Reviews for NVIDIA’s GeForce RTX 4090 FE graphics card are out, and so far the results are impressive. Reviewers who are usually critical of NVIDIA’s graphics card seem to be impressed with the performance of the RTX 4090. Not only that, these new RTX 4090 graphics cards are physically big, but performance gains are (significantly) huge as well. Unfortunately, size and performance are not the only “big” things about it. Its price is also on the expensive side, starting at $,1599. Some custom cards from AIB partners are selling for around $2,000. Anyway, here’s an overview comparison between the new RTX 4090 vs the RTX 3090 Ti and RTX 3090.
GeForce RTX 4090 vs RTX 3090 Ti and RTX 3090 – Almost Double the Performance
When the RTX 4090 was first announced by NVIDIA, the company claimed that it is up to 4x faster than the RTX 3090 (Ti). Of course, many of us were very skeptical about the claim. However, there’s a big asterisk there. Those up to 4x faster are when DLSS 3.0 is enabled. This is one of the key main features of the RTX 40 series GPUs based on the new Ada Lovelace architecture.
DLSS 3.0 offers a significant improvement and a somewhat different approach compared to DLSS 2.0. However, DLSS 3.0 is only available on the RTX 40 series GPUs. The older RTX 30 series and RTX 20 series will not support DLSS 3.0 (at least maybe for now).
However, when reviews for the RTX 4090 came out recently, the new flagship GPU is still significantly faster than the RTX 3090 Ti and RTX 3090 by a huge margin. It’s almost double in some game titles and scenarios.
Another key feature of the RTX 40 series GPUs is Dual NVIDIA Encoders (NVENC). This feature is very useful for streamers or gamers who record their gameplay. The RTX 40 series GPUs can decode and encode AV1 format, and the same can be said with H.264 and H.265. Not only that, it supports the AV1 format, but it also cuts export times by up to half. The NVENC AV1 encode is being adopted by OBS, Discord, Blackmagic Design DaVinci Resolve, and many more.
GeForce RTX 40 Series GPUs New Features
- Streaming multiprocessors with up to 83 teraflops of shader power — 2x over the previous generation.
- Third-generation RT Cores with up to 191 effective ray-tracing teraflops — 2.8x over the previous generation.
- Fourth-generation Tensor Cores with up to 1.32 Tensor petaflops — 5x over the previous generation using FP8 acceleration.
- Shader Execution Reordering (SER) improves execution efficiency by rescheduling shading workloads on the fly to better utilize the GPU’s resources. As significant innovation as out-of-order execution was for CPUs, SER improves ray-tracing performance by up to 3x and in-game frame rates by up to 25%.
- Ada Optical Flow Accelerator with 2x faster performance allows DLSS 3 to predict movement in a scene, enabling the neural network to boost frame rates while maintaining image quality.
- Architectural improvements tightly coupled with custom TSMC 4N process technology results in an up to 2x leap in power efficiency.
There are a couple of things that we expected from the RTX 40 series GPUs, but sadly these features were absent. The RTX 4090 didn’t have support for DisplayPort 2.0, which is unfortunate since it can easily handle 4K resolution with well above 100 frames per second. It’s also using PCIe 4.0, although the latest motherboards and platforms have support for PCIe 5.0.
Below is a side-by-side specifications comparison of the RTX 4090, RTX 3090 Ti, and RTX 3090.
GeForce RTX 4090 vs RTX 3090 Ti and RTX 3090 Specifications
| RTX 4090 | RTX 3090 Ti | RTX 3090 | |
|---|---|---|---|
| Architecture | Ada Lovelace | Ampere | Ampere |
| GPU | AD102 | GA102 | GA102 |
| Fab. Process | TSMC 4N | Samsung 8N | Samsung 8N |
| Tansistors (Billions) | 76.3 | 28.3 | 28.3 |
| CUDA Cores | 16384 | 10752 | 10496 |
| Ray Tracing Cores | 128 3rd gen | 84 2nd gen | 82 2nd gen |
| Tensor Cores | 512 4th gen | 336 3rd gen | 328 3rd gen |
| ROPs | 176 | 112 | 112 |
| TMUs | 512 | 336 | 328 |
| Boost Clock | 2.52 GHz | 1.86 GHz | 1.7 GHz |
| Base Clock | 2.23 GHz | 1.67 GHz | 1.4 GHz |
| Memory Size | 24 GB GDDR6X | 24 GB GDDR6X | 24 GB GDDR6X |
| Memory Interface | 384-bit | 384-bit | 384-bit |
| Memory Speed | 21 Gbps | 21 Gbps | 19.5 Gbps |
| NVIDIA DLSS ver. | 3 | 2 | 2 |
| Graphics Card Power | 450W | 450W | 350W |
| Required PSU | 850W | 850W | 750W |
| Release Date | October 2022 | March 2022 | September 2020 |
| Launch Price | $1,599 | $1,999 | $1,499 |
Check GeForce RTX 4090 graphics cards latest pricing and availability on Amazon.com here.
GeForce RTX 4090 vs RTX 3090 (Ti) Benchmarks from Various Reviews
Different media outlet and independent reviewers have already published their respective reviews of NVIDIA’s GeForce RTX 4090 Founders Edition graphics card. If you are looking for the custom RTX 4090 graphics card reviews, they will go live tomorrow. So far, most of the reviews I have watched and read align with each other. Even some of the most critical reviewers were impressed with the RTX 4090’s performance.
GeForce RTX 4090 Benchmarks by Linus Tech Tips
Based on LTT’s testing, the RTX 4090 is significantly faster, on Cyberpunk 2077, than the RTX 3090 Ti and RTX 3090 in traditional rasterization alone. The same can be said in Shadow of the Tomb Raider with Ray Tracing enabled, and on Forza Horizon 5. Although the gains are not that significant anymore for the latter two game titles.
UPDATE: LTT’s settings on their Cyberpunk 2077 benchmark were all over the place. That’s why the RTX 4090 got way more frames than it should be. They posted the correct numbers:
- *No RT, no DLSS (1%low, 5%low, avg):
- RTX 4090: 54, 69, 81
- RTX 3090 Ti: 43, 46, 56
- RTX 3090: 35, 43, 50
- RX 6950 XT: 30, 39, 46
Source: Linus Tech Tips
GeForce RTX 4090 Game Benchmarks by TechPowerUp
Meanwhile, based on TechPowerUp’s average results across three different resolutions, the RTX 4090 performs better in 4k resolution or 3840×2160. The performance improvement diminishes as the resolution lowers.
In terms of power consumption, the RTX 4090 is actually more power efficient than the older RTX 3090 Ti and RTX 3090.
Source: TechPowerUp
GeForce RTX 4090 Review Benchmarks by TechSpot
TechSpot/HardwareUnboxed also shared a 13-game average of the RTX 4090 across three different resolutions. And the results that they got are in line with TechPowerUp’s results. The RTX 4090 is no doubt the fastest gaming graphics card to date.
Source: TechSpot
Finally, do not buy an RTX 4090 if you only have a 1080p display. I wouldn’t even recommend it for a 2560×1440 monitor. Maybe yes for a 3440×1440 or 5120×1440 monitor. But if you are gaming at 4K resolution or planning to get a 4K monitor, the RTX 4090 will not disappoint. You can check out some of the best 4K high-refresh rate gaming monitors here.
Check GeForce RTX 4090 graphics cards latest pricing and availability on Amazon.com here.
