Nvidia‘s GeForce RTX 3000 series GPUs based on the Ampere architecture represent a major generational leap in performance and capabilities. The flagship RTX 3090 and slightly cut-down RTX 3080 offer enthusiasts, gamers and creators immense graphics horsepower. But how do these two graphics cards stack up against each other? This comprehensive comparison will examine all the key specs, real-world gaming benchmarks, power draw and thermal performance to help you decide which of these GPU giants is the better choice.
A Brief History of Nvidia GPUs Leading up to Ampere
Before diving into the Ampere cards, it‘s worth providing some context on Nvidia‘s GPU roadmap over the past decade. Each new architecture brought major performance and efficiency improvements.
- Fermi (GTX 400, 2010): Fermi introduced key compute capabilities for GPU-accelerated workflows. However, it ran extremely hot and power hungry.
- Kepler (GTX 600/700, 2012-2014): Kepler focused on power efficiency and GPU boost clock speeds. It provided strong gaming performance and became popular in data centers.
- Maxwell (GTX 900, 2014-2015): Maxwell further improved efficiency with benefits for thermals and noise. The GTX 970 and 980 excelled for 1080p and 1440p gaming.
- Pascal (GTX 1000, 2016-2019): Pascal pushed into the high end with its flagship GTX 1080 Ti. It also debuted Nvidia‘s current GPU Boost technology for dynamic overclocking.
- Turing (RTX 2000, 2018-2020): Turing introduced dedicated ray tracing and AI cores. Despite the new capabilities, its pricing was controversial at launch.
Ampere aims to combine the strengths of Turing‘s features with dramatically improved rasterization and compute performance. For the first time, Nvidia‘s x80 class card can deliver solid 4K gaming. Next, let‘s examine Ampere‘s architectural enhancements.
Ampere Architecture Overview
The Ampere architecture includes both improvements to existing cores and new functionality. Key highlights include:
- Enhanced CUDA Cores: Higher clock speeds and efficiencies leading to up to 2x the throughput over Turing for traditional rasterized graphics.
- 2nd-Gen RT Cores: Improved ray tracing acceleration, up to 2x faster than Turing.
- 3rd-Gen Tensor Cores: New AI capabilities and up to 2x the performance for deep learning workloads compared to Turing.
- PCIe Gen 4 Support: Ampere GPUs utilize the faster bandwidth of PCIe 4.0 for additional performance gains.
- New Streaming Multiprocessors: Higher FP32 throughput and double FP16 performance forcompute-heavy workflows.
- Faster GDDR6X Memory: Paired with a 320-bit memory bus on 3080 and 384-bit bus on 3090 for overall higher memory bandwidth.
- AV1 Decode: Hardware acceleration for the efficient AV1 video codec, important for streaming high resolution video.
- HDMI 2.1: Support for 4K 120Hz, 8K HDR and variable refresh rate on compatible displays.
Ampere provides a massive generational jump over the already powerful Turing cards. Now let‘s see how the specs compare between the 3080 and 3090.
RTX 3080 vs RTX 3090: Specifications Comparison
| Specification | RTX 3080 | RTX 3090 |
|---|---|---|
| CUDA Cores | 8704 | 10496 |
| Boost Clock Speed | 1.71 GHz | 1.70 GHz |
| Memory Size | 10GB GDDR6X | 24GB GDDR6X |
| Memory Interface | 320-bit | 384-bit |
| Memory Bandwidth | 760 GB/s | 936 GB/s |
| RT Cores | 68 | 82 |
| Tensor Cores | 272 | 328 |
| TDP | 320W | 350W |
| Price | $699 | $1499 |
On paper, the RTX 3090 looks considerably more capable. The 3090 has 20% more CUDA cores and 10GB more VRAM than the already impressive 3080. Memory bandwidth sees a 23% increase from the wider 384-bit interface. The 3090 distinguishes itself with titanium and gold colored accents, opposed to the all-black 3080 design.
But how do these spec differences actually affect gaming, creative workloads and compute performance? Let‘s investigate with some real-world benchmarks.
