AMD is expected to release the Radeon RX 6700 series of graphics cards based on the Navi 22 GPU later than previously expected. According to new information provided by Cowcotland, AMD plans to launch the Radeon RX 6700 series by the end of March.

NVIDIA is currently expected to unveil the mobile RTX 3060 6GB and desktop RTX 3060 12GB at CES 2021 next week, but the availability of the desktop variant has yet to be confirmed.

Cowcotland reports that the Radeon RX 6700 series will feature a Navi 22 GPU with support for up to 40 compute units (2560 SP). Thus, the RX 6700 XT is considered to offer half of the flagship stream processors of the RX 6900 XT. It is also reported that the card will have 12 GB of memory, with a 192-bit memory bus. However, the exact specifications of the video cards have not yet been confirmed.

The non-XT RX 6700 is also expected to have 12GB of GDDR6 memory, which could be the reason why NVIDIA is now pushing the RTX 3060 with 12GB of memory despite releasing the 3060GB 8 Ti just a month earlier.

An image of the RX 6700 series reference design was leaked a few months ago, even before the first Navi 21 GPUs were announced. The graphics card is expected to feature a dual-fan design with dual 8-pin power connectors. The graphics card must have a PCIe Gen4 interface and up to 12 GB of GDDR6 memory. The Navi 22 GPU is also expected to debut in a mobile form factor for future gaming laptops.

Add a comment

The site https://videocardz.com/ Information has appeared about the new mobile processor Ryzen 5000U.

AMD is set to announce the Ryzen 5000 series with Zen3 core next week. The company is expected to unveil its U series with a rated thermal design power (TDP) of 15W and its high-end gaming H series with a TDP of 45W. Ryzen 7 5800U belongs to the first group. This processor will be based on the Cezanne APU.

The Ryzen 7 5800U is an 8-core, 16-thread processor with a base clock of 2,0 GHz and a boost clock of up to 4,4 GHz. Compared to the current series, the Zen3-based APU will have twice the L16 cache at 200 MB. The processor should have a XNUMX MHz higher base frequency and increased clock speed compared to its predecessor.

In CPU-Z software, the Ryzen 5000U scored 592 points in the single-threaded test and 3812 points in the multi-threaded workload. The multi-threaded score may not be very impressive, but the single-threaded score of 592 puts this processor even higher than Intel's current mainstream desktop flagship, the Core i9-10900K, which scores 584.

When it comes to multi-threaded evaluation, it is obvious that there is a problem with SMT. Even the Ryzen 7 4800U scored 5k points, so it's safe to assume that either SMT does not work properly, or this particular laptop still has a crude BIOS.

The Ryzen 7 5800U scored 509 points in the Cinebench R20 single-thread test and 3614 points in the multi-thread test. In Cinebench R23, the processor scored 1311 and 9326 points, respectively. 

 

Add a comment

AMD recently added Adobe Premiere Pro results to its official Radeon RX 6900 XT website. The company used PugetBench performance testing software provided by Puget Systems to obtain Adobe Premiere Pro results. This may have prompted the company running the test to test the graphics card with other "content creation" workloads, such as DaVinci software.

In our latest review of DaVinci Resolve Studio, the commercial version of the highly popular video editing software, Puget Systems tested several graphics cards from NVIDIA and AMD released over the past four months. According to the results, AMD still has a lot of work to do to remain competitive in this particular application.

The AMD Radeon RX 6900 XT is the flagship model of the RDNA2 series for $999. This video card is now almost nowhere to be found, but Puget Systems, a company specializing in the production of workstation PCs, managed to find a device for its test. This model, like the rest of the current Radeon RX 6000 stack, should be of interest to content creators for one specific reason - 16GB of memory. This memory is ideal for editing 4K content, but it's probably not enough to handle raw 8K data. For these tests, the professional series is preferred, which usually offers more memory than the gaming series. For this reason, the NVIDIA RTX 3090 graphics card seems better suited for 6K or 8K video editing as it comes with 24GB of high-speed GDDR6X memory.

