RTX 4070 vs RTX 3080

Which is better? the 4070, or the 3080? Let's find out.

Table of Contents Table of Contents

RTX 4070 vs RTX 3080: Which is better? The RTX 4070 is due to be released soon, and with that, we’re sure you’re wondering how exactly the RTX 4070 stacks up against the RTX 3080, one of the best, well-rounded GPUs of the previous generation. The RTX 4070 is certainly a powerful budget card, but will the generational jump be enough to one-up the RTX 3080?

As with anything, it’s your use case that matters most. It doesn’t matter how good a GPU is at a certain thing, if it’s not right for you, you shouldn’t get it. These GPUs are primarily gaming GPUs, but they can perform other tasks well, like encoding and workstation tasks.

Without further delay, here’s RTX 4070 vs RTX 3080.

---

RTX 4070 vs RTX 3080

Here we will compare the RTX 4070 Ti vs the RTX 3080, this is where we will gather everything we know about the RTX 4070 and stack them against the RTX 3080.

RTX 4070 specifications

Here are the RTX 4070 specifications as we know them so far.

• Process: 4nm TSMC
• CUDA cores: 5,888
• Base clock: 1,920 MHz
• Memory Size: 12 GB
• Boost clock: 2,475 MHz
• Memory clock: 21 Gbps
• Memory bandwidth: 504 GB/s
• TBP: 200W
• PCIe interface: Gen4 x 16
• MSRP $599

RTX 3080 specifications

Here we will outline all the relevant specifications for the RTX 3080.

• Process: 8nm TSMC
• CUDA cores: 8,704
• Base clock: 1,440 MHz
• Memory Size: 10 GB
• Boost clock: 1,710 MHz
• Memory clock: 19 Gbps
• Memory bandwidth: 760 GB/s
• TBP: 320W
• PCIe interface: Gen4 x 16
• MSRP $699

As you can see, the RTX 4070 is one heck of a GPU, but does the 3080 have more power? It would be easy to think so based on the specifications.

---

RTX 4070 vs RTX 3080: Things to consider

As is the case with any purchasing decision, the main determining factor is you and your specific use case. The abundance of advanced features and specifications on a GPU may be irrelevant if it does not meet your needs. Therefore, it is essential to carefully evaluate all relevant factors before deciding whether the RTX 4070 or RTX 3080 is the best fit for your specific use case.

GPU CUDA core count

The number of CUDA cores in a GPU is a crucial factor that can have a significant impact on its performance, especially when dealing with workloads that can be parallelized using CUDA, such as machine learning, scientific simulations, and video processing.

CUDA cores are tiny processing units that operate in parallel within a GPU, resulting in faster computation of data-intensive tasks. The more CUDA cores a GPU has, the more parallel computations it can perform simultaneously, possibly leading to better performance and quicker processing times.

However, it’s essential to consider that the actual performance gains from increasing the CUDA core count can be influenced by several factors, including the software implementation, the type and speed of GPU memory, and the efficiency of the cooling system. Furthermore, a higher CUDA core count typically comes at the cost of higher power consumption and heat generation, which can impose limits on how much the core count can be increased without risking stability or reliability. The RTX 3080 has many more CUDA cores than the RTX 4070, making it ideal for gaming.

GPU clock speed

The maximum rated boost GPU clock speed is a critical specification that can have a significant impact on the performance of a GPU. This clock speed indicates the highest frequency that the GPU can operate at under optimal conditions, usually achieved when the workload is fully utilizing all available processing resources and the GPU is functioning within its power and temperature limits.

A higher maximum rated boost GPU clock speed generally implies that the GPU can process more data and perform calculations more quickly, resulting in improved performance and higher frame rates in games, or faster processing times in other applications.

However, the actual performance gains from increasing the maximum rated boost GPU clock speed depend on several factors, such as the workload, software implementation quality, and cooling system efficiency. It is also essential to recognize that increasing the maximum rated boost GPU clock speed comes with a trade-off in terms of power consumption and heat generation, similar to CUDA core count. This implies that the GPU may need a more powerful and effective cooling system to avoid thermal throttling and ensure steady operation at higher clock speeds.

Furthermore, increasing the clock speed can increase the risk of hardware failure, particularly if the GPU is overclocked beyond the manufacturer’s specifications. The RTX 4070 has a higher clock speed, meaning its cores work faster than the 3080s.

