NVIDIA GeForce RTX 5080 Founders Edition Review – GeForce RTX 4080 Ti Super with Blackwell genes

1 - Introduction and details of the Blackwell GB203-400-A1 GPU 2 - Test system and equipment 3 - Gaming: Full-HD 1920x1080 Pixels (Rasterization Only) 4 - Gaming: WQHD 2560x1440 Pixels (Rasterization Only) 5 - Gaming: Ultra-HD 3840x2160 Pixels (Rasterization Only) 6 - Gaming: WQHD 2560x1440 Pixels, Supersampling, RT & FG 7 - Gaming: Ultra-HD 3840x2160 Pixels, Supersampling, RT & FG 8 - DLSS4 and MFG: Cyberpunk 2077 in detail 9 - DLSS4 and MFG: Alan Wake 2 in detail 10 - PCIe Gen5 problems, power consumption and standards 11 - Load peaks and power supply recommendation 12 - Cooler, temperatures, thermography, noise development 13 - Summary and conclusion

The cooler of the RTX 5080 Founders Edition

The GeForce RTX 5070 Founders Edition (FE) is characterized by an innovative cooling design that combines high efficiency with a compact design. Compared to its predecessor, the RTX 4080 Super, which required three slots, the RTX 5080 only needs two slots. This was achieved by introducing a double-sided flow-through cooling system. Two axially arranged fans draw in cool air from below and discharge heated air upwards. This “double flow through” design significantly improves airflow, allowing for more even heat dissipation while reducing noise.

As with the RTX 5090, the design of the card is based on a short PCB placed centrally in the middle. This positioning creates space for an optimized arrangement of heat pipes and large cooling fins. The heat generated by the GPU is efficiently dissipated by the surrounding heatpipes, while the fin structure optimally directs the airflow. These measures not only reduce the temperature of the card, but also ensure even cooling of all critical components.

In addition, the fan design has been revised to further increase cooling performance. The axial fans are designed to ensure high airflow with low noise levels. The combination of central PCB positioning, innovative heatpipes and the improved fan design makes it possible to keep the high power consumption of the card stable without overheating or a significant increase in noise, as will be seen in a moment. Nevertheless, the whole thing looks rather “overengineered” and makes the card more of a collector’s item. But the gap dimensions are cool, even if it’s all just simple and surface-treated aluminum.

Temperature curves

The temperature curve of a GPU varies greatly depending on whether it is operated in a constant stress test or in a gaming scenario with highly fluctuating loads. These differences are due to the complex power management of modern GPUs, which is made possible by technologies such as power gating, separate power rails and accelerated frequency switching, which are integrated into the NVIDIA Blackwell architecture. During a constant stress test, the GPU operates under a constant full load, resulting in constant power consumption and stable heat generation. In this scenario, the temperature curve remains relatively smooth, as all functional units of the GPU are permanently active and the energy-saving mechanisms are hardly used. The even heat development results from a constant energy supply and stable heat dissipation through the cooling system.

Gaming scenarios, on the other hand, are characterized by dynamic load changes. Some frames require intensive calculations, while others are less demanding. Technologies such as power gating allow the GPU to selectively switch off unused units, reducing energy consumption and heat generation in less demanding phases. Separate power rails ensure that critical components such as the memory remain continuously supplied, while the GPU cores can be almost completely deactivated during idle periods. This targeted control leads to temperature fluctuations, as the GPU constantly adapts to the changing performance requirements.

Accelerated Frequency Switching is a central mechanism here. This technology enables the GPU to adapt its clock frequency and voltage to the respective workload within microseconds. This optimizes energy consumption considerably, but generates irregular heat development. If the GPU switches between low and high frequencies several times in a short period of time, these dynamic changes are reflected in a strongly varying temperature curve. Such fluctuations occur because the heat development reacts directly to the abrupt changes in energy consumption. These differences in the temperature curve between constant and dynamic loads illustrate the efficiency and complexity of modern power management technologies. While constant loads favor high thermal stability, dynamic scenarios demonstrate the GPU’s ability to adapt flexibly and energy-efficiently to variable requirements – a key factor in the performance and longevity of modern GPUs.

The memory temperatures of the GeForce RTX 5080 remain remarkably constant thanks to the cooling design, even during the dynamic load changes caused by the GPU. This stability is ensured by a separate power supply via separate power rails, which ensures that the memory is continuously supplied with a stable power supply, regardless of the fluctuations in the GPU load. Another decisive factor is the central placement of the printed circuit board (PCB), which is surrounded by large heat pipes and an efficiently designed cooling fin array. These components enable even heat dissipation and help to avoid thermal hotspots in the memory area. While the GPU core exhibits significant temperature fluctuations due to its dynamic adaptation to load changes, the memory area remains thermally stable thanks to the constant power supply and optimized air circulation.

The cooling design of the RTX 5080 effectively minimizes temperature fluctuations in the memory area, even under demanding conditions such as intensive gaming or workloads with high memory usage. With maximum memory temperatures below 76 °C during gaming, the card easily stays within a thermally safe range. This efficient cooling not only ensures high stability and longevity of the memory chips, but also contributes to the overall performance and reliability of the card, even in scenarios with highly fluctuating performance requirements.