The testing platform was equipped with an AMD Ryzen Threadripper 3970X (24-core) processor with 64 GB of DDR4-3200 memory.

 

AMD Radeon RX 6900 in DaVinci Resolve (Extended Overall Score)

      

AMD Radeon RX 6900 in DaVinci Resolve (4K and 8K editing)

The overall extended rating is a good indicator of the performance of the graphics cards, but it is likely that they are more related to the CPU. For this reason, it is also necessary to test the performance of the graphics card with GPU filters applied, such as temporal noise reduction or OpenFX effects, which are accelerated by the GPU:

 

AMD Radeon RX 6900 in DaVinci Resolve (GPU effects)

Reviewer's conclusion:

From an overall perspective, the Radeon 6900 XT comfortably trails the significantly less expensive GeForce RTX 3060 Ti by about 10% and trails the cheaper (though still cheaper) RTX 3080 by 23%. And when you look at tasks especially GPU-related, such as noise reduction and OpenFX, the results are even more in NVIDIA's favor, with the GeForce RTX 3080 scoring 56% higher than the Radeon 6900 XT.
- Matt Bach, Puget Systems

The full review includes individual benchmarks. It looks like the RX 6900 XT can beat the RTX 3060 Ti in a few tests. Otherwise, the AMD Radeon RX 6000 can't keep up with the GeForce RTX 30 in DaVinci Resolve. The reviewer doesn't list the drivers used, but it's important to note that NVIDIA has special "Studio Drivers" that come with support for accelerated workloads. GPU this DaVinci.

AMD has yet to release a Radeon Pro series based on Navi 2X GPUs. It remains to be seen whether these graphics cards will be faster at handling content creation workloads like DaVinci Resolve.

Add a comment

Fixed issues

• [GeForce Experience]: FPS counter is enabled and in applications Windows Mail and Calendar overlay appears. [200682565]
• [GeForce GTX 750 Ti]: A blue screen crash may occur (UNEXPECTED KERNEL MODE TRAP). [3196164]
• [GeForce GTX 1080 Ti]: The desktop screen may flicker. [3200599]
• [GeForce GTX 1080 Ti]: The system may freeze and the screen will appear solid color. [3203038]
• [HDR]: When HDR is enabled, black levels are incorrect, especially on LG OLED TVs. [200682795/3201305]
• [Laptop] [Lenovo Y740]: The laptop shows corruption after resuming from sleep mode. [3178279]
• A blue screen crash (0x116) can occur when resuming from sleep mode when three 4K monitors are connected. [3054239/3195884]
  
  

Known Issues

• [X4: Foundations] [Vulkan]: Game may crash on GeForce RTX 30 series GPUs. [3220107]
• [X4: Foundations] [Vulkan]: The HUD in the game is broken. [3169099]
• [Batman Arkham Knight]: Game crashes when turbulent smoke is enabled. [3202250]
• [Detroit: Become Human]: Game crashes randomly. [3203114]
• [Steam VR Game]: There is stuttering and lag when starting the game. [3152190]
• [G-SYNC] [NVIDIA Ampere/Turing Architecture GPU]: GPU power consumption may increase in idle mode on systems using some G-SYNC monitors with higher refresh rates.[200667566]
• [YouTube]: Video playback stops when you scroll down the YouTube page. [3129705]
• [Laptop]: Some Pascal-based laptops with high refresh rate displays may randomly drop to 60Hz while gaming. [3009452]
• [G-Sync] [Vulkan Apps]: Performance degradation occurs when using G-SYNC and switching from full-screen to windowed mode using in-game settings. [200681477]
  To resolve this issue, either run the game directly in windowed mode or disable G-SYNC.
• [GeForce RTX 3070] [Clone Mode]: When the resolution is set to 2560×1440 at 144Hz, the performance status is stuck at maximum performance. [200678414]
• [NVIDIA Turing or newer] [Windows Movies and TV Player]: When playing 4k video in full screen on a 2560×1440 HDR monitor, the video extends beyond the edge of the screen. [3186830]
  
