My 4K at 100 FPS! Only requires 300W of power, NVIDIA GeForce RTX 4080 unboxing test report/flip RTX 3080 Ti

NVIDIA GeForce RTX 4080

NVIDIA’s second gaming graphics card that can reach 4K and 100 FPS, NVIDIA GeForce RTX 4080 uses a new generation of Ada Lovelace GPU architecture, bringing Tensor Core and RT Core upgrades, as well as DLSS 3’s AI supplementary frame technology and AV1 dual encoding engine, Satisfies the powerful performance required by creators and gamers, and has an excellent performance per watt upgrade compared to the previous generation, but the price starts at $1199 in US dollars and NT$ 42,990 in Taiwan dollars, which is also a must-have for this generation of outstanding performance. necessary evil.

The second 4K, 100 FPS gaming graphics card NVIDIA GeForce RTX 4080

The second NVIDIA Ada Lovelace GPU generation graphics card that can meet 4K, 100 FPS gaming graphics card NVIDIA GeForce RTX 4080, with 76 groups of SM, 9728 CUDA cores, 304 Tensor Core and 76 RT Cores also have a Boost clock rate higher than 2.5GHz, and 16GB GDDR6X high-speed memory, and only need 320W TGP power consumption.

The RTX 4080 uses the AD103 GPU but does not use a complete core. The RTX 4080 only has 76 sets of SM units, which means that the future RTX 4080 Ti may use the full AD103 GPU to have 80 sets of SM units.

RTX 4090 and RTX 4080 specification comparison.
AD103 complete core architecture diagram.

Compared with the previous generation RTX 3080 Ti, the RTX 4080 can provide a performance upgrade of 1.28-1.59x times compared with the previous generation RTX 3080 Ti, and has lower power consumption and better GPU cooling performance; compared with the RTX 4090 performance The gap is about -20%, so the RTX 4090 and RTX 4080 have the same price/performance ratio, allowing flagship players to have more choices in high unit price graphics cards.

Ada Lovelace: Process clock upgrade, advanced ray tracing and dual AV1 Encode

In addition to improving the SM unit, this generation of Ada architecture also has GDDR6X high-speed memory, the 4th generation Tensor Cores to improve AI inference performance, the 3rd generation RT Core to improve the quality of ray tracing, and the 8th generation video encoder to support AV1 hardware encoding features, as well as a 2-4x performance upgrade brought by DLSS 3.

First of all, Ada’s 4th generation Tensor Core can bring 2x times the Tensor TFLOPS performance improvement of FP16, BF16, TF32, INT8 and INT4. At the same time, the FP8 Transformer Engine with Hopper architecture can provide a Tensor Core performance of 1.3 PetaFLOPS.

Ada Lovelace.

The third-generation RT Core can bring 2x faster Ray-Triangle Intersection output performance (compared to the previous generation Ampere), and at the same time add new technologies such as the new Opacity Micro map Engine, Displaced Micro-Mesh Engine and Shader Execution Reordering, which can once again Improve the performance of ray tracing.

Opacity Micromap Engine, which allows objects to have transparent, translucent, and opaque attributes to speed up the performance of ray tracing processing.
Displaced Micro-Mesh Engine, the object is represented by a simpler BVH, and the ray tracing effect of the object is quickly calculated based on the vector map.
Shader Execution Reordering, which can optimize the scheduling performance of SM processing ray tracing.

In terms of creation, Ada Lovelace has the 8th generation dual NVENC encoding engine, which mainly adds the video and audio encoding function of AV1, and this generation of audio and video output can achieve a 2x performance improvement, which requires the support of video editing software such as DaVinci Resolve, Voukoder, and Jianying. The dual encoding engine of RTX 40 is supported for the first time, and the mainstream Adobe Premiere Pro will have to wait for an update in the future.

8th generation dual NVENC encoding engine.

