According to two generally reputable leakers, Nvidia has brought forward the release date of its GeForce RTX 5050 graphics card, the little sister of the Blackwell family, which is positioned in the entry-level segment. It will apparently be released on July 1st—a timing that risks sowing discord among the brand's partners.
This information, relayed by VideoCardz, was initially revealed by hongxing2020 and MEGAsizeGPU, two insiders generally well-informed on this type of subject. They were the first to announce that the group's youngest would be released at the very beginning of the month, much earlier than expected.
Availability problems... and no reviews?
This rush is likely to have some rather unfortunate consequences, particularly on two levels. The first is that integrators who are used to working with the brand may have a very difficult time being ready on time. It's a safe bet that the majority of them won't be in a position to offer this card on the day, and availability is likely to be chaotic.
This may have a second unfortunate consequence: since integrators won't be ready in time, they won't have the opportunity to ship cards to reviewers to allow them to dissect them in advance. This means that the first customers will be left in the dark, and they will probably have to pay with very little information on the product's real-life performance.
As a reminder, this isn't the first time the public has found themselves in this situation with the Blackwell generation. Nvidia had already skipped reviews of its RTX 5060 and 5060 Ti models, forcing some specialist media to organize “wild” tests in their hotel rooms on the sidelines of Computex to get some concrete information before the official release. A decidedly worrying approach. Let's hope that Nvidia will make some efforts in terms of transparency in the future...
A (probably) substantial performance gain
This is all the more regrettable since these entry-level models are generally very popular, in a context where the price of these components continues to increase with each generation. Furthermore, as a reminder, Nvidia had decided not to release an RTX 4050 for desktop computers; the only models available were specifically designed for laptops. To find a card in this segment, you have to go back two generations, to the RTX 3050.
As for the spec sheet, the GeForce RTX 5050 should be entitled to 2560 CUDA cores, as many as the RTX 3050 and the portable RTX 4050. But don't panic: the switch to the Blackwell architecture should still bring a significant performance boost.
The big question concerns memory. We know that it will come with 8 GB of VRAM — a modest figure, but not surprising for an entry-level model. On the other hand, several sources suggest that the desktop version will only be entitled to old-generation GDDR6 with a bandwidth of 320 GB/s, unlike its big sisters which have switched to the new GDDR7.
The real impact of this choice is difficult to quantify; it's typically the kind of point that independent reviewers would undoubtedly have liked to verify before the release... but this will probably be impossible for the reasons mentioned above.
In any case, in the end, we can still expect a substantial performance gain on the entry-level model - but we should proceed with caution. See you on July 1st for the first answers.
0 Comments