  

Product Compatibility

• GEFORCE RTX 30 SERIES
  GeForce RTX 3090, GeForce RTX 3080, GeForce RTX 3070, GeForce RTX 3060 Ti
• GEFORCE RTX 20 SERIES
  GeForce RTX 2080 Ti, GeForce RTX 2080 SUPER, GeForce RTX 2080, GeForce RTX 2070 SUPER, GeForce RTX 2070, GeForce RTX 2060 SUPER, GeForce RTX 2060
• GEFORCE 16 SERIES
  GeForce GTX 1660 SUPER, GeForce GTX 1650 SUPER, GeForce GTX 1660 Ti, GeForce GTX 1660, GeForce GTX 1650
• GEFORCE 10 SERIES
  GeForce GTX 1080 Ti, GeForce GTX 1080, GeForce GTX 1070 Ti, GeForce GTX 1070, GeForce GTX 1060, GeForce GTX 1050 Ti, GeForce GTX 1050, GeForce GT 1030
• GEFORCE 900 SERIES
  GeForce GTX 980 Ti, GeForce GTX 980, GeForce GTX 970, GeForce GTX 960, GeForce GTX 950
• GEFORCE 700 SERIES
  GeForce GTX 780 Ti, GeForce GTX 780, GeForce GTX 770, GeForce GTX 760, GeForce GTX 760 Ti (OEM), GeForce GTX 750 Ti, GeForce GTX 750, GeForce GTX 745, GeForce GT 740, GeForce GT 730, GeForce GT 720, GeForce GT 710
• GEFORCE 600 SERIES
  GeForce GTX 690, GeForce GTX 680, GeForce GTX 670, GeForce GTX 660 Ti, GeForce GTX 660, GeForce GTX 650 Ti BOOST, GeForce GTX 650 Ti, GeForce GTX 650, GeForce GTX 645, GeForce GT 640, GeForce GT 635, GeForce GT 630

Driver download link

Add a comment

2021 begins with an increase in prices for ASUS PC components.

ASUS has officially announced an immediate price increase for its graphics cards and motherboards. The prices have been adjusted due to increased component, logistics and operations costs. An ASUS marketing manager claims that the manufacturer has worked closely with partners to minimize the increase, but apparently it wasn't enough.

 

ASUS ROG STRIX graphics cards are already some of the most expensive on the market, so it would be disappointing to learn that they could cost even more.

Update regarding MSRP price of ASUS components in 2021.

This update applies to video cards and motherboards*

We have an announcement regarding MSRP price changes that will take effect in early 2021 for our award-winning series of graphics cards and motherboards. Our new MSRP reflects increased component costs. operating costs and logistics activities plus maintaining import tariffs. We have worked closely with our sourcing and logistics partners to minimize price increases. ASUS greatly appreciates your continued business and support as we navigate this time of unprecedented market change.

*additional models may increase as we move into the first quarter.

- Juan Jose Guerrero III, Technical Product Marketing Manager, ASUS

The new prices are already reflected in the lists of retailers. For example, a significant increase can be seen on Newegg, where prices for the water-cooled Radeon RX 6800 XT ROG STRIX LC jumped from $899,99 to $1079,99.

ASUS's own store also sees new increased prices:

• ROG-STRIX-RTX3090-O24G-WHITE: $2109,99
• ROG-STRIX-RTX3090-O24G-GAMING: $1979,99
• RTX3090-24G-EK: $2014,99
• TUF-RTX3090-O24G-GAMING: $1839,99
• ROG-STRIX-RTX3080-O10G-WHITE: $1049,99
• ROG-STRIX-RTX3080-O10G-GAMING: $929,99
• RTX3080-10G-EK: $999,99
• TUF-RTX3080-O10G-GAMING: $859,99
• ROG-STRIX-RTX3070-O8G-WHITE: $779,99
• ROG-STRIX-RTX3070-O8G-GAMING: $699,99
• RTX3070-8G-EK: $729,99
• TUF-RTX3070-O8G-GAMING: $649,99
• KO-RTX3070-O8G-GAMING: $639,99
• DUAL-RTX3070-O8G: $599,99
• ROG-STRIX-RTX3060TI-O8G-GAMING: $599,99
• TUF-RTX3060TI-O8G-GAMING: $529,99
• KO-RTX3060TI-O8G-GAMING: $519,99
• DUAL-RTX3060TI-O8G: $499,99