DLSS 3 and Optical Flow Accelerator

RTX 40’s unique “DLSS 3” is based on DLSS 2 technology, adding the concept of “AI supplementary frame”, which is the function of an Optical Flow Accelerator. Optical Flow is an optical flow method used in computer vision, which is used to calculate the moving direction and moving amount of each pixel in continuous images.

DLSS 3 technology requires the game engine to provide: lower-resolution rendering images and Motion Vectors, deduce high-resolution images through the deep learning network of DLSS, and provide the images to the Optical Flow Accelerator to calculate the moving direction of each pixel and the amount of movement, and finally through Optical Multi Frame Generation to generate an AI supplementary frame picture.

DLSS 3 can provide a 2-4x improvement in game performance through AI supplementary frames while maintaining a similar image quality to native rendering, but it will also increase the overall delay of the game, so NVIDIA forces DLSS 3 to include Reflex technology, by cancelling Render Queue allows the CPU to process the GPU immediately to take over the rendering, achieving lower system latency.

Therefore, DLSS 3 combines technologies such as AI Super Resolution, Frame Generation and ReFlex, relying on the 4th generation Tensor Core, Optical Flow Accelerator, and NVIDIA’s supercomputer used to train AI, to meet the ultimate performance of 4K and 100 FPS for next-generation gamers.

DLSS 3.

NVIDIA GeForce RTX 4080 Founder’s Edition graphics card out of the box / The back is the front Classic and then enhanced

NVIDIA GeForce RTX 4080 Founder’s Edition continues the same design as the RTX 4090 Founder’s Edition, “the back is the front”, “less but better”. And the new outer packaging is also quite special, using 2 triangular, cardboard outer boxes, concisely printed with the lines of RTX 4080 and classic X frame.

After opening, the RTX 4080 Founding Edition is placed on a slope with radial lines, which looks like a design with a background of a stem. It is not so much buying a graphics card as buying a computer boutique.

Founder’s Edition’s distinctive box.
RTX 4080 with a radial background.
There are accessories and power adapter cables in the small inner drawer.

The RTX 4080 Founders Edition uses a strong and durable aluminium alloy to create the X-Frame frame, and the surface is anodized for a high-end texture and gold metal finish.

The inside of the frame is filled with heat dissipation fins, and the inside is a vapour chamber to dissipate heat for GPU and VRAM, and then the heat pipe guides the waste heat to the heat dissipation fins. The RTX 4080 Founding Edition uses a larger 116mm, FDB, 7-blade dual fan, increases the thickness of the graphics card to 3-Slot, and reduces the length of the graphics card to 30.48cm (12 inches).

This generation of vapour chamber is also optimized and has a dedicated cutout for the memory so that the vapour chamber can be more evenly in contact with the GPU, and the heat conduction pad of the memory is reduced to 1.5mm for better heat conduction effect; this The first-generation radiator can support up to 650W Qmax heat dissipation capacity.

The front appearance of the RTX 4080 is more refined and detailed than the previous generation.
RTX 4080 radiator, it can be seen that there are cooling fins embedded in the X frame.
On the upper side of the graphics card, there is a Logo light with the words GEFORCE RTX and a PCIe 12+4 Pin (12VHPWR) power supply interface.
The magnet on the front of the graphics card absorbs the hidden fixing lock hole.
The lower side of the graphics card.

The RTX 4080 and the RTX 4090 use the same PCIe 12+4 Pin (12VHPWR) power supply interface, which can transmit a maximum power consumption of 600W. The RTX 4080 accessories provide a 12VHPWR to 3 PCIe 6+2pin cables.

It is recommended to connect at least 3 PCIe 6+2pins for conversion when installing the machine. If you buy a new power supply, it is recommended to choose a new power supply that meets the ATX12 V3.0 and EPS12V V2.92 specifications. A 12VHPWR cable can provide the power required by the graphics card.

RTX 4080 12VHPWR to 3 PCIe 6+2pin cables.
When connecting the wire, make sure that the entire plug is inserted into the socket.
And don’t bend too much at the outlet of the wire.
Or with a native 12VHPWR cable, once and for all.

