Edited By
Marcelo Rodriguez

In a recent discussion, a growing group of people are debating the merits of the RTX 3090 against the newer 5060 Ti, stirring up conversations around future-proofing and performance in demanding tasks like handling Comfyui.
With the rise of graphic-intensive applications and generative models, users are grappling with the speed and efficiency of GPU performance. One enthusiast, currently using an AMD RX 7900 GRE with 32GB RAM, is considering making a switch to either a used RTX 3090 or a new 5060 Ti due to inadequate generation speeds.
The focus on performance has sparked some contrasting opinions:
VRAM Capacity: Many comments emphasize the significance of the 3090's 24GB RAM. One user stated, "24GB RAM really matters," highlighting the advantage in memory-intensive tasks like model training.
Speed of Performance: The RTX 3090 is praised for being faster overall, with comments noting its ability to handle larger models effectively. One enthusiast claimed, "The 3090 is faster it can handle bigger models."
Value vs. New Technology: While several advocate for the 3090, some believe the 5060 Ti offers newer technology with features like native FP4 and FP8 support. A user remarked, "Though the 5060 Ti does have native support, the slow memory bandwidth is a downside."
"Get whichever is the best deal" - One user suggests when comparing prices of these GPUs.
Community sentiment leans towards the RTX 3090, with many highlighting its superior specs despite the newer options available.
๐ฝ The 5060 Ti considered for its newer architecture but is slower
๐ฌ "Performance-wise, the 3090 wins hands down" - A userโs verdict on the competition
As technology continues to advance, the question remains: Is the additional price for better performance worth it? With reports of rapid developments in the GPU market, many savvy buyers are opting for the 3090 in hopes of a solid return on investment.
As technologies evolve, there's a strong chance we'll see a wider gap in performance between high-end and mid-range GPUs, like the RTX 3090 and the 5060 Ti. Experts estimate around a 60% increase in performance demand in AI applications over the next few years, pushing consumers to seek more capable hardware. This suggests that while the RTX 3090 remains a strong choice now, the rate of advancement in GPU architecture may mean that new models could surpass its capabilities sooner than expected. Consequently, buyers who invest in the 3090 may find their decision vindicated in the short term, while those looking for long-term solutions could benefit from waiting for future releases featuring advanced technologies like DDR6 memory or improved cooling systems.
Consider the VHS and Betamax rivalry from the 1970s and 80s. Initially, Betamax boasted superior video quality, much like the RTX 3090's specs. However, VHS won the battle largely due to its longer recording time and cost-effectiveness. This situation draws a compelling parallel to today's GPU debate. While the RTX 3090 may currently outperform the 5060 Ti in specific niches, upcoming models may shift the focus towards features and capabilities that prioritize broader accessibility and future-proofing over sheer power. History shows that sometimes, the battle is not won by the product with the best specs, but rather by the one that understands and adapts to consumer needs.