Gaming Benchmarks: 1080p, 1440p and 4K Gameplay
Gaming is the most common application for high end consumer GPUs like the 3080 and 3090. Here‘s a look at benchmark results across a range of popular games at the mainstream 1080p, 1440p and 4K resolutions.
1080p Performance
At 1920 x 1080, the RTX 3080 and 3090 are neck and neck when it comes to gaming. Both can far surpass 144 FPS in even the most demanding titles. The cards are simply overkill for Full HD gaming. The 3090 holds miniscule leads in some games, but the differences amount to just a few FPS.
1080p gaming benchmark results (Image credit: Tom‘s Hardware)
1440p Performance
When increasing resolution to 2560 x 1440, the performance gap remains very small. Frame rates are still exceptionally high on both cards. The 3090 continues to maintain insignificant FPS advantages in some games, along with parity in others. Overall, either GPU provides an excellent 1440p gaming experience.
1440p gaming benchmark results (Image credit: TechSpot)
4K Performance
At 3840 x 2160 pixels, the higher specced 3090 finally gets to stretch its legs a bit. However, the performance delta remains smaller than the price difference between the two cards. The 3090 does deliver solidly higher frame rates in demanding titles like Red Dead Redemption 2 and Horizon Zero Dawn. Yet in other games, performance is closer between the GPUs. Both handle 4K gaming very well.
4K gaming benchmark results (Image credit: Tom‘s Hardware)
Performance for Creators: Video Editing, 3D Modeling and Rendering
For creative professionals, the advanced capabilities of Ampere translate into huge performance gains in applications like video editing, 3D modeling/animation and rendering. Often these apps are optimized to take advantage of GPU acceleration.
The 3090‘s extra VRAM, memory bandwidth and CUDA cores give it a definitive edge in creator workflows. However, the 3080 still marks a major upgrade over previous generation cards like the 2080 Ti, thanks to Ampere‘s architectural improvements.
In Puget Systems testing, the 3090 achieved premier prosumer performance for DaVinci Resolve and Adobe Premiere Pro 4K video editing and effects. It scored up to 27% higher than the 3080 in these tests.
For 3D rendering with Blender, the 3090 completed benchmark scenes 25-30% quicker than the 3080. The gap widened further in Blender benchmarks at 4K output resolution.
The 3090 dominates in professional visualization apps as well. For example, in SPECviewperf 2020 testing, the 3090 delivered 47% higher score in SolidWorks and 41% higher performance in Maya.
For creators seeking the ultimate productivity, the RTX 3090 is easily worth the premium over the already speedy RTX 3080.
Power Draw, Thermals and Noise
Having fast hardware is only part of the equation. Heat, noise and power usage determine the overall user experience. Let‘s see how the 3080 and 3090 compare when it comes to these real-world factors.
As expected given its 350W TDP, the RTX 3090 consumes more power than the 320W RTX 3080. However, the Tom‘s Hardware review recorded a lower than expected 45W difference under gaming loads (305W vs 260W). The 3090‘s beefier cooler keeps it a few degrees Celsius cooler as well (69C vs 72C).
Anandtech measured peak power consumption between the cards as a larger 80W gap, with the 3090 topping out at 420W! Both cards exhibit reasonable thermals and noise levels thanks to Nvidia‘s vastly improved cooling system. Just make sure your power supply can keep up with these hungry graphics cards.
Memory Size and Bandwidth Analysis
One of the most prominent differences between the RTX 3080 and 3090 is VRAM – 10GB versus 24GB. For gaming today, 10GB remains perfectly adequate for smooth 4K experiences. Even demanding titles like Microsoft Flight Simulator 2020 used under 8GB in TechSpot‘s testing at 4K ultra settings.
However, for creators working with high resolution videos or complex 3D models, the 3090‘s extra memory capacity removes any limitations. Applications can utilize more assets for enhanced realism and details.
The 3090 also provides 55% more total bandwidth than the 3080 (936 GB/s vs. 760 GB/s). This speed advantage directly translates to better performance in memory intensive workloads as we saw in previous benchmarks.
While unnecessary today, the 3090‘s 24GB buffer may better stand the test of time as games continue increasing texture sizes and world detail. Only time will tell whether 10GB enters a chokepoint before the end of this GPU generation.
