Comparing Nvidia GeForce RTX 4090 and RTX 4080: Which is Better?

Comparing Nvidia GeForce RTX 4090 and RTX 4080: Which is Better?

Nvidia’s RTX 4090 and RTX 4080 are incredibly powerful graphics cards. Built on the latest Ada Lovelace architecture, they feature increased CUDA cores, higher clock speeds, and enhanced memory capabilities. These cards require significant power and cooling solutions, but their exceptional performance justifies these requirements.

For 4K gaming, these GPUs are among the most robust available and are expected to maintain their top performance status despite new competition from AMD's RX 7000 series.

If you're seeking top-tier gaming performance shortly, Nvidia’s RTX 4090 and RTX 4080 are excellent choices. But which one should you choose, and how do they perform in practical scenarios?

Key Features and Performance Analysis

The Nvidia RTX 4000 series introduces a new generation of GPUs based on the Ada Lovelace architecture, utilizing TSMC's 4N (5nm) process node. Only two GPUs from this series have been officially released, with one announced and subsequently canceled. Nevertheless, the initial performance results are promising for the RTX 40-series.

The Ada Lovelace architecture significantly boosts CUDA core counts compared to the RTX 30-series and supports higher clock speeds. It also features advanced RT cores for ray tracing and Tensor cores for deep learning super sampling (DLSS). Enhancements to the ray tracing pipeline and the inclusion of DLSS 3 with its innovative frame generation technology are notable upgrades in this generation.

This series also removes NVLink support, effectively ending SLI gaming, which had already received limited support in recent GPU generations.

Despite concerns about high power consumption, Nvidia has adjusted its TDP estimates for this generation. The TDP represents a power limit rather than a peak performance target, suggesting that these GPUs will not exceed their TDP in typical gaming scenarios.

New Power Standards and Considerations

These GPUs utilize the new 16-pin power connector, or 12VHPWR connector, which can deliver more power than traditional six—and eight-pin cables. While some new power supplies support this connector, adapters are provided for older PSUs to connect multiple eight-pin PCIe power cables. Although there have been isolated incidents of melting adapters, these are usually due to user error. Ensuring the adapter is securely plugged in without any wiring kinks is crucial.

Here's a comparison chart between the Nvidia RTX 4090 and RTX 4080

A specifications table is an excellent starting point for comparing two cards from the same generation. While it may not be as useful for comparing cards across different generations due to varying designs, it can provide valuable insights into the potential of cards that share the same process and similar GPU architecture.

FeatureNvidia RTX 4090Nvidia RTX 4080
GPU ArchitectureAda LovelaceAda Lovelace
CUDA Cores16,3849,728
Boost Clock Speed2.52 GHz2.51 GHz
Base Clock Speed2.23 GHz2.21 GHz
Memory24 GB GDDR6X16 GB GDDR6X
Memory Speed22 Gbps21 Gbps
Memory Bandwidth1,008 GB/s672 GB/s
TDP (Thermal Design Power)450W320W
Transistors76.3 billion45.9 billion
Ray Tracing Cores144112
Tensor Cores576448
Output PortsHDMI 2.1, DisplayPort 1.4aHDMI 2.1, DisplayPort 1.4a
PCI ExpressPCIe 4.0PCIe 4.0
Release DateQ4 2022Q4 2022
Price (MSRP)$1,499$1,199

Detailed Comparison: Nvidia RTX 4090 vs RTX 4080

Spec Analysis

The differences between the RTX 4090 and RTX 4080 are immediately apparent.

The RTX 4090 boasts more CUDA cores than the 4080, promising a significant performance edge in most games, particularly since their clock speeds are similar, with the 4090 even slightly faster. This marks a shift from previous generations, where higher-tier cards didn't always have higher clock speeds.

The disparity extends to RT and Tensor cores. Both cards use the latest third-generation RT and fourth-generation Tensor cores, but the 4090's superior performance suggests it will excel in ray-traced games, far surpassing the 4080.

