[ad_1]
There will be four players on the market soon, the competition will be interesting.
Regarding Intel, we have known for a long time, practically since mid-2018, that they are preparing for a great return, at least in the independent graphics card market that they left more than 20 years ago. However, this company will not be the only new participant, so next year promises to be very interesting.
Intel Graphics on the first day of May is a short entry It illustrates that despite the restrictions imposed by the new corona virus, serious work continues, which obviously also applies to the development of Intel Xe Graphics. In this regard, it is important to note that its latest solution was the i740, which debuted in 1998 and has since tried to strive for better positions with its more integrated versions, which promised much more modest performance and functionality as part of desktop and mobile chips, although sometimes it was also enough to run certain games. Still, discrete (standalone) copies are really interesting to the masses, and here there seems to be this third player in the battle between AMD and nVidia, while Huawei would become the fourth with the launch of GPU for data centers. competitor, which can be followed by user developments at a later date.
AMD will release RDNA 2 and nVidia RTX 3080 this year, while for Intel, the new copy will actually be the twelfth generation, if we include the integrated solutions introduced over the past decade. The exact technical details and hardware specifications are still not completely clear, but we know that the manufacturer will produce 3 different versions, which will include LP (low power), HP (high performance) and HPC, which is already specifically for work in centers of data. The LP will appear on desktop and mobile chips, as one of these (experimental) copies is found in members of the Tiger Lake processor family to be released soon, which will be the eleventh generation in the Core line. The big question is how many processing units will have space on the chip, because the performance depends on it, while it does not matter how much space these components can be grouped. It is expected that a bandwidth of 10 nanometers will be used, the second generation of the chip will appear at 7 nanometers, but this is obviously even further.
In terms of the number of cores, there would be an upper limit of 4096, while end users could rely on up to 1024 such drives. The clock would be between 1.5 and 2 GHz, they would have between 4 and 16 GB of memory, and the commercial price would be between $ 150 and $ 600, where the most expensive and powerful copies would also end up in the data centers. However, there are still many question marks, so we will have to wait a few months to get the final data.
[ad_2]