AMD Navi 24 core photo first exposed
AMD expects to release the entry-level new card RX 6500 XT on January 19, and a lower-end RX 6400 for the OEM market, both of which will feature the new Navi 24 core, which will also be the first gaming GPU to use the 6nm process.
Now, the core rendering of the Navi 24 has been exposed for the first time. At this point, all four members of the RDNA2 architecture family have appeared. The largest, Navi 21 and the smallest, Navi 24, are both vertically cored, while the Navi 22 and Navi 23 in the middle are all tilted by 45 degrees.
The core area of the Navi 24 is estimated to be only 141 square millimeters, slightly smaller than the previous generation of 7nm Navi 14 158 square millimeters, a full 40% smaller than the Navi 23, and only a quarter of the navi 21 core.
The Navi 24 core is basically cut in half in size by the Navi 23, with only 16 compute units, 1024 stream processors, 16MB of unlimited cache, and 64-bit 4GB of GDDR6 video memory.
The second generation of 5G chips has been mass-produced
On December 27th, ZhanRui held an online conference with the theme of "People's 5G", announcing that Zhanrui's second-generation 5G chip platform Tanggula T770 and Tanggula T760 have achieved mass production of customer products, and have reached the highest industry standard of 500ppm in terms of mass production quality. This marks that ZhanRui has fully equipped the research and development and commercial capabilities of advanced process chips.
Tanggula T770 integrates eight core CPU, including an A76 large core, three A76 medium cores, four A55 small cores, integrated Mali-G57 GPU, AI hash rate up to 4.8TOPS, support 4K60fps video encoding, 4K60fps 10-bit video decoding, up to 108 million pixel cameras, quad-core ISPs, support LPDDR4X memory.
Tanggula T760 is a quad-core A76 plus quad-core A55 CPU, Mali-G57 GPU, AI hash rate up to 2.4TOPS, support 4K30fps video encoding, 4K30fps 10-bit video decoding, up to 64mp camera.
Zhan Rui stressed that Tanggula T770 and Tanggula T760 are the world's first successful chip platforms for chipback, which have increased performance by more than 100% compared with the previous generation, increased the integration degree by more than 100%, reduced the overall power consumption by about 37%, increased the endurance by 30%, and supported cutting-edge technologies such as 5G R16 and 5G network slicing, and the 5G rate was increased by more than 30%.
At the same time, 6nm will be the beginning of Zhanrui in advanced processes, and Zhanrui has also had enough technical accumulation to lay the foundation for the next generation of products to cut into more advanced processes such as 5nm.
It is reported that the second-generation 5G chip platform of Zhan Rui has a complete 5G main platform set, as well as an optional 5G RF front-end chip, etc., with a total of more than 10 chips, each of which has reached the mass production standard. Among them, the main platform set includes 7 chips such as main chips, transceiver chips, power management chips, and interconnect chips, and the 5G RF front-end chips contain multiple chips such as PA.
Next, Zhanrui's 5G products will enter the fast lane of development, Tanggula 6, 7, 8, 9 and other series will go hand in hand, and at least one new product will be launched every year.
Intel 2nd Generation Iris Graphics or Mid 2023 Release
Intel Arc, as Intel's new high-performance gaming graphics brand, the first generation of products is Alchemy Graphics (DG2), based on the new Xe Core (Xe Core), manufactured using TSMC N6 process, will support hardware-based ray tracing and artificial intelligence-driven super sampling (XeSS), providing full support for DirectX 12 Ultimate. The first Razak discrete graphics cards are said to be mobile-ready, launched with Alder Lake Mobile in the first quarter of 2022, and not expanded to desktops until the second quarter of 2022.
According to Intel's announced development plans, Alchemist is followed by Battlemage, Celestial, and Drid, with four generations. Unlike the previous three generations, the fourth generation of Druid will adopt a new Xe architecture to replace the original Xe-HPG architecture. If the Xe-HPG architecture used in Alchemy graphics is considered Intel's RDNA architecture, then the Xe2-HPG architecture used by Battlemage is equivalent to the RDNA 2 architecture.