The RTX 4080 display output provides 1 HDMI 2.1a supporting VRR, 4K120Hz / 8K60Hz HDR, 3 DisplayPort 1.4a DSCs supporting 12-bit 4K240Hz HDR / 12-bit 8K60Hz HDR and other output capabilities and can connect up to 4 screen outputs at the same time.

RTX 4080 display output.

NVIDIA GeForce RTX 4080 creative audio and video output, GPU rendering performance test

This test includes creation tests such as Adobe Premiere Pro 2020, DaVinci Resolve 18 and Blender, games are tested at 2160p, and 1440p resolution, special effects fully open, e-sports, AAA games and lighting Chasing the performance of the game, as well as the related tests of DLSS 3, and comparing the RTX 4090 and RTX 3080 Ti at the same time so that players have more data reference.

Test Platform
Processor: Intel Core i9-13900K
Motherboard: ASUS ROG MAXIMUS Z790 HERO 0502
Memory: G.SKILL TRIDENT Z5 NEO DDR5-6000 16GBx2
Graphics Card: NVIDIA GeForce RTX 4090 Original Edition, NVIDIA GeForce RTX 4080 Original Edition, NVIDIA GeForce RTX 3080 Ti Original Edition
System Disk: Solidigm P41 Plus 1TB PCIe 4.0 SSD
Cooler: Phanteks Glacier One 360MPH
Power Supply: Seasonic PRIME PX-1000
Operating System: Windows 11 Pro 21H2 64bit, Resizable BAR On
Driver Version: NVIDIA 526.72

GPU-Z has not been able to view the information of NVIDIA GeForce RTX 4080, which uses AD103 GPU with 4nm process, with 9728 rendering CUDA cores, and 16384 MB GDDR6X (Micron) memory, while the GPU preset clock frequency is 2205 MHz and Boost 2505 MHz.

GPU-Z.

DaVinci Resolve 18 is a purely GPU-accelerated video editing program, including powerful colour correction and special effects functions, and directly uses CUDA core computing, so that the playback and output of video clips have very good performance. The beta version includes support for NVIDIA AV1 encoding.

DaVinci Resolve 18.

First of all, the first test project uses 4K Blackmagic RAW images and has a Wedding_Heavy_Styles timeline. This video uses a lot of Resolve effects, such as OFX: Light Rays / Glow / Sketch, etc., to output a very high-style video type.

Bride_FaceRefine_Selective_Color uses Face Refinement for face tracking and highlights the main bride with colour; both 50% Retime and Optical Flow Enhanced Better use Optical Flow technology to reduce the image speed by 50%.

SuperScale2x 4K Source uses 4K ProRES source video to produce 4K video output of 2x Zoom In subject; SuperScale4x HD_Source uses HD H.264 source video and uses Resolve Super Scale to output 4K videos.

The output performance is definitely better than the RTX 4090, but the performance of the RTX 4080 should not be underestimated. It depends on whether your project will use such a high memory capacity of the RTX 4090. If it is a common video type, the RTX 4080 can still give a Not bad performance.

DaVinci Resolve 18, the shorter the better.

The second test is the AV1 and HEVC encoding test with dual NVENC encoding. The test project is a 44-second short film from the Blender Open Movie Project “Tears of Steel”, and has 8k Prores442HQ 30FPS and 4K Prores422HQ 30FPS videos, available To test the output performance of HEVC, AV1 encoding.

Output settings mainly use NVIDIA Encoder, Quality: Restrict to 80000 Kb/s, Encoding Profile: Main, Rate Control: Constant Bitrate, Preset: Faster, Tuning: High Quality, Two Pass: Disable and other output settings.

In terms of performance, there is not much difference between RTX 4090 and RTX 4080 in 4K30 output, but compared with the previous generation RTX 3090 Ti, the HEVC encoding time is saved by as much as 2x times.

