Nvidia RTX 4090 vs. 3090 - Which Should You Buy in 2023?
NVIDIA’s RTX 4090 series sets a new standard for graphics cards and takes the moniker of the company’s flagship GPU from the wildly popular RTX 3090 series.
Launched on October 12th 2022, the RTX 4090 range is the first generation of NVIDIA’s RTX 40 series GPU range which is set to replace all models in the RTX 30 series that have been on the market for more than two years.
So, how does the RTX 4090 stack up against the RTX 3090 range, and is it a worthy successor? Let’s take a look at the specs from performance and power to price and value proposition to determine whether the RTX 4090 is a worthy upgrade for the RTX 3090.
RTX 4090 vs. 3090 - Side by Side Spec Comparison
Before we delve into all of the spec differences between the 3090 and 4090 graphics cards, here is a quick overview of the key differences between the technical specifications.
Technical Specs | RTX 4090 | RTX 3090 |
GPU | AD102 | GA102 |
Architecture | Ada Lovelace | Ampere |
Pipelines / CUDA cores | 16384 CUDA cores | 10496 CUDA cores |
Thermal design power (TDP) | 450 Watt | 350 Watt |
Manufacturing process technology | 4 nm | 8 nm |
Boost clock speed | 2520 MHz | 1695 MHz |
Core clock speed | 2235 MHz | 1395 Mhz |
Memory Capacity | 24GB GDDR6X | 24GB GDDR6X |
Memory speed | 21 Gbps | 19.5 Gbps |
Bandwidth | 1018 GBps | 936 GBps |
Tensor cores | 576 | 328 |
GeForce RTX 4090 Specifications |
Release Date: Sep 20th, 2022 |
GPU Name: AD102 |
Architecture: Ada Lovelace |
Process Size: 4 nm |
Transistors: 76,300 million |
Die Size: 608 mm² |
Number of CUDA Cores: 16384 |
Base Clock: 2235 MHz |
Boost Clock: 2520 MHz |
Memory Capacity: 24GB GDDR6X |
Memory Clock: 1313 MHz (21 Gbps effective) |
Shading Units: 16384 |
TMUs: 512 |
ROPs: 176 |
SM Count: 128 |
Tensor Cores: 512 |
RT Cores: 128 |
L1 Cache: 128 KB (per SM) |
L2 Cache: 72 MB |
Features |
DirectX: 12 Ultimate (12_2) |
OpenGL: 4.6 |
OpenCL: 3.0 |
Vulkan: 1.3 |
CUDA: 8.9 |
Shader Model: 6.6 |
Design |
Slot Width: Triple-slot |
TDP: 450 W |
Suggested PSU: 850 W |
Outputs: 1x HDMI 2.1 and 3x DisplayPort 1.4a |
Power Connectors: 1x 16-pin |
Board Number: PG139 SKU 33 |
GeForce RTX 3090 Specifications |
GeForce RTX 3090 Specifications |
Release Date: Sep 1st, 2020 |
GPU Name: GA102 |
Architecture: Ampere |
Process Size: 8 nm |
Transistors: 28,300 million |
Die Size: 628 mm² |
Number of CUDA Cores: 10496 |
Base Clock: 1395 MHz |
Boost Clock: 1695 MHz |
Memory Capacity: 24GB GDDR6X |
Memory Clock: 1219 MHz (19.5 Gbps effective) |
Shading Units: 10496 |
TMUs: 328 |
ROPs: 112 |
SM Count: 82 |
Tensor Cores: 328 |
RT Cores: 82 |
L1 Cache: 128 KB (per SM) |
L2 Cache: 6 MB |
Features |
DirectX: 12 Ultimate (12_2) |
OpenGL: 4.6 |
OpenCL: 3.0 |
Vulkan: 1.3 |
CUDA: 8.9 |
Shader Model: 6.6 |
Design |
Slot Width: Triple-slot |
TDP: 350 W |
Suggested PSU: 750 W |
Outputs: 1x HDMI 2.1 and 3x DisplayPort 1.4a |
Power Connectors: 1x 12-pin |
Board Number: PG132 SKU 10 |
RTX 4090 vs 3090 Performance Comparison
RTX 4090 Performance Specs
- No NVLink or SLI support
- Dominates previous GPUs with its performance
- Can become limited by high-end CPUs even at 4K resolution
- Requires a highly capable modern CPU, such as AMD Ryzen 7950X or Intel Core i9-13900K
- DLSS 3.0 support with more powerful hardware and optical flow accelerators
- Offers impressive gains in demanding games such as Flight Simulator 2020 with DLSS 3.0
RTX 3090 Performance Specs
- Suitable for 1440p and below resolutions
- Performance gains under 4K on the RTX 4090 aren't significant
- Not powerful enough to convincingly pull off DLSS 3.0
- Great choice for gaming at 4K on a 60Hz display
Performance Comparison Summary
Side by side, the RTX 4090 is more powerful than the RTX 3090 and RTX 3090 Ti, with extremely impressive 4K performance that can sometimes be bottlenecked by CPU capability. The lack of NVLink and SLI support on the RTX 4090 is a slight disappointment, however, the advanced optical flow accelerators and DLSS 3.0 technology go some way to making up for the absence of NVLink and SLI support.
So what about the RTX 3090 – where does it stand with the advent of the RTX 4090 series?
Well, there is no getting away from it – the RTX 3090 series still offers brilliant performance for 1440p and 4K gaming, with less dependence on the CPU. If money is no object and you’re looking for outright performance with no sacrifices, then the RTX 4090 is the better choice for demanding games like Cyberpunk 2077 that demand higher resolutions. On the other side of the coin, the RTX 3090 GPUs still offer outstanding 4K gaming performance that will be more than enough for most.
