Rumors that Nvidia is preparing the GPU GA103 of the Ampere line have been circulating for a long time, and now there are new details about it. A user of the largest American forum, registered under the nickname Faintsnow, is sure that the new GA103 appears in the drivers under the code name GN20-E8 with the ID 2420, and it will be used in the GeForce RTX 3080 Ti for laptops.
Another insider believes that the GA103 will serve as the basis for the RTX 3090M (M here is from the word Mobile, that is, also a laptop). Finally, the defunct GeForce RTX 3080 Super was previously spotted in an unanimated laptop configuration. In other words, GA103 has three possible designations, and it’s not clear which Nvidia will settle on. But it is clear that this will be the company’s most powerful 3D accelerator for notebooks.
According to preliminary data, GA103 will receive exactly the same memory bus as mobile GeForce RTX 3070 and 3080 – 256-bit. The GDDR6 memory will be 16 GB, and the CUDA cores will be noticeably more than the RTX 3080: 7680 versus 6144.
The premiere of the graphics card on GA103, whatever it is called, is logical to expect in January at CES 2022. As part of the event, Intel will present Core 12 (Alder Lake-P) processors for laptops, and Ryzen 6000 mobile APUs (Rembrandt) should also debut. The new 3D card is intended for laptops on these CPUs and APUs.
The cooling system for Palit GeForce RTX 2060 Dual series graphics cards with 12 GB of memory includes two 90 mm fans
Palit Microsystems, which calls itself the largest partner of Nvidia in the production of video cards, introduced the GeForce RTX 2060 Dual and GeForce RTX 2060 Dual OC models with 12 GB of memory. As you know, these are improved versions of the well-known model built on the Nvidia Turing architecture. They differ from the original version in the increased number of CUDA cores and doubled the amount of memory. The cards support Nvidia DLSS, Nvidia Reflex, real-time ray tracing and many other modern technologies. Contrary to the letters of the OS in the name, usually indicating factory overclocking, in both cases the base frequency is 1470 MHz, the increased one is 1650 MHz. GDDR6 memory communicates with the GPU over a 192-bit bus, transferring data on each line at 14 Gbps.
The cooling system design for the new graphics cards includes two 90mm fans. The mounting plate has a Dual-Link DVI-D, HDMI 2.0b and DP1.4a video outputs connector. The card measures 235 x 115 x 40 mm and occupies two slots in the PC case. It consumes 184 watts. It has one 8-pin PCIe auxiliary power connector.
Zotac introduced the GeForce RTX 2060 graphics card with 12 GB of memory
Like Nvidia’s other graphics card partners, Zotac has unveiled the GeForce RTX 2060 with 12GB of memory. According to the source, Nvidia has “called from the stock” the GeForce RTX 2060 model, increasing the number of CUDA cores and doubling the memory from the previous 6 to 12 GB to meet AMD’s recently unveiled Radeon RX 6600.
The Turing graphics architecture at the heart of the GeForce RTX 2060 is still modern as it offers full DirectX 12 Ultimate support. The GPU configuration includes 2,176 CUDA cores, 34 RT cores, and 272 Tensor cores. The GDDR6 memory is connected to a 192-bit bus.
As for the Zotac board, this is a fairly simple and compact model with a cooling system that includes an aluminum heatsink and a couple of fans. The card is powered by a single 8-pin PCIe power connector. No pricing information yet.
There is confusion in the ranks of the new Intel CPUs: the processor with fewer cores is older than the model with more cores. It’s all about the cores themselves
The heterogeneous structure of Intel processors, when there are fast large and small energy efficient cores, in theory can lead to situations where a processor with fewer cores will be better than a model with a larger one. Judging by the latest data, this situation will take place in Intel’s lineup.
The specifications of the mobile CPU Core i5-12650H show that it has fewer cores than the younger Core i5-12500H. But at the same time, the Core i5-12650H received its number for a reason.
Yes, it has 10 cores versus 12 cores for the younger model. But the ratio is important. While the Core i5-12500H has 4 large cores and 8 small cores, the Core i5-12650H has 6 large and 4 small cores. And for most, if not all, this configuration will be faster.
Judging by what we now know about future Alder Lake, there will be no such confusion in the desktop segment, but still we need to wait for the official announcement of all these processors, which will take place early next year.
Elon Musk told how humanoid robots Tesla Bot will help us
Earlier this year, Elon Musk announced plans to create a humanoid robot Tesla Bot, equipped with artificial intelligence, designed to...
The most productive low-cost Android smartphones around the world. There are two well-deserved leaders and one new contender for the throne in the AnTuTu ranking
The AnTuTu benchmark team has published a ranking of mid-range Android smartphones for November 2021 for the international market. It...
“Hopefully we see better performance out of the box,” PCMag compares Snapdragon 8 Gen 1 to Apple A14 Bionic and Apple A15 Bionic
PCMag had the opportunity to test the new Snapdragon 8 Gen 1 SoC and compare it with the Apple A14...
The most productive Android flagships around the world. A promising newcomer burst into the Top 3 of the Antutu ranking
The team of the popular AnTuTu benchmark has published a fresh ranking of the most productive smartphones based on the...
Security6 days ago
Ex-Ubiquiti employee posing as a hacker may be imprisoned for 20 years
News4 days ago
Dramatic Bitcoin collapse: the rate fell from 57 to 45 thousand dollars. Bitcoin pulled other cryptocurrencies with it
Phones5 days ago
33 million Samsung Galaxy S22, 267 million Galaxy A23 and Galaxy A33, and 92 million Galaxy A53 and Galaxy A73. Samsung has set plans for the year
Phones7 days ago
The world’s first Snapdragon 8 Gen 1 smartphone will go on sale on December 15th. A video from his production has been published