Add a comment

Lenovo has included new unreleased GeForce RTX graphics cards in the specifications of its Legion desktop gaming system. The specs show that the machine will feature 10th-gen Intel Core processors along with NVIDIA GTX 16, RTX 20, and 30 graphics.

The Lenovo Legion R5 28IMB05 will be offered with up to GeForce RTX 3070 graphics card based on NVIDIA Ampere architecture with GA104 GPU. The desktop system is not currently offered with the RTX 3060 Ti, which is also GPU-based.

As mentioned earlier, NVIDIA recently changed one SKU in its roadmap called RTX 3060 6GB. This model is now known as the RTX 3050 Ti, and Lenovo has such a graphics card in its listing. Both models, however, are based on the GA106 GPU, one of which is expected to have 3840 CUDA cores, while the other is rumored to have 3584 cores instead. The listing seems to confirm that the higher-end model will have 12GB of GDDR6 memory, which was presumably dictated by the upcoming Radeon RX 6700 series, which is also expected to have that amount of memory.

The list is closed by the RTX 3050, the only model based on GPU GA107: This card will have 2304 CUDA cores and up to 4GB of GDDR6 memory, but the memory bus will be limited to 128 bits. NVIDIA is expected to unveil the GA106 GPUs in January, possibly during a special GeForce livestream on January 12.

 

Lenovo Legion R5 28IMB05 GPU options

Add a comment

STALKER 2 received the first in-game teaser
STALKER Shadow of the Chernobyl was released in 2007. After almost 13 years, the company is finally ready to introduce the fourth game in the franchise.

STALKER 2 will follow Call of Prypyat, released in 2009. The company had already announced STALKER 2 back in 2010, but the game was ultimately canceled in 2012. Six years later, the company confirmed that development of the game had resumed.

The game uses Unreal Engine 4, which was revealed by Sergey Galyonkin (who also coincidentally recently appeared in a report on Epic Games Launcher). The game is confirmed to be released for PC and Xbox X and S consoles. The PlayStation version will not be what it seems.

The game will feature next-gen graphics, but support for technologies like ray tracing has yet to be confirmed. STALKER 2 doesn't have a release date yet, but it has already been confirmed that the game will be released next year.

"Get your first look at the first-person sequel STALKER 2 and its in-engine Zone through the eyes of protagonist Skif. According to developer GSC Game World, this teaser represents "what the game actually feels like: rapid scene changes, ominous landscapes, and a constant sense of imminent danger, accompanied by a subtle guitar soundtrack." STALKER 2 is coming to Xbox Series X | S in 2021 (Xbox Day XNUMX Game Pass) and PC".

Add a comment

Yesterday, the first laptop manufacturer put its gaming devices with the GeForce RTX 30 Mobile series up for pre-order. The company has since removed the listing, but some information is known.

NVIDIA will be introducing three graphics cards based on the 8nm Ampere architecture. The flagship RTX 3080 Mobile SKU is expected to feature a full GA104 GPU with 6144 CUDA cores enabled, according to Notebookcheck, which recently updated its GPU database. The RTX 3080 will feature a GA104-775 GPU with a clock speed of up to 1,7 GHz. The clock speed will vary depending on the variant, mainly due to the different specifications of the Max-P and Max-Q models. This GPU is expected to feature up to 16 GB of GDDR6 memory, which will be the first mobile gaming GPU to feature this capacity. However, variants with 8 GB of memory are also expected.

