*Updated on 24th June, 2025

The NVIDIA GeForce RTX 5050 marks the 7th entry in the Blackwell GeForce RTX 50 lineup and comes next to the GeForce RTX 5060. The GPU has been in the leaks for the past several months, with more specs surfacing in recent days.

Although NVIDIA hasn't officially said anything about the RTX 5050 yet, it has been spotted in multiple laptop listings and has today unveiled it silently. The GPU exists for both the desktop and laptop platforms, and now we have the official specifications as of 24th June. As far as the performance is concerned, there have been a couple of reports indicating its true prowess, which we will be explaining in a dedicated section.

Here is all we know about the GeForce RTX 5050 till now.

GeForce RTX 5050: Smooth 1080p Performance on a Budget

The NVIDIA GeForce RTX 5050 is a 50-class card that succeeds the RTX 4050. While the RTX 4050 wasn't released for the desktop, it exists in the form of a laptop SKU. The last time NVIDIA launched a 50-class card for the desktop was in 2022. It was the GeForce RTX 3050, based on the Ampere architecture.

Before that, NVIDIA launched the GTX 1650 and the GTX 1650 Super in the Turing lineup. After skipping the 50-class card in the Ada series, NVIDIA is expected to release the GeForce RTX 5050 for the entry-level segment. It is expected to replace the RTX 3050 on the desktop and RTX 4050 on the laptop, positioning itself as a new budget GPU for 1080p gaming.

Specifications

The speculated specifications of the GeForce RTX 5050 include the GB207 die (GB207-300-A1), which has been reported multiple times. The RTX 4050 was based on the AD107 die, and naturally, the RTX 5050 is expected to follow the same suit. NVIDIA usually uses the XX6-class die for the 60-class cards, as we know from the recent generations. The only exception was the RTX 3050, which utilized the cut-down AD106 GPU.

The GeForce RTX 5050 is expected to feature 2560 CUDA cores, which is the same count as found on the RTX 3050 and RTX 4050, except that the architecture is now more refined and power-efficient. The RTX 5050 is reportedly getting the PG152-SKU50 board and will feature 8 GB of VRAM. There have been multiple conflicting reports on the memory type, but today it was confirmed. Some suggested GDDR7 while some suggested older GDDR6, but as NVIDIA launched both laptop and desktop variants today, it confirmed that both variants will have different memory types.

For the laptop variant, NVIDIA went with the GDDR7 memory , which will run at 28 Gbps compared to 20 Gbps on the desktop variant. Specs-wise, there isn't any difference between the desktop RTX 5050 and its laptop variant as far as we have seen. Still, it's usual to see a slower SKU for laptops, but the difference won't be as large as you see for other RTX 50 series GPUS.

Theoretically, if the RTX 5050 got GDDR7 memory, the GPU would be naturally faster with a 28 Gbps memory speed, bringing roughly 448 GB/s of memory bandwidth. Since it is shipped with GDDR6 memory, running at 20 Gbps, which is noticeably higher than the GDDR6 memory found on RTX 40 series cards, the memory bandwidth is 320 GB/s.

It should be noted that NVIDIA hasn't added more VRAM capacity since the RTX 3050, but thankfully, the laptop version is also getting 8 GB VRAM , which is 2 GB higher than the RTX 4050. Hence, the memory bus on both will be 128-bit, as NVIDIA hasn't officially started adopting the 3 GB GDDR7 modules.

Coming to the power requirements, various reports indicated that the GeForce RTX 5050 will be rated at 130W of TDP, but as recently seen on NBD shipping manifest, the GB207 was rumored to feature a TDP of 100W. With the official launch, NVIDIA confirms that the laptop RTX 5050 will have a 115W TDP while the desktop variant will run at a slightly higher 130W TDP rating. It will therefore, ship with a single 8-pin PCIe connector for power requirements and NVIDIA recommends at least a 550W of reliable PSU.

Performance

GeForce RTX 5050 is expected to be closer to the RTX 4060 in performance and should be noticeably better compared to the RTX 4050. Regarding its actual performance, there are only a handful of reports that indicate the performance level of RTX 5050. The laptop RTX 5050 GPU scored 88,727 points in recently, which is 11% higher than the RTX 4050. One should still take these numbers with a grain of salt, as Geekbench scores do not show the whole picture and might not translate similarly in gaming benchmarks.

Even though the RTX 5050 brings fewer cores than the RTX 4060, the architectural improvement and faster VRAM should contribute to its faster performance. Since both GPUs have 8 GB VRAM capacity in common, the RTX 5050 won't be ideal for 1440p resolution. Nonetheless, both the desktop and laptop RTX 5050 SKUs are targeted for 1080p resolution, and the laptop variant should be a nice jump from RTX 4050, which only brings 6 GB VRAM capacity.

There isn't much related to RTX 5050's performance, but in the coming days, expect more leaks to surface. We will update the page once we find more info on the RTX 5050's performance.

Pricing

The official MSRP of GeForce RTX 5050 is $249, which we expected going by the previous launches. At this price tag and performance somewhat close to RTX 4060, the RTX 5050 may be on par with the latter in value for price. However, a price tag of around $200 suits the RTX 5050 more, particularly when the RTX 5050 already has 50% fewer cores than the RTX 5060 and is equipped with slower GDDR6 VRAM.

Release Date & Availability

The GeForce RTX 5050 has been released on 24th June 2025 , for both laptops and desktops. NVIDIA initially planned to launch the card in late July, but has released the laptop variant on the 24th of June. As for the desktop variant, the card was officially unveiled alongside but will see a retail launch in the second half of July 2025.