RTX 5090 Laptops - I Noticed Something...

Summary

In this video, the speaker discusses the power consumption and efficiency of Nvidia's RTX 50 Series GPUs, focusing on how the 50 Series doesn't showcase the same power efficiency improvements as the previous 40 Series. Highlighting the struggles of integrating these high-power desktop GPUs into laptops, the speaker compares the power limits of gaming laptops and the limitations this imposes on GPU performance. The speaker also critiques Nvidia's marketing around the 5070 laptop GPU, stressing that new hardware should only be purchased if the games you want to play struggle on your current setup, regardless of the new features being marketed.

Takeaways

• 😀 The user upgraded to an RTX 590 GPU, which requires a significant power boost to 1200W, compared to the 450W needed by the previous RX 490.
• 😀 Nvidia's 50 Series GPUs do not seem particularly power-efficient, which was not highlighted during their CES presentation.
• 😀 The 40 Series GPUs had a major focus on power efficiency, as Nvidia showcased their significant node shrink and energy improvements.
• 😀 Desktop GPUs can consume more power, but laptops are severely limited in terms of power supply and cooling capacity, especially for GPUs.
• 😀 High-end gaming laptops are generally capped at 175W for GPU power, such as MSI's Titan or the Asus ROG Strix Scar, which impacts their performance.
• 😀 The Nvidia 50 Series GPU for laptops did not see a significant improvement in power efficiency or performance over the 40 Series, despite a minor increase in core count.
• 😀 Nvidia’s heavy reliance on AI technologies like DLSS 4 and multiframe generation to drive gaming performance, as opposed to relying on further node shrinks.
• 😀 DLSS 4 introduces promising new features, including improved image quality and a new Transformer model, but many games still don’t support this technology.
• 😀 There is an argument that the gaming industry needs AI-driven solutions like DLSS to progress, as hardware improvements via process node shrinks hit a limitation.
• 😀 When considering new hardware, it’s crucial to evaluate whether the games you want to play are poorly supported by your current setup, rather than jumping to the latest GPU models based on marketing claims.

Q & A

Why did the speaker need to upgrade the power supply when installing the RTX 590?

The speaker needed to upgrade the power supply to 1200 watts because the RTX 590 requires 575 watts of power, which is significantly higher than the 450 watts required by the previous GPU, the RX 490.

What concerns did the speaker have about the new 50 Series GPUs?

The speaker was concerned that the new 50 Series GPUs might not be power-efficient, as Nvidia did not provide details on power efficiency during their CES presentation, unlike previous generations.

What did Nvidia emphasize during the launch of the 40 Series GPUs?

During the 40 Series launch, Nvidia highlighted the significant improvements in power efficiency and energy throughput, especially due to a major node shrink in the fabrication process.

How does power efficiency differ between desktop and laptop systems?

In desktop systems, power efficiency is important for performance improvements but not critical, as users can simply upgrade the power supply. However, in laptops, the power supply is limited, making power efficiency a much more critical factor for performance.

What is the power cap for laptop GPUs, and how does it affect performance?

Laptop GPUs are typically capped at 175 watts for high-performance models, such as MSI's Titan and Asus's Strix Scar, and even lower for thinner models like the G16. This cap severely limits GPU performance compared to desktop systems, where more power can be fed into the GPU.

Why did Nvidia focus on power efficiency for the 10 Series and 40 Series GPUs?

Nvidia emphasized power efficiency for the 10 Series and 40 Series because these generations benefitted from significant node shrinks in the fabrication process, allowing for better energy efficiency, reduced heat, and overall improved performance.

What makes the 50 Series GPUs different from previous generations in terms of node shrink?

The 50 Series GPUs did not receive a significant node shrink like the 10 and 40 Series. The fabrication process only shrank from 5 nanometers to 4 nanometers, which did not lead to the same substantial improvements in power efficiency or performance as previous generations.

What role does AI and software like DLSS play in the performance of the 50 Series GPUs?

Since the 50 Series GPUs did not see a major node shrink, Nvidia relies heavily on AI-driven technologies like DLSS (Deep Learning Super Sampling) and multiframe generation to improve gaming performance, despite the hardware's limitations.

How does the marketing of the 5070 laptop GPU compare to its actual performance?

The marketing of the 5070 laptop GPU claimed that it offered 4090 laptop performance for half the power, but the speaker believes this is only true in very specific situations with DLSS4 enabled. This may not accurately reflect the overall performance of the two products in all games.

What advice does the speaker give when considering purchasing new hardware?

The speaker advises buying new hardware only if the games you want to play do not run well on your existing system. This principle is especially important in years when there are many cool features, as marketing can sometimes be misleading.