Video Cards & GPUs News - Page 10
Arm has just announced its next-gen flagship GPU with a really interesting name that I like the sound of more and more as I see it: Immortalis.
The new Arm Immortalis-G715 GPU is the successor to Arm's super-successful Mali GPU -- which has shipped 8 billion GPUs to be precise -- packing hardware-based ray tracing (RT), variable rate shading (VRS), and more. Arm says that its new Immortalis-G715 GPU will support ray tracing and variable rate shading across all of its Immortalis GPUs.
Arm's new Immortalis-G715 GPU is the first Arm GPU to feature hardware-based ray tracing support, designed specifically for mobile. We have ray tracing on the PC and consoles and have had it for a while now, but Arm is wanting to get ray tracing support into smartphones and now it's on its way.
AYANEO has just teased its next-gen gaming handheld, which is also the world's first gaming handheld system that will include either Intel or AMD discrete graphics cards.
The new AYANEO NEXT 2 will include an Intel CPU and Intel GPU, or an AMD CPU and AMD GPU together, for the first time in history. AYANEO CEO Arther Zhang confirmed the news himself, with the AYANEO NEXT 2 being the first to use a non-integrated GPU -- a discrete GPU -- inside of one of its designs, giving gamers the choice between Intel and AMD.
AYANEO will be using an Intel Alder Lake CPU with an Intel DG2 Mobile GPU for its Intel-powered AYANEO NEXT 2 handheld gaming system, while the AMD-powered version will include an AMD "6000 series" CPU and an AMD 6000 series mobile GPU.
A fresh new rumor on AMD's next-gen RDNA 3 GPU architecture has floated up into the rumor mill, with AMD software engineers pushing out new patches for the new GFX11 architecture -- otherwise known as RDNA 3.
The new patch teases AMD is working on its own instructions that can operate on matrixes, with support for advanced artificial intelligence algorithms, to super-power their super-resolution technologies on RDNA 3-based graphics cards, spotted by @0x22h on Twitter. Effectively, this would be like Tensor cores on an NVIDIA GPU, hardware-based assistance for DLSS.
AMDGPU is a backend for AMD GPUs for LLVM compiler library, which gets updated by AMD employees, and if you're keen-eyed enough you can get a sneak into what next-gen GPU architectures will bring. In this case, the Wave Matrix Multiply-Accumulate was added to the new GFX11 architecture, GFX11 being the codename of AMD's next-gen RDNA 3-based Radeon RX 7000 series GPUs.
EVGA's new PowerLink 41s has been announced, acting as a next-gen cable management accessory for your EVGA GeForce RTX 3090 Ti FTW3 graphics card.
The new EVGA PowerLink 41s provides maximum stability, improved case airflow, and of course it glows with fully-customizable ARGB LED lighting. EVGA has designed its PowerLink 41s exclusively for its GeForce RTX 3090 Ti FTW3 series of GPUs, which rock the new 16-pin PCIe power connector... this is where PowerLink 41s helps.
PowerLink 41s connects to the end of your EVGA GeForce RTX 3090 Ti FTW3 graphics card, and turns the single 16-pin PCIe power connector into 4 x 8-pin PCIe power connectors, which then get connected individually to your power supply. You'll also get some beautiful RGB lighting on the back of it, so the EVGA PowerLink 41s connected to an EVGA GeForce RTX 3090 Ti FTW3 will look beautiful... cleaner, and has RGB lighting to boot.
The closer we get to AMD's next-gen RDNA 3 GPU launch, the more details that are leaking out: today, we have news that the rearchitected RDNA 3 GPU Compute Units will have "enhanced ray tracing" capabilities, and even faster GPU clocks.
AMD itself has confirmed the news of the boosted ray tracing capabilities and higher GPU clocks during its Financial Analyst Day 2022. AMD's SVP of Engineering at Radeon Technologies Group, David Wang, explained that the company has architectured the compute units inside of RDNA 3 that will increase ray tracing capabilities.
We don't know what these "enhanced ray tracing capabilities" are going to do, but it should be obvious that ray tracing performance and upgraded (and new) features inside of the next-gen AMD RDNA 3-based Radeon RX 7000 series GPUs will be plenty. AMD will better compete against NVIDIA in the ray tracing department, as NVIDIA is the clear dominating GPU in ray tracing, but RDNA 3 won't be messing around.
We only just heard about Intel's first Arc workstation GPU -- the Arc Pro A40 -- just yesterday, and now we're already hearing about the Intel Arc Pro A50 workstation GPU.
Intel recently launched its new desktop Arc GPUs in a China exclusive, with benchmarks and more being revealed over the last few days, then news of the Arc Pro GPUs started. First, we heard about the Intel Arc Pro A40 workstation GPU, and now news of the Arc Pro A50 workstation GPU is here.
In new listings on SiSoft Sandra, the Intel Arc Pro A40 and Arc Pro 40 workstation GPUs are found -- with the new Arc Pro A50 workstation featuring the full ACM-G11 GPU with 8 Xe-Cores (128 Execution Units and 1024 FP32 cores). Both the Intel Arc Pro A50 and Arc Pro A40 workstation GPUs seem to be packing 6GB of GDDR6 memory, which should be pumping through a 96-bit memory bus.
AMD isn't too far away now from launching its next-gen RDNA 3-based Radeon RX 7000 series GPUs, and now we're hearing about higher power consumption on the RDNA 3 chips... because NVIDIA is using more power.
The news is coming from Tom's Hardware, which spoke with AMD Senior VP Sam Naffziger, who was one of the minds behind AMD's revolutionary chiplet architecture. AMD has recently said that it would have a 50% projected performance-per-watt uplift with its new RDNA 3 GPUs over RDNA 2, but now those numbers are getting clearer.
Naffziger said that since NVIDIA is cranking up the power consumption on its next-gen Ada Lovelace-powered GeForce RTX 40 series GPUs, that it would be boosting the power consumption on its RDNA 3-based Radeon RX 7000 series GPUs.
NVIDIA's next-gen GeForce RTX 4090 