4090 vs RTX 3090 Architecture Comparison
As gamers, we’re always searching for a smoother, more efficient gaming experience. Well, the RTX 4090 will give you that. Using the new Ada Lovelace microarchitecture, the RTX 4090 is a big leap from the Ampere (NVIDIA’s 2nd gen RTX architecture), used in the TRX 3090.
The arrival of the NVIDIA GeForce RTX 4090 is quite literally a game changer, and with this upgrade to the latest Ada Lovelace architecture, means more transistors, faster clock speeds and improved Tensor cores for AI computation.
This architectural enhancement is a huge win for gamers and here’s some reasons why:
- Ada Lovelace microarchitecture uses 4 nm fabrication process from TSMC’s N4 technology, compared to the 8nm RF process used in RTX 30-series GPUs
- The smaller processor size allows for higher clock speeds and 1.7x more transistors (76 billion) than the RTX 3090 (28 billion)
- The greater number of Tensor cores, along with improved technology, has allowed for the introduction of DLSS 3.0, which brings up to 4 times better performance than brute-force rendering
- The Ada Lovelace architecture introduces Shader Execution Reordering (SER)
- The RTX 4090 runs cooler (60°C) compared to the RTX 3090 (66°C)
At BPC Tech, we love the data. Seeing figures like this gets us excited, and it’s impressive to see the architectural improvements NVIDIA have made.
4090 vs RTX 3090 CUDA Core Comparison
When comparing the RTX 3090 with the 4090, the CUDA Core count is a crucial element to consider. With a significant increase in the number of CUDA Cores in the new 4090 model, it is intriguing to investigate what this means in terms of performance.
When you take a closer look, you’ll notice that there is a clear difference in the core count, clock speeds, and a variety of vital specs between these two GPUs. Here are some of the key enhancements when it comes to the RTX 4090 CUDA Core:
Number of CUDA Cores - RTX 4090 has 16,384 CUDA cores while the RTX 3090 has 10,496 CUDA cores, representing a 64% increase.
Better Specs - RTX 4090 has better specs in all aspects, including higher base and boost clock, slightly faster memory, and more RT and Tensor cores.
Memory Speed and Bandwidth – Both the RTX 4090 and the 3090 have memory capacity of 24GB GDDR6X, so there is no increase in this regard. However, the memory speed takes a jump from 19.5GT/s on the 3090, to 21 GT/s on the 4090, while the memory bandwidth has reached 1,008 GB/s, rather than 936.2GB/s. As you can see, the 3090 is still a brilliant GPU. But who wouldn’t like greater bandwidth and faster speed?
Architecture - The new architecture and 5nm process allows for higher core count and clock speeds.
4090 vs RTX 3090 Power Use Compared
Depending on what you’re looking for, or what your priorities are, there is a couple of ways you could look at the power usage of the RTX 4090, compared to the 3090.
The updated 4090 model has a TDP of 450 Watts, which is 100W more than its predecessor. This requires an 850W power supply for the RTX 4090, rather than 750W for the 3090. So, this is something to keep in mind when purchasing the RTX 4090 – you will need more power. However, this isn’t done by mistake. The extra 100W of power generates more heat and more importantly, results in greater performance.
When it comes to the power usage of your GPU and what it means for your gaming, here is the main differences between the RTX 4090 and 3090:
- RTX 3090 has a TDP of 350W, while the RTX 4090 has a TDP of 450W
- You need at least an 850W power supply for the RTX 4090, whereas the 3090 needs a 750W power supply
- RTX 4090 consumes 100W more power than the RTX 3090 and generates more heat
- Taking a game like Horizon Zero Dawn as an example, the RTX 4090 is consuming 427.7W when running, compared to 361.7W for the RTX 3090
As you can see from the above, it depends on your goals and concerns regarding the power usage. If you are seeking better performance, the 4090 will give you that. However, it comes at a cost - you will be using more power to achieve it.
RTX 4090 vs 3090 Size Comparison
Despite its smaller size, the updated RTX 4090 packs a powerful punch. Even with a die size reduction from 628 mm² to 608 mm², the 4090 is still a goliath compared to the RTX 3090. Let’s take a closer look at size difference between the RTX 3090 and the new 4090 model.
- The RTX 4090 has a smaller die size of 608mm², compared to 628 mm² on the 3090 model
- Despite the smaller size, the 4090 has 76.3 billion transistors, versus the RTX 3090's 28.3 billion transistors
- The RTX 4090 is built on the smaller Ada/Lovelace architecture with the AD102-300 chip, whereas the 3090 uses the larger Ampere architecture
- The RTX 4090 is a more compact build with its smaller fabrication process, leading to improved performance
Don’t be fooled by the compact build of the RTX 4090, this GPU delivers in terms of the innovation, technology, and performance.
Is it Worth Upgrading from the RTX3090 to the RTX4090?
Let’s get this straight, the NVIDIA RTX 4090 and the RTX 3090 are both exceptional Graphic Processing Units. When it comes to making your decision on whether to upgrade your RTX 3090, or to purchase the 4090 as your first ever GPU, it all comes down to how much you are willing to spend, and how much you would like to invest in your gaming.
If you have the money to spend, and you want to optimise your gaming set up, it’s a no brainer – the RTX 4090 performs better than the 3090 in every single category. Looking at the specs above, and comparing the two GPUs against one another, we can see that NVIDIA have given us an outstanding new product that out-preforms their previous model.
But don’t be mistaken, the RTX 3090 is still a fantastic GPU. If you find yourself on a budget and you are unwilling to pay the extra money, you will not be disappointed by the 3090.




