Recently, some foreign netizens said that they got news about the Battlemage (DG3) graphics card codenamed "Elasti". The Battlemage graphics card is said to be aimed at the high-end GPU market, and the release time will be about mid-2023, and will compete with NVIDIA's GeForce RTX 40 series based on Ada Lovelace architecture and AMD's Radeon RX 7000 series based on RDNA 3 architecture.
Intel is adopting a modular design on Meteor Lake, and the CPU will have at least three different modules, namely the compute module, the SOC-LP module (responsible for I/O), and the GPU module. These modules can be stacked with modules from different process nodes, and the package may contain the first modules manufactured at other fabs (TSMC) and interconnected using EMIB technology. It is understood that Intel intends to adopt a similar idea on Battlemage graphics, and is currently working on a design method called "Tiled-GPU" on the DG3 series.
Motorola Edge 30 Pro exposure
According to the latest reports, Motorola has a model of Moto Edge 30 Pro that appears in the Geekbench database, and from the configuration point of view, the machine is still equipped with the Snapdragon 8 platform, equipped with 12GB of memory and 256GB of storage space.
From the naming of the product model, the edge 30 Pro should be higher than the previously released edge X30 positioning, belonging to the top flagship series, in the image, fast charging and other aspects have been strengthened.
In fact, the current edge X30 specifications have been relatively comprehensive, using a 6.8-inch full screen, resolution of FHD+, refresh rate of 144Hz, front 60 million pixels, rear 50 million main camera, 50 million ultra-wide angle, 2 million vice camera, battery capacity of 5000mAh, support 68W fast charging.
Edge 30 Pro as a super cup, may be upgraded to a 2K screen, the image upgrade to the rear three main camera system, fast charging may also break through the 100W, at least equipped with 120W specifications, in terms of hardware impact on the first echelon, but the price may bring surprises.
Toshiba develops HDD expansion technology
HDD hard disk compared to SSD hard disk the main advantage is only the unit price under the larger capacity, at present, Western Digital, Seagate has begun to ship 18TB or even 20TB hard disk, Toshiba's 18TB hard disk has also begun to go on the market, they have also developed HDD expansion technology, is expected to manufacture 30TB hard disk.
According to reports, Toshiba has developed a technology that can greatly increase the storage capacity of digital device storage devices "mechanical hard disk (HDD)" and will be used as the main data center product, with the goal of developing a product with a storage capacity of 1.7 times the current storage capacity of more than 30TB.
Toshiba has verified the technology of using microwaves to improve storage capacity, mechanical hard disks write data by changing the direction of the magnets on the disk, and by irradiating the disks with microwaves, the magnets can be easily reversed, thereby improving the storage capacity, which can save more data in the same unit area.
Toshiba began trial supplying of products with a maximum storage capacity of 18TB in March, mass production began in October, and in recent years, it has gradually expanded its storage capacity in 2TB units, and will continue to promote large-capacity development in the future.
Toshiba believes that it is possible to develop products with a capacity of more than 30 TB using new technologies, so it will be practical as soon as possible.
Exposure Samsung Exynos 2200 CPU performance increased by only 5%
According to SamMobile, Samsung is about to release a new generation of high-end Soc Exynos 2200, and like the Qualcomm Snapdragon 8 platform, the Exynos 2200 chip is also based on Samsung's 4nm process. The report pointed out that the CPU performance of the Samsung Exynos 2200 is only 5% higher than that of the previous generation, and the GPU performance is increased by 17%.
SamMobile was slightly disappointed with the upgrade of the Samsung Exynos 2200, as the previous generation of Exynos 2100 chips, both in CPU and GPU performance, has a significant improvement over the Exynos 990.
Specifically, compared to the Exynos 990, the Exynos 2100 has 30% better CPU performance and 40% more GPU performance.
Compared with the upcoming Exynos 2200, the cpu and GPU performance of the former is limited, and it seems that the improvement is not much.
It is worth noting that the Exynos 2200 GPU is based on AMD's RDNA 2 architecture, and the operating frequency may be 1250MHz. Judging from the exposed news, the Exynos 2200 GPU is not as powerful as the Snapdragon 8's integrated Adreno 730 GPU, which is more powerful.
It is reported that the Samsung Exynos 2200 will be mass-produced and commercially available in February next year, and will be launched worldwide by the Galaxy S22 series, and the National Bank Edition is expected to be equipped with the Snapdragon 8 platform.