Conclusion
In today’s review, we took the new NVIDIA GeForce RTX 4080 Founders Edition video card and overclocked it as high as we could. We tested 4K performance with and without DLSS and Ray Tracing to see where the strengths lie in its overclocking ability. You can read our full launch review of the NVIDIA GeForce RTX 4080 Founders Edition as well. The GeForce RTX 4080 FE has an MSRP of $1,199.99 and was launched in November of 2022.
The GeForce RTX 4080 FE sits right underneath the GeForce RTX 4090 but does have a pretty big gap between them. We found that performance-wise, the GeForce RTX 4080 competes with AMD’s Radeon RX 7900 XTX. In raster performance, it is competitive, but in Ray Tracing performance the GeForce RTX 4080 FE dominates the scene. It also has very good upscaling support and performance with DLSS and DLSS 3.
Overclocking Summary
The default GPU Boost on the NVIDIA GeForce RTX 4080 Founders Edition is 2505MHz. We found that the actual real-world gaming frequency, thanks to NVIDIA GPU Boost dynamically clocks the Founders Edition up to 2775MHz while gaming, providing a big 11% higher GPU clock frequency, just at stock settings thanks to this enormous cooling on the video card and power efficiency. The default, stock board power hit 308W, so there is plenty of headroom for overclocking with the three 8-pin PCIe connectors utilized. Therefore we fired up EVGA Precision X1 software to overclock and set the fans to 100% for good measure.
With EVGA Precision X1 software we were able to increase the Power Target by 11%, from 100% to 111%. This is much lower than the RTX 4090 FE, which was able to be increased by 33%. Once again, just like the GeForce RTX 4090 FE Overclocking, we were able to actually hit 3GHz on this GPU with the GeForce RTX 4080 FE. However, we could not sustain this frequency over time. In the end, it turns out +170 on the GPU Clock was the best setting in our overclocking attempts. This allowed a stable and consistent clock frequency at 2955MHz.
Interestingly enough, this is also the exact same clock speed we found stable on the GeForce RTX 4090 FE, and the ASUS ROG STRIX RTX 4080 video card. It seems 2955MHz is a hard physical limit of this GPU. While it can peak a bit higher, depending on the card, it cannot sustain that frequency without either more Voltage or power headroom. We are finding, in our testing 2955MHz is a common overclocking “red-line” for the RTX 4080 and RTX 4090 GPUs.
We were able to increase the memory frequency from 22.3GHz up to 24GHz. The default memory bandwidth is 716.8GB/s, and with our overclock at 24GHz for the memory it is increased to 768GB/s, a 7% bandwidth improvement. We were quite impressed by the level of memory overclock, and the RTX 4080 FE benefits from that.
All of this increased the total board power up to 324W in our testing, while gaming. That is only a 5% increase in power, from overclocking, which is quite good. The power draw at default is quite efficient, and as you raise the power the performance goes up on a 1:1 ratio with it, so it isn’t out of line.
The other good part is the cooling, this cooling on the RTX 4080 Founders Edition is enormous, but it does an excellent job keeping the GPU cool while overclocked. It is the same cooler as the RTX 4090 FE. You don’t have to ramp the fans up to 100% to achieve this overclock, which is good because the fans are super loud at 100%. This cooler does not hold back overclocking, just like the RTX 4090 FE.
Final Points
Overall, overclocking the GeForce RTX 4080 Founders Edition at 4K provided around a 4-6% performance improvement depending on the game. The average was definitely around 5%, with 4% being on the low-end gains. The game is what really matters here, in the more graphically demanding games, the percentage will be higher, around 6%, but the less demanding games, more CPU dependent, will be a bit under that.
Overall, these are very low gains in terms of performance uplift from manual overclocking of the NVIDIA GeForce RTX 4080 Founders Edition. None of these performance increases actually helped create a new gameplay experience that wasn’t playable before. It barely helped with increasing gameplay to a playable level in situations that weren’t playable.
For the added potential of instability issues over time, power draw, and effort, it does not seem worth it to overclock the GeForce RTX 4080 Founders Edition. NVIDIA GPU Boost is already taking the GPU very high by default, and pushing it that extra 5% doesn’t get you much in performance. We are also finding that around 2955MHz seems to be a common denominator in terms of maximum GPU clock frequency potential on GeForce RTX 4080 GPUs, and that is also extended to GeForce RTX 4090.
Overclocking the NVIDIA GeForce RTX 4080 Founders Edition is easy, but the effective result doesn’t seem worth it. What does seem worth it, is DLSS upscaling technology. DLSS is what makes settings that are not playable at 4K, actually playable at 4K on the GeForce RTX 4080 FE. It makes Ray Tracing playable at 4K in many graphically demanding games and is in fact required in order to achieve that. DLSS adds more to the performance profile than overclocking does.
Discussion (9 replies)
Join Discussion →Nice read, thank you. :)
I can't remember where I read it but I do remember someone saying sometime before the 40 series launch that NVIDIA's dynamic boost had reached a point where OC gains would be at a minimum for this gen. It was something to the effect that there's not even much for board partners to squeeze out with custom models. This card is definitely an example of that. Granted, it's been getting that way for a while now but this might be the least we've seen in around 10 years. Even with my 4090 I usually keep it just under 2900 and that's only ~100 more than the out-of-the-box settings. Still a fun ride though. I'd say the other big take away is the power efficiency of these GPUs in comparison to Ampere. It's almost like night and day.
CPUs and GPUs are coming factory maxed now; there's still value in tuning, and still some performance on the table, but it comes with significant exploratory testing.
Which is why we give them to Brent to test :D
I remember the time when my Geforce3 TI200 overclocked and performed better than my friends Geforce 3 Ti500.
IMO OC has been dead for years, you don't get much more performance since GPUs are already maxxed out to begin with.
Its really not worth the higher temps and powerdraw.
And that applies to CPUs too.
Man I remember those candy bar celerons... overclocked like a BEAST.
I had the legendary celeron 300 that could easily OC to 450mhz and a duron 600 that reached 1.1ghz.
...but did you have a Golden Orb on it?
:D
The golden orb....
Overclocking is dead unless you are using exotic cooling methods. DI or LN2. Even on high end custom water the gains are minimal. OEM's aren't leaving anything on the table anymore. And people aren't so quick to apply a custom firmware to remove the power, voltage and temp limits on the $1000+ GPU's.