The news follows and does not quite look the same on the NVIDIA side with some unpublished information regarding the RTX 3090 Ti.

During the next CES in Las Vegas – early 2022 – NVIDIA should unveil new graphics cards to keep players and investors patient while the next generation of GPUs – codenamed Lovelace – should make us languish for a few more months.

At the top of the list of cards that NVIDIA should launch, there is talk of a new flagship to use the English terminology. In other words, the new flagship of the Ampere range intended to replace nothing less than the GeForce RTX 3090.

Until now, on the naming side, rumors have hesitated between GeForce RTX 3090 Super and GeForce RTX 3090 Ti. Our colleagues from Videocardz seem convinced, however, that it is this second name that is retained by NVIDIA.

The most important is however elsewhere and, in particular, in the new TDP of this card “maousse” which must come to tease the 450 Watts when the previous flagship of NVIDIA was satisfied with 100 W less, at 350 W. A monster is in preperation !

Still according to our colleagues, this increase in TDP is largely linked to changes in the video memory side. The amount – 24 GB – should not change, but there is talk of a bandwidth of 21 Gb / s (against 19.5 Gb / s on the RTX 3090). In addition, NVIDIA would have opted for 2 GB modules – against 1 GB currently – and these same chips should also be found on the RTX 3070 Ti 16 GB.

On the GPU side, we are talking about a GA102-350 engraved using the 8 nm process. It would be roughly well equipped compared to the GA102-300 of the RTX 3090: it would thus be a question of 10,752 CUDA cores on the RTX 3090 Ti against 10,496 on the RTX 3090, of 128 ROPs against 112, 84 RT cores against 82 and of 336 Tensor cores against 328. What leads to the best in the world of graphics cards.

However, as a consequence of these changes on the memory and TDP side: NVIDIA would adopt a new card design and would have set its sights on a single power connector – like its other Founders Edition – but with 16 pins.

Finally, note that for the moment, there has been no question of any pricing for this dreaded card and it is also not known if NVIDIA has in its plans support for PCIe 5.0. We doubt it, however.

Microsoft has agreed to launch a study on the environmental impact of better repairability of its devices and to act on its results by the end of 2022.

A recognized player in the “memory” market in the very broad sense of the term, ADATA manufactures DRAM modules as well as products with NAND Flash, mobile accessories, gaming products and much more.

DDR5 memory
is about to land at our dealers: Corsair is simply one of the first to draw.

If Intel remains relatively spared by the global shortage of semiconductors, it is because the group manufactures its processors.
in its own factories. And precisely, he still plans to build a new one in Europe.

AMD will soon complete its graphics card offering
RNDA 2 from below, with a Radeon RX 6600 expected on October 13. YouTube channel author Moore’s Law is Dead says launch pushes NVIDIA to unsheathe RTX 2060 with 12GB of VRAM

Whether single-core or multi-core, the Alder Lake Core i7-12700K atomizes the Core i9-11900K and even comes to tease the Ryzen 9 5900X
in multicore.

AMD announced that its processors
Windows 11 compatible
had performance issues on the new version of the operating system.

Pat Gelsinger, P.-D.G. Intel, do not go with the back of a spoon to extol the merits of its new CPUs
general public and servers.

Are you looking for an EVGA RTX 3000 and haven’t had the opportunity to buy one yet? Know that the brand wants to put you at the head of its queue. The manufacturer has announced that it will now prioritize orders placed by customers who have not yet purchased.

In an interview with the American site Cnet, Michael Dell, P.-D.G. and founder of Dell, returned to his relationship with Steve Jobs. We learn that at the end of the 90s, the boss of Apple was considering an agreement with Dell to preinstall macOS
on its competitor’s PCs.