Especially in the case of 8K HEVC output, RTX 4090 and RTX 4080 are directly released to the previous generation of RTX 3080 Ti, which undoubtedly shows the advantages of dual encoding engines, but relative software support is required to liberate this performance.

DaVinci Resolve 18, the shorter the better.

Adobe Premiere Pro 2022 video editing software is accelerated by the self-developed Mercury Playback Engine GPU, which can accelerate the video output speed with the help of the GPU encoding engine. The project used in the test is the company’s 1080p60fps out-of-the-box video, and the BigMix4K project uses three FinalAdjusted_MPE 1920×1080 images to form a 4K timeline for H.264 and HEVC format output.

(The tested Premiere Pro 2022 does not yet support the RTX 4090 dual encoding function.)

Since Premiere Pro 2022 does not yet support the RTX 40 dual encoding engine, the performance of the test is not significantly different from the previous generation RTX 3080 Ti. We need to wait for Adobe to provide software Only after the update can it show the output performance of the new generation GPU.

Adobe Premiere Pro 2022.
Adobe Premiere Pro 2022 output, the shorter the better.

Blender is a cross-platform, open-source 3D creation tool that supports various 3D operations: Modeling, Rigging, Animation, Simulation, Rendering, Compositing and Motion Tracking, etc. For testing, use Blender Benchmark 3.3.0 to test the rendering work of the Demo project.

According to the Blender Benchmark 3.3.0 test, the RTX 4080 has a 1.48x increase in computing performance compared to the RTX 3080 Ti, but a -24% reduction in computing performance compared to the RTX 4090.

Blender, the more performance the better.

V-Ray Benchmark is developed by Chaos Group. V-Ray is a ray rendering software designed based on the laws of physics, and this tool can perform calculation tests on ray-traced rendered images for CPU and GPU respectively.

According to the V-Ray test, RTX 4080 has a 1.4x performance improvement compared to RTX 3080 Ti, and a -30% reduction in computing performance compared to RTX 4090.

V-Ray Benchmark, the higher the performance, the better.

SPECviewperf 2020 is a standard drawing performance test tool developed based on professional applications to test various professional computer graphics software such as: 3ds Max, Catia, Creo, Energy, Maya, Medical, SNX, SolidWorks and other drawing tests and engineering simulations.

Tested at 1920 x 1080 resolution and scored in FPS. The performance depends on the tools used. The performance difference between RTX 4080 and RTX 4090 is about -15%; but compared with RTX 3080 Ti, it has about 1.3x performance improvement.

SPECviewperf 2020.

NVIDIA GeForce RTX 4080 – 3DMark Benchmark Performance Test

3DMark Fire Strike performance test is a test scenario for the mainstream DirectX 11 API, testing the performance of 1080p, Extreme 1440p and Ultra 2160p respectively.

RTX 4080 achieved a score of 46013 points in Fire Strike, while Ultra Graphics was 1.37x faster than RTX 3080 Ti, and lost about -31% to RTX 4090; while Extreme Graphics was about 1.39x faster than RTX 3080 Ti , Losing to RTX 4090 about -25%.

3DMark Fire Strike, the higher the score, the better.

3DMark Time Spy is a test scenario designed using DirectX 12 API, which is also locked at the AAA game level, and tests the performance of 1440p and Extreme 2160p respectively.

RTX 4080 achieved a total score of 27569 points in Time Spy, which has a 1.4x performance improvement compared to RTX 3080 Ti, and also lost about -26% to RTX 4090.

3DMark Time Spy, the higher the score, the better.

For ray tracing tests,3DMark Port Royal adds ray tracing function to scenes in AAA games, testing the ability of the new generation GPU to accelerate hardware ray tracing. At the same time, the DXR test is a functional test using the DirectX Raytracing API.

Even without DLSS, RTX 4080 can have amazing ray tracing performance. Port Royal achieved 82.3 FPS and DXR 84.2 FPS. Compared with RTX 3080 Ti, it has a 1.4x improvement in ray tracing performance, but it also loses to RTX 4090 by about -35% ray tracing performance.