The clock rates under full gaming load are at a maximum of 2670 MHz and only reach almost the 2900 MHz mark at minimum load. I am very sure that board-parter cards should easily exceed these values.

Thermography during load operation

Thermography with the Optris PI 640 enables precise temperature measurements even with an undisassembled GeForce RTX 5080 and provides detailed insights into the heat distribution on the card surface. This compact infrared camera offers a high optical resolution of 640 x 480 pixels and a thermal sensitivity of less than 75 mK, allowing even the smallest temperature differences to be visualized. Thanks to the metallic and well-coated surface of the graphics card, temperatures can be reliably recorded, provided the emissivity is set correctly. The housing material has defined emissivity values that enable accurate measurement if comparative measurements have previously been carried out with high-quality K-type sensors. With careful calibration of the camera system, the heat flow can be precisely mapped, making hotspots and areas with even heat dissipation clearly identifiable.

The GeForce RTX 5080’s “Double Flow Through” cooling design ensures efficient and even heat dissipation. This property is reflected in the thermography by homogeneous temperature zones, which confirm an effective distribution of waste heat. Despite the card not being open, the measurements make it clear that the cooling structure of the card as a whole works excellently. The GPU and voltage converters in particular, which were identified as areas with higher heat development, show moderate, well-controlled temperature peaks. The memory temperatures, on the other hand, remain stable and are evenly distributed, which further underlines the efficiency of the cooling system. The Optris PI 640, which operates at a frame rate of up to 32 Hz, also allows dynamic temperature changes to be recorded, which is ideal for scenarios such as gaming and stress tests. The card was tested in three scenarios: in idle mode, in gaming mode and during a torture test, each after 30 minutes of operation.

Noise emissions (“volume”) and fan speeds

The analysis of the GeForce RTX 5080’s fan curves and their correlation with the audio measurements in an acoustic measurement chamber provides valuable insights into the optimization of the card’s noise development. The fan curves of the RTX 5080 are designed to ensure a good balance between cooling performance and noise level. This is supported by the double-sided flow design of the fans, which enables efficient airflow at a relatively moderate speed. Nevertheless, the cooler on the GeForce RTX 5080 FE is significantly less powerful than the one on the RTX 5090, as illustrated by the measured fan speeds in the respective load scenarios:

The measurements in the acoustic chamber show that the noise development of the GeForce RTX 5080 depends significantly on the fan speed and the general load. At low to medium speeds, the background noise remains acceptable, which is due to the aerodynamically optimized fan blades and the low-vibration bearing. These optimizations effectively reduce air turbulence and mechanical resonance. Under intensive load, however, when higher fan speeds are required, the noise level increases.

A specific acoustic feature of the RTX 5080 FE is a slight humming at around 200 Hz, which was detected in the tests. This humming is caused by minimal vibrations or resonances, which can be triggered at certain fan speeds or by the voltage converters in the power supply area. It is characterized by a low, humming frequency that is perceptible but should not be disturbing for most users. Resonances of this type often occur at fixed speed ranges or constant load conditions, as mechanical and electronic components reach certain natural frequencies.

The spectrogram, which visualizes the noise frequencies over time, clearly shows the presence of this 200 Hz component. Interestingly, the curve of the peak frequency is partially interrupted over time, which indicates that the humming does not occur continuously with varying fan speed (oscillation). The noise levels measured in dB(A) correlate closely with the temperature curves, which indicates that the fan curves of the RTX 5080 were specifically designed to achieve a balance between efficient cooling and minimal noise pollution. Despite the perceptibility of the humming, the overall noise development remains just pleasant even under load, but it has also been quieter in the past. Just under 38 dBA is almost too much.

The phenomenon of coil beeping in the GeForce RTX 5080 occurs somewhat less in the RTX 5080 FE and can also be indirectly attributed to the load changes. The coil noises are caused by mechanical vibrations in the induction coils of the VRMs, which are caused by the rapid changes in the current flow. The intensity of the coil beeping thus depends heavily on the load situation of the GPU. Under high loads, especially in applications with very high frame rates (FPS), the current consumption of the GPU increases significantly. In such scenarios, the voltage converters operate at maximum efficiency, which keeps the switching frequencies constantly high and increases the mechanical vibrations in the coils. This makes the noise more noticeable, especially in open housings or in very quiet environments.

Interestingly, the use of Multi-Frame Generation (MFG), a technology of the NVIDIA Blackwell architecture, reduces coil noise. MFG allows the GPU to process multiple frames simultaneously, flattening the load peaks. This results in less abrupt power consumption and smoother operating conditions for the voltage converters, significantly reducing the intensity of coil noise. This technology not only helps to increase the performance of the GPU, but also to improve the acoustic user experience. Of course you can listen to it now (slightly amplified):

Here again the direct comparison to the slightly beeping RTX 5090 FE: