Components
Nvidia is preparing a stripped-down GeForce RTX 3050 with 6 GB of memory and a 96-bit bus
While Intel is finally going to release the Arc A580 video card, introduced more than a year ago, which will compete with the GeForce RTX 3050, Nvidia is going to update the RTX 30 line by introducing the cheapest model. This will be the RTX 3050, but with 6 GB of memory.
The RTX 3050 6GB will be powered by the GA107-325 GPU. The core configuration is still unknown: it may either remain unchanged, that is, the new product will receive 2560 CUDA cores, or offer a smaller number of computing units. For example, the mobile RTX 3050 Laptop has 2048 CUDA cores.
It is known that the GPU frequency will be only 1470 MHz, that is, it will be 17% lower than the current RTX 3050. In addition, the bus width will be reduced to 96 bits, and the TGP indicator will be reduced to 70 W, that is, the card will not require additional food.
It is likely that the performance of such an adapter will be significantly lower than that of a full-fledged RTX 3050, especially if the GPU is also cut down.
It is expected that the new model will go into production only in January 2024, that is, it will appear on sale, apparently, by spring. There is no information on the price, but the RTX 3050 in its current form is the cheapest Nvidia video card among the previous generation, and the new one does not even have an analogue.
Most modern solid-state drives have the M.2 2280 form factor, but sometimes you need a much more compact SSD – M.2 2242. Corsair did just that.
In the photo you can see the MP600 Micro model in the specified form factor. The length of such a drive is 42 mm, although there is also an M.2 2230 that is only 30 mm long.
The new product supports PCIe 4.0 and is characterized by data transfer speeds of up to 5100 MB/s for reading and 4300 MB/s for writing. Performance is respectively 600,000 and 890,000 IOPS. These parameters are typical for the 1 TB model, which also has a resource of 600 TBW.
It remains to add that the new product is based on 176-layer Micron 3D TLC NAND memory and is priced at $70.
Components
Huawei defeated American sanctions: the Chinese company now has its own 5-nanometer processor
Five years under strict American sanctions, which cut Huawei off from advanced single-chip systems, have yielded results: the company has its own 5-nanometer processor made in China.
There is an interesting story with this processor. Five days ago, Huawei introduced the Qingyun L540 laptop based on the Kirin 9006C SoC: and at that time, all the characteristics of the device were revealed, but nothing was reported about the central element of the hardware platform. Now Huawei has updated the description on the website, which clearly states that the Kirin 9006C is manufactured using the 5 nm process technology.
The official description says that the single-chip system has eight cores and a maximum frequency of 3.13 GHz. The Gizmochina resource adds: the CPU is made according to a two-cluster design – with four Arm Cortex-A77 cores and four more Arm Cortex-A55 cores.
Let’s remember that earlier Huawei had a 7-nanometer SoC Kirin 9000S, and it is used in the flagships of the Mate 60 line. Kirin 9000S is produced by SMIC, and the same company most likely produces the 5-nanometer Kirin 9006C.
Components
This is the noisiest and hottest RTX 4090, albeit with a unique cooler. Acer has equipped the video card with liquid CO without an external radiator or pump
Acer showed an interesting development earlier this year: a GeForce RTX 4090 video card with liquid cooling, but without an external radiator and pump. It turned out that Acer failed to make such a design effective.
The video card is not available at retail, but is available as part of the finished Predator Orion-X PC. Judging by KitGuru’s tests, such a PC with such a video card is hardly worth buying.
The Acer cooler turned out to be the most unsuccessful of all that the authors tested on the RTX 4090. Firstly, the card turned out to be the noisiest, and secondly, the hottest. Moreover, if the GPU temperature, although the highest, is not much higher than that of the same Asus RTX TUF, then the memory temperature is immediately much higher than that of any other RTX 4090, and in absolute terms is too high (96 degrees).
As a result, the card operates at relatively low frequencies, is very noisy and suffers from throttling.
Small, but remote. Corsair unveils SSD MP600 Micro in M.2 2242 form factor
Most modern solid-state drives have the M.2 2280 form factor, but sometimes you need a much more compact SSD –...
Huawei defeated American sanctions: the Chinese company now has its own 5-nanometer processor
Five years under strict American sanctions, which cut Huawei off from advanced single-chip systems, have yielded results: the company has...
Intel, where is the gain from new architectures? Core Ultra 9 185H was slower than Core i9-13900H with the same number of cores
Yesterday we took a look at the first tests of the Core Ultra 7 155H, which at first glance are...
Windows 12 will be released in September-October 2024, but its capabilities will be limited on many computers
The next big Windows 11 update is called Hudson Valley. It is so big that it will be released as...
-
Components7 days agoGamer’s Lexus TX presented: with RGB lighting, powerful acoustics and a pair of large monitors in the cabin
-
Phones4 days agoNew Redmi smartphone, cheap. Redmi 13C 5G presented
-
News4 days agoPolaris Slingshot 2024 unveiled
-
News7 days agoXiaomi 10G Router and Xiaomi Router BE7000 began to receive the update to Wi-Fi 7