Memory is another key differentiator, with the 4090 featuring 8GB more than the 4080. Both use GDDR6X memory, exclusive to Nvidia through a partnership with Micron. Interestingly, the 4080's memory speed is slightly faster at 22.4 Gbps than the 4090's 21.2 Gbps. However, the 4090's wider memory bus results in significantly higher overall bandwidth, demanding more power and pushing its limits, according to most reviews. Conversely, the 4080's typical power consumption around 300W makes it more efficient, although it can't match the 4090's performance.

Real-World Performance of the RTX 4090 and 4080

The RTX 4090 and 4080 deliver exceptional power, outperforming previous generations at all resolutions. However, how do they fare against each other? As expected, the RTX 4090 is the more powerful card, achieving higher frame rates across all resolutions and detail settings, especially with ray tracing enabled.

In synthetic benchmarks like 3DMark Time Spy, the RTX 4090 scores over 31,000 points at stock, setting new records. The RTX 4080, paired with a decent processor, scores around 25,000, significantly improving over previous generations but still far behind the 4090.

In 3DMark Port Royal, which focuses on ray tracing, the RTX 4080 reaches nearly 18,000 points, a 50% increase over the RTX 3090 and 3080 Ti. However, the RTX 4090 outperforms with scores nearing 26,000 points.

Gaming Benchmarks

In actual gaming scenarios, the RTX 4090 continues to lead. In Cyberpunk 2077 at 4K Ultra settings, the RTX 4080 nearly hits a 60 FPS average, a 10-20% improvement over the previous generation. The RTX 4090, however, achieves close to 80 FPS, solidifying its superior performance.

With ray tracing maxed out (without DLSS), the RTX 4090 is the only card averaging over 30 FPS, reaching around 40 FPS, whereas the best last-generation cards barely manage 20 FPS.

Performance in AMD-Optimized Games

In games optimized for AMD, such as Assassin’s Creed Valhalla and Forza Horizon 5, the RTX 4080 shows a 35% performance increase over its predecessors. However, the RTX 4090 still outperforms, achieving over 120 FPS at 4K Ultra settings. However, AMD’s RX 6950 XT and 6900 XT are within 10% of the RTX 4080 in these titles, indicating game-specific performance considerations.

At 1440p, the RTX 4090 is about 20% faster than the 4080, with the former often exceeding 150 FPS, making it overkill for all but the most demanding esports scenarios. The RTX 4090 is primarily a 4K GPU, where its power truly shines.

DLSS 3 Frame Generation: Is It Worth It?

DLSS 3, Nvidia's latest upscaling technology, introduces frame generation, creating new frames based on game data and previous frames, significantly boosting frame rates with minimal impact on frame times. This can dramatically increase performance, such as elevating the RTX 4080 from below 30 FPS to over 120 FPS in Cyberpunk 2077 with ray tracing, and enhancing the RTX 4090’s performance in Plague Tale: Requiem by over 100%.

Image comparing the FPS difference in Cyberpunk 2077 between using DLSS and not on a 4090 GPU

However, DLSS 3’s image quality isn't perfect, with generated frames sometimes displaying visual artifacts, particularly in games with heavy HUD elements. While it’s an exciting feature for future-proofing and lower-end cards, it may not be immediately essential for high-end GPUs like the RTX 4090 and 4080.

Conclusion

The Nvidia RTX 4090 and 4080 are currently the most powerful graphics cards available. They offer substantial performance improvements over previous generations and are arguably the best choice for playing demanding games.

However, choosing between the two depends on your needs. The RTX 4090 offers the best performance value, but its extreme power is often unnecessary for most games. If budget is not an issue and you want the best, the RTX 4090 is an easy choice. The RTX 4080 and AMD's upcoming cards are also strong contenders for everyone else.

As the demand for GPU resources continues to surge, especially for AI and machine learning applications, ensuring the security and ease of access to these resources has become paramount.

Spheron’s decentralized architecture aims to democratize access to the world’s untapped GPU resources and strongly emphasizes security and user convenience. Let’s unpack how Spheron protects your GPU resources and data and ensures that the future of decentralized compute is both efficient and secure.

Interested in learning more about Spheron’s network capabilities and user benefits?Review the whitepaper in full.