10900k in gaming.

Additionally your ram choice of 2400mhz is slowing the 3600x down and making the cpu bottleneck more likely at 1080p. I'd recommend some good 3600mhz c16 memory to go with that cpu for max performance.

 

[Goiabaman]

Your GPU temps are normal.

What benchmark results are you looking at ? Did you mirror their settings exactly ? I ask because if you have settings like anti-aliasing maxed and they didn't...or if they ran DLSS and you did not...results will vary wildly depending on game.

Also since you're playing at 1080p your cpu can really become a deciding factor in how high your FPS will be. Most online reviews run very high end CPUs in their reviews. Your 3600x is a solid gaming cpu but it won't match a 5900x or i9-10900k in gaming.

Additionally your ram choice of 2400mhz is slowing the 3600x down and making the cpu bottleneck more likely at 1080p. I'd recommend some good 3600mhz c16 memory to go with that cpu for max performance.

Yes, I made sure to make the graphic settings the same as the benchmark. So, I researched the benchmark with my settings, the only difference would be the models (ASUS, GIGABYTE, etc.) of the 3080, the same processor, the only difference would be the memories, they used much higher frequencies, however, would they influence both FPS ? (divisions of 20 to 30fps). So, I don't have much interest in taking a lot of frames in FPS games, I bought this card with this processor to play it in the future (when I buy a monitor) in 4K, I try to keep in mind that in this resolution this processor has practically no difference from a top processor of line.

 

[dorsai]

memory plays a big role in the performance Ryzen cpus provide.

View: https://www.youtube.com/watch?v=azFYdiH9Hns

 

[The_Ginja_Ninja]

I'm watching this thread with interest.

CPU: Ryzen 9 3900X
GPU: Zota Amp Holo RTX 3080
Mobo: MSI Mag Tomahawk X570
RAM: 2x16gb (3600C18)

I only really play CoD Warzone and CoD Cold War. In warzone, I get roughly the same fps as you (maybe 5 - 10 more) and around 90 - 110 in Cold War multiplayer.

Also feel that these should be higher. My userbenchmark of GPU, CPU and memory are all in the higher percentiles. I am playing 2k at ultra settings (incl. ray tracing).

 

[RTX 2080]

memory plays a big role in the performance Ryzen cpus provide.

View: https://www.youtube.com/watch?v=azFYdiH9Hns

This 100%.

I can't tell you how many times on here I've had a Ryzen 2600/3600 with sub-3000 MHz RAM and resulting low fps. Higher frame rates are CPU-dependent, and when you hamstring your CPU with slow RAM (which Ryzen CPUs are very sensitive too) it kills your frame rates, and not just at 1080p either, it affects all resolutions.

 

[Goiabaman]

memory plays a big role in the performance Ryzen cpus provide.

View: https://www.youtube.com/watch?v=azFYdiH9Hns

So, I will try to buy memories of higher frequencies, but I saw some benchmarks like this

View: https://www.youtube.com/watch?v=DDxRtQV1pS8&t=7s&ab_channel=XeonHacker
, increased "only" about 10 FPS.

 

[RTX 2080]

So, I will try to buy memories of higher frequencies, but I saw some benchmarks like this

View: https://www.youtube.com/watch?v=DDxRtQV1pS8&t=7s&ab_channel=XeonHacker
, increased "only" about 10 FPS.

Well since it's "only" a 10 fps difference, I guess there's no point in upgrading your RAM.

sarcasm

 

[Goiabaman]

I think you didn't understand the quotes, because yes, 10 fps are relevant, however, as I already said above, there is a difference of 20 to 30 fps with the benchmarks in computers with identical hardware, except for the frequency of the ram

 

[RTX 2080]

I think you didn't understand the quotes, because yes, 10 fps are relevant, however, as I already said above, there is a difference of 20 to 30 fps with the benchmarks in computers with identical hardware, except for the frequency of the ram

Unless 2 different benchmarks are comparing the same game at the same resolution and graphical settings (even if they are using the same hardware) the results will be different. Some benchmarks will show larger improvements in fps than others due the game and the way the game is being run and what sequence of the game is being benchmarked. And even if absolutely everything but the RAM was the same, the benchmarks are likely not comparing the exact same hardware to what you have; your results will probably be different. The purpose of benchmarks for me isn't to measure the exact performance uplift I should expect for my PC so much as quantify whether a part of my build is likely slowing me down enough to warrant an upgrade.

 

[Goiabaman]

I understand what you mean, I am thinking of contacting the support of the store where I bought it for them to carry out tests to confirm that the part cannot be with any problem. However, there is the possibility that this model, specifically the RTX 3080 BLACK, has a lower performance than the others. However, I believe that the chances of this last chance are small, I believe that there are no such great differences between pieces of the same line.

 

[dorsai]

So, I will try to buy memories of higher frequencies, but I saw some benchmarks like this

View: https://www.youtube.com/watch?v=DDxRtQV1pS8&t=7s&ab_channel=XeonHacker
, increased "only" about 10 FPS.

He's running a 1660...not a fast GPU so it will perform much lower than a 3080 even with the same exact memory.