3DMark Port Royal, the higher the better.

3DMark DLSS Feature Test can perform performance tests for DLSS 3 and DLSS 2, set to 3840 x 2160, Performance acceleration settings.

RTX 4080 can achieve 102.14 FPS about 2.6x performance improvement in DLSS 2, and DLSS 3 can achieve 149.69 FPS about 3.8x performance improvement by using AI supplementary frame technology.

3DMark DLSS Feature Test, the higher the better.

NVIDIA GeForce RTX 4080 – 4 e-sports games performance test

The 4 e-sports games “Rainbow Six: Siege”, “League of Legends”, “APEX Heroes” and “CS:GO” are all skills-heavy, team-based tactical competitive shooting, and DOTA-type games. When the quality and details are not high, the game FPS is also an average performance of more than 100 frames. Tests were conducted at 2160p, 1440p, and special effects maximum settings.

For e-sports games, the performance of RTX 4080 is still quite powerful, but the performance in “CS:GO” is currently low, and NVIDIA is solving it. For e-sports games, 4K and 400 FPS are not a problem.

2160p e-sports game test, the higher the FPS, the better.
1440p e-sports game test, the higher the FPS, the better.

NVIDIA GeForce RTX 4080 – 11 Games Performance Test

The average performance of 11 AAA games, also tested 2160p, 1440p, special effects fully turned on for testing, this test only F1 2021 uses the ray tracing function, the rest of the games have no ray tracing, no DLSS acceleration, Tests the GPU’s actual traditional rendering game performance.

The game test list includes the entry-level “F1 2021”, “Forza Horizon 5” racing game, “Tomb Raider: Shadow”, movie game “Death Stranding”, “Gears of War 5”, “The Division 2″, ” Horizon: Waiting for Dawn, as well as tests such as “Borderland 3”, “Assassin’s Creed: Viking Age”, “Blood Killing 2” and “God of War” that are heavy on performance.

The RTX 4080 achieves an average of 114.4 FPS in 2160p and AAA games. Compared with the average 89.3 FPS of the RTX 3080 Ti, it can achieve an average performance upgrade of about 1.2x. Compared with the RTX 4090, it is a small loss of -24% in game performance.

Then at 1440p resolution, RTX 4080 averages 186 FPS, which is 1.2x faster than RTX 3080 Ti, and loses about 15% to RTX 4090.

2160p AAA game test, the higher the FPS, the better.
1440p AAA game test, the higher the FPS, the better.

NVIDIA GeForce RTX 4080 – 8 ray chasing game tests

8 light-chasing DXR game tests, using the most popular “Dian Yu Ren Ke 2077”, “Control”, “Watch Dogs: Liberty Legion”, “Thrilling Deep: Exile”, “Marvel Spider-Man Remake”, Games such as “Marvel Interstellar”, “Ghostwire: Tokyo” and “Far Cry 6” are tested. Test 2160P, 1440p resolution, special effects/ray tracing highest setting, DLSS acceleration will also be enabled, please refer to the chart for detailed settings.

The RTX 4080 can achieve an average of 113 FPS at 2160p and DLSS 3 acceleration of “Dian Yu Ren Ke 2077”, while the average of 8 light-chasing games reaches 111.6 FPS, which is 1.4 times the game performance upgrade compared to the RTX 3080 Ti. Compared with the RTX 4090 about -21% performance reduction.

As for the 1440p resolution, the RTX 4080 has an average of 163.4 FPS, which has a 1.ˇx times game performance upgrade compared to the RTX 3080 Ti, and a -14% reduction in performance compared to the RTX 4090.

2160p ray tracing game test, the higher the FPS, the better.
1440p light-chasing game test, the higher the FPS, the better.

NVIDIA GeForce RTX 4080 – DLSS 3 performance measurement

RTX 40 update is a major focus of “DLSS 3”, the games tested include “Microsoft Flight Simulator”, “A Plague Tale: Requiem”, “Marvel Spider-Man Remake”, ” F1® 22″, “Unity Enemies” and “Dian Yu Ren Ke 2077”, using 2160p resolution and the highest setting of ray tracing.

