- Nvidia’s neural texture compression characteristic has become more effective
- He obtained this boost thanks to the new cooperative vector of Microsoft in DXR 1.2
- The tests with a demo showed a reduction of almost 90% of the imprint of the video RAM for the textures, but that will not apply in all the scenarios (by all means)
Do you fear that your NVIDIA graphics card will be undernourished in terms of Video RAM (VRAM) that it has on board? You may know Team Green is working on technology to alleviate these concerns, and it seems that making considerable progress in this regard.
The WCCFTECH reports that the NVIDIA (or NTC) RTX neuronal texture compression has a big boost thanks to the new cooperative vector of Microsoft in DirectX Raytracing 1.2.
Does it look like a charge of technical nonsense? Well, yes, and if you dare to look at the in -depth explanation of what is underway here, you will probably be very lost in a country of confusion – see the post on X below of Osvaldo Pinali Doderlein (which tested the functionality with a graphic pre -pilot).
First, look at NVIDA’s neural texture compression with DXR1.2 COOPERATIVE VECTOR! First, it needs a preview pilot (590.26), I installed that you do not have too much and he corrupted the screen, only after a few hard reintegrated, he decided to work.July 15, 2025
Thus, in short, the Nvidia texture compression deception, which makes it possible to make textures smaller, which means that more can adapt to a modest assistant of VRAM, now works much more effectively thanks to the cooperative vector.
As Doederlein observes, in the demo he has passed as a test, NTC offered an economy of almost 90% in terms of reduction in the VRAM imprint, which is a massive matter. If, and it is a key warning, the game takes care of NTC (the developers must therefore code it, as is the case with Nvidia DLSS, of course).
Since the textures generally occupy half of the VRAM in a given game session (or much more), the advantages should result in a considerable victory for games where to grasp textures from the outside of video memory becomes a chore that slows down everything.
Analysis: the wider image
NTC uses AI (neural networks, hence the name: neuronal texture compression) to intelligently compress graphic textures – in theory without much notable difference from native textures, when they are decompressed to display the player – and, as mentioned, the cooperative vectors of Microsoft in DXR 1.2 Ensure that this process is more effective.
Much more efficient, in fact, and I think Nvidia had an eye on how this type of technology is progressing on the line when Team Green made decisions about the deployment of certain current blackwell graphics cards and to stick to 8 GB of RAM video.
This was a very unpopular decision in certain quarters, so maybe Nvidia counts – and counts strongly – on NTC to prove a major difference in the resistance to the future of these GPUs. (Like the RTX 5060 TI 8 GB and the RTX 5060 Vanilla – The RTX 5050 also has 8 GB, but this memory configuration is more understandable on a budget graphics card).
That said, we must be quite cautious in terms of confidence of certain early results in a single demo as a reason to celebrate – as promising as it may seem. The figure of a (close) reduction to 90% will not be the case in each scenario, you can put yourself in the bank, but even more modest (for example, 50%) will always be excellent news.
The other point to keep in mind is that NTC could have a side effect of loss of performance in the frequency of images – at least in certain scenarios. Hopefully such a frequency of images decreases – or effects on the quality of the image elsewhere – would be light and probably was worth the compromise in terms of easier life in life in VRAM (if not, what would be the point?).
In addition, even if I make more cautious noises here, as already indicated, the textures are not in terms of what is crowded on VRAM in a game session, and wider concerns remain in the future of 8 GB of graphics cards to date.
However, NTC could be a key part of the puzzle for Nvidia, and it looks more like it given these test results. As you can guess, this technology is for Team Green graphics cards, and only the most modern offers for that. (While in theory, NTC can be used by rivals GPUs, such as the Radeon of AMD, NVIDIA in a realistic way recommends that only its RTX 4000 or 5000 graphics cards have the punch and the material necessary to execute it).
So where does that leave AMD and Intel products? These GPU manufacturers have their own neuronal compression technology to a boil, and the right thing is that they will also benefit from Cooperative Vectors DirectX (everyone will). Although at the moment, Nvidia has your head in this department – but we are always a way to see this technology carried out. As indicated, game developers must support technology in their titles – it is not a functionality in the pilot that can be applied in all games. (Well, unless Nvidia plans a watered -down version on a later date – a kind of equivalent of what is the gentle movement in DLSS).
In short, Nvidia has a persuasion to do, once again in the same direction as DLSS, because if the developers do not adopt this, it is a non-starter. However, the sales arguments seem to be seriously convincing.