Ray Tracing and DLSS Performance
The second generation RT cores enable Nvidia‘s Ampere cards to deliver more performant real-time ray tracing. By offloading ray intersection calculations to dedicated hardware, the GPU‘s CUDA cores are free to render the rest of the scene while ray tracing runs in the background.
Early benchmarks show the 3080 is capable of playing games with ray tracing enabled at over 60 FPS finally, a great leap over Turing. With 82 RT cores, the 3090 further accelerates ray traced games, achieving up to 15% faster ray tracing performance.
DLSS leverages the tensor cores for intelligent upscaling, boosting frame rates with minimal image quality loss. Quality and performance DLSS modes are now joined by Ultra Performance mode on Ampere cards. This pushes DLSS to render at just 25% of the target resolution for the absolute highest FPS.
While the 3080‘s 272 tensor cores handle DLSS well, the 3090‘s added throughput provides a bit more headroom for further AI advancements like Nvidia Broadcast.
PCIe 4.0 Support
Both new GPUs take advantage of PCI Express 4.0 support provided by AMD‘s recent processors and chipsets. PCIe 4.0 doubles the available bandwidth for transfers between the CPU and GPU.
Early testing showed limited real-world impact from upgrading to Gen4. However, future multi-GPU setups may benefit more from the extra headroom. At minimum, the platform PCIe 4.0 compatibility ensures a degree of future-proofing.
Ports and Connectivity Comparison
The RTX 3080 and 3090 Founders Edition models provide an array of output options via their rear port configuration:
- 1x HDMI 2.1
- 3x DisplayPort 1.4a
- USB Type-C for VR headsets
HDMI 2.1 enables support for 4K 120Hz, 8K 60Hz HDR and variable refresh rates with compatible televisions or monitors. DisplayPort is geared for super high resolution, high refresh rate PC displays. Meanwhile, the USB-C port simplifies connecting next-gen virtual reality headsets.
Both cards include the same highly versatile range of outputs for single displays or multi-monitor gaming and productivity setups. The only difference is the RTX 3090 FE‘s second USB Type-C connector lacking on the 3080 FE.
Physical Size Comparison
As expected for its beefier specs, the RTX 3090 has a noticeably larger physical footprint than the 3080. See the dimensions below:
- RTX 3080: 285mm x 112mm, 2-slot
- RTX 3090: 313mm x 138mm, 3-slot
Make sure to measure clearance inside your PC case before opting for the massive 3090. Some mid tower cases won‘t be able to fit the full length of the card. Aesthetically, the 3090FE oozes luxury appeal with its huge die exposed through a partial cutout along the gold-plated cooling fins.
Price to Performance Breakdown
Let‘s break down the value proposition between these flagship GPUs when pricing enters the equation:
- RTX 3080: $699 MSRP – Excellent performance for 4K gaming. Major upgrade over previous x80 cards.
- RTX 3090: $1499 MSRP – Ultimate power for creators and 8K enthusiasts. Around 30% more performance for 2x the cost.
There‘s no question the RTX 3090 is the reigning performance champion. However, the RTX 3080 delivers incredible performance in its own right, even outpacing the former flagship 2080 Ti for hundreds less.
For pure gaming, the 3080 is likely the better buy for most. But professionals who can truly leverage 24GB VRAM and higher clocks in production workloads may find the 3090‘s premium justifiable.
Verdict: RTX 3080 Offers More Bang-For-Buck, 3090 is Ultimate Powerhouse
To conclude, when comparing the RTX 3080 versus RTX 3090, there are reasons to recommend either GPU depending on the use case:
- The RTX 3080 gives gamers and general enthusiasts nearly top-tier performance for $800 less than the 3090. It‘s the sensible value choice for 4K gaming.
- The RTX 3090 is unmatched when it comes to horsepower. The 24GB VRAM, increased cores and memory bandwidth empower creators with prosumer-level performance. If budget allows, the 3090 is the GPU of choice for productivity and future-proofing.
Hopefully this detailed breakdown has helped provide clarity between these two graphics card behemoths. With performance achievements far surpassing their predecessors, both the RTX 3080 and RTX 3090 represent exciting new milestones for PC gaming and creation workflows under Nvidia‘s Ampere architecture.