RTX 3070 Mobile will feature 5120 CUDA cores and GA104-770 GPUThis model will run at up to 1,62 MHz clock speeds and offer up to 8 GB of GDDR6 memory. Unfortunately, the memory clock speed has not yet been confirmed, but there has been no information so far to suggest that the mobile Ampere will have faster memory than Turing (up to 14 Gbps).

The NVIDIA GeForce RTX 3060 Mobile will actually have fewer streaming multiprocessors than its Turing predecessor. The card is supposedly equipped with 3072 CUDA cores, which means it has 24 SMs rather than the 30 included in the RTX 3060. Otherwise, the card is expected to have 6GB of GDDR6 memory and a 192-bit memory bus, just like the Turing model.

 

The GeForce RTX 30 Mobile is now scheduled to launch on January 12th. NVIDIA has confirmed that it will be hosting a "special GeForce broadcast" where new mobile SKUs are expected to debut. The new SKUs will expand the list of RTX mobile series to 8 models (not counting the countless and confusing Max-P and Max-Q variants).

Add a comment

Gigabyte NVIDIA is preparing a large list of next-gen graphics cards based on yet-to-be-announced NVIDIA SKUs. The manufacturer has confirmed that it will release a full lineup of GeForce RTX 3080 Ti models, including AORUS Xtreme, AORUS Master, Vision, Gaming, Eagle, and Turbo. This appears to be a complete refresh of the RTX 3080 model lineup without the Ti.

A few non-Ti RTX 3060 models have also been added to the list, though there appear to be far fewer cards on the list compared to the 3080 Ti. For now, the company will be releasing the Gaming, Eagle, and Vision series with and without factory overclocks. In fact, it would be surprising if Gigabyte offered more options such as AORUS cards as this model is expected to be a mid-range card despite having more memory than the 3060 Ti.

 

Listing Gigabyte EEC with RTX 3080 Ti, RTX 3060

 

GeForce RTX 3080 Ti

The NVIDIA GeForce RTX 3080 Ti is a new high-performance model that will fill the gap between the GA102-200 and GA102-300 based models. This card is expected to have the same number of cores as the RTX 3090 and twice the memory of the RTX 3080. With 20GB of GDDR6X VRAM, the card should serve as a good alternative to the Radeon RX 6900 XT (which is still impossible to buy) that ships with a full-featured Navi 21 GPU, 5120 stream processor and 16GB of memory.

With double the CUDA cores per streaming multiprocessor, the RTX 3080 Ti is expected to have 10496 CUDA cores. This new SKU is rumored to cost up to $999 for the Founders Edition, and likely much more for non-standard variants like the Gigabyte. The launch date has not yet been confirmed, but the map was originally scheduled for January (hence, Gigabyte and other AIB brands have started leaking information about the new part number), but the 3080 Ti may be delayed until after the Chinese New Year.

Add a comment

The NVIDIA AD102 (Lovelace) GPU can offer up to 18432 CUDA cores.

A week ago it was known that NVIDIA would introduce a new graphics architecture named after Ada Lovelace. The series is expected to be NVIDIA's next gaming architecture, which could arrive even before Hopper.

The architecture, codenamed Lovelace, will see the number of graphics processing clusters increase to 12, which is double that of the Turing GA102 GPU, according to a new tweet from a well-known and trusted leaker @ kopite7kimi. This means that the GPU can contain up to 72 texture processor clusters (TPC) and 144 streaming multiprocessors. With 144 SMs, the GPU can see up to 18432 CUDA cores (144 x 128), which is 71% more than GPU GA102.

It is unclear when the new architecture will be released. Kopite7kimi only started sharing information about this architecture this month, so it seems to be a fairly recent topic. This may also never happen since roadmaps tend to change quite frequently at NVIDIA HQ.

The NVIDIA Lovelace AD102 GPU could be coming to the GeForce RTX 40 series, possibly after the RTX 30 SUPER update (which is yet to be confirmed). NVIDIA is now expected to update its lineup (all three months after release) with the RTX 3 Ti and RTX 3080 Ti graphics cards in the next quarter.

Source:

Add a comment