In the DLSS 3 game settings, there will be clear options for “Super Resolution” and “Frame Generation”. Both of these functions must be enabled at the same time to use the technology of DLSS 3, while RTX 30 / 20 series players can only enable it Super Resolution function, Frame Generation will not be enabled.

“Dian Yu Ren Ke 2077” DLSS 3 settings.

Accelerated by RTX 4080 through DLSS 3, “Dian Yu Ren Ke 2077” can achieve an average performance improvement of 109 FPS about 4x times, and the Enemies movie animation released by the Unity engine can also reach 75 with DLSS 3 under real-time light tracing rendering About 3.2x performance upgrade of FPS.

Under the setting of DLSS 3 Performance, the RTX 4080 can achieve a performance improvement of about 1.9x~4x times, and the average is about 2.48x times.

DLSS 3 game performance test, the higher the better.

NVIDIA GeForce RTX 4080 power consumption and temperature measurement

The power consumption and temperature test of the graphics card, using Time Spy Stress test, Furmark and “Dian Yu Ren Ke 2077” for testing. When measuring power consumption, use the PACT tool provided by NVIDIA to monitor the wattage provided by the PCIe slot and the power supply 12V.

In terms of graphics card temperature, the RTX 4080 Founder Edition maintains a maximum temperature of 66.1°C in the stress test, and the temperature of the 2077 game will be slightly lowered by 63°C. Compared with the 73°C of the previous generation RTX 3080 Ti, the temperature performance of this generation can be maintained. Said to be pretty good.

RTX 4080 Founder Edition GPU temperature.

Graphics card TBP power consumption test, in the Time Spy Stress test, the RTX 4080 reached an average power consumption of 293.3W, while the Furmark 4K Xtreme burn-in test reached a maximum of 317.2W, but 2077 only consumes 280W of power consumption during games, compared to the RTX 3080 With Ti’s 360W power consumption, it can only be said that the RTX 4080 has been upgraded quite beautifully.

RTX 4080 Founder Edition GPU power consumption.

Summarize

NVIDIA once again handed over the second 4K, 100 FPS game graphics card GeForce RTX 4080, 11 AAA games averaged 114.4 FPS, 8 ray chasing games achieved brilliant results of 111.6 FPS, including DLSS 3 can also bring 1.9x~ 4x performance improvement, and firmly beats the comparative RTX 3080 Ti by about 1.4x performance.

The performance of RTX 4080 is about -25% lower than that of RTX 4090, but the price difference between the two is also about 24%. Therefore, when the price and performance ratio of the two are comparable, it depends on the performance requirements of creators and gamers. , The capacity of video memory is determined, but RTX 4080 only needs 300W power consumption to have 4K, 100 FPS performance.

The price of the first wave of RTX 4080 in Taiwan also ranges from a suggested price of NT$ 42,990 to a maximum of NT$ 49,990. Although the price in US dollars is comparable to that of the RTX 3080 Ti at that time, the price of the RTX 4080 in Taiwan is also increasing. , For players who pursue 4K games, undoubtedly need more budget for buying cards.

According to this price range, the future RTX 4070 may start at 30,000, and the RTX 4060 will start at 20,000. Can this really meet the expectations of ordinary players for game graphics cards? The RTX 40 series adopts the TSMC 4N process and the new Ada Lovelace architecture, which brings a solid performance improvement but also makes the price of the graphics card soar, so I ask the flagship gamers whether the budget is enough.

If this article is helpful for you, please share this article with your friends on social media. Thank you!!

This article is based on the personality of the reviews. You are responsible for fact-checking if the contents are not facts or accurate.

Title: My 4K at 100 FPS! Only requires 300W of power, NVIDIA GeForce RTX 4080 unboxing test report/flip RTX 3080 Ti

en_GBEnglish