GPU memory size

The size of a GPU’s memory, also known as VRAM (Video Random Access Memory), plays a crucial role in its ability to process large amounts of data. VRAM is utilized by the GPU to store data and instructions that are essential for its computations. A larger memory size can lead to faster processing times and improved performance, particularly when dealing with large datasets, such as high-resolution graphics rendering, video processing, and deep learning.

It’s worth noting that the impact of memory size on performance can vary based on the workload and the software used to leverage the GPU’s processing power. In some instances, increasing the memory size beyond a specific threshold may not result in significant performance gains as the GPU might not be able to utilize the additional memory fully due to other limiting factors such as processing power or memory bandwidth.

Larger memory sizes tend to be more expensive, both in terms of the GPU’s cost and the power consumption required to operate it. As a result, selecting a GPU with a larger memory size should be a carefully considered decision based on the specific requirements of the intended workload. The RTX 4070 has a higher memory size, making it ideal for larger textures and higher resolutions.

GPU TBP

The TBP, or Total Board Power, of a GPU, is a critical specification that affects its performance in multiple ways. Essentially, the TBP is the maximum power that the GPU can draw from the PSU and dissipate through its cooling system. If the TBP of a GPU is too low for the workload, it may not be able to maintain the required clock speeds, leading to slower processing times and lower frame rates. This is especially problematic for workloads with high levels of GPU utilization, such as gaming and machine learning.

On the other hand, a high TBP can lead to increased power consumption and heat generation, potentially impacting system stability and increasing the risk of thermal throttling and hardware failure. These issues can be addressed by using an efficient cooling system and ensuring that the PSU can provide the necessary power without overheating or causing voltage drops.

Moreover, a higher TBP typically indicates lower power efficiency, meaning that a GPU with a higher TBP may require more power to achieve the same level of performance as a GPU with a lower TBP. Therefore, selecting a GPU with an appropriate TBP is crucial to achieving optimal performance and efficiency in various workloads. The RTX 4070 has a much smaller TBP, making it more efficient to run over the RTX 3080.

---

RTX 4070 vs 3080 performance

Looking at TechSpot’s review of the 4070 we get a good performance comparison of what they have to offer. There we get a good look at the performance differences between the generations. And overall we see a slight increase in framerate on average over the previous popular card.

In particular, at 1080p the RTX 4070 averages 175 FPS with a 1% low of 139, whilst the 3080 10GB gets 161 with a low of 129. Then up to 1440p, the difference decreases, with the 4070 averaging 126 FPS with a 1% low of 104, with the RTX 3080 averaging 121 FPS with a 97 low. With the 4k flipping the charge, the 4070 averaged 69 FPS with a 1% of 57, whilst the 3080 gets 71 FPS with a low of 58.

So in fact for the highest resolution, the older card still performs better. This is likely from the better bandwidth of the 3080, which means it can handle the higher resolutions much better. But for the lower resolutions, the 4070 still has the better value for those.

---

RTX 4070 vs RTX 3080: Final Word

In conclusion, both the RTX 4070 and the RTX 3080 are powerful GPUs that offer impressive performance for a wide range of workloads. While the RTX 3080 offers improved performance in some areas, such as memory bandwidth and CUDA core count, the RTX 4070 remains a strong contender with its larger memory size and more efficient power consumption.

Ultimately, the decision between these two GPUs will depend on your specific needs and use case. If you require the latest and greatest hardware for workloads such as gaming or machine learning, the RTX 4070 may be the better choice. However, if you value efficient power consumption and a larger capacity for processing, the RTX 3080 may be the more practical option.

Regardless of which GPU you choose, it’s important to consider factors such as TBP, memory size, CUDA core count, and clock speed to ensure that your system is optimized for your specific workload. With either GPU, you can expect top-tier performance and an excellent user experience for years to come.

RTX 4070 vs RTX 3080 FAQs

Is RTX 4070 faster than 3080?

The RTX 4070 is quite close in the benchmarks to the RTX 3080 and RX 6800XT. While it is much faster than the RTX 3070, the difference is less noticeable between the RTX 4070 and 3080.

How much faster is the 4070 than the 3080?

The RTX 4070 has a base clock of 1,920 MHz and a boost clock of 2,475 MHz. Whilst the RTX 3080 has a base of 1,440 MHz and a boost of 1,710 MHz. This means the 4070 is 33% and 44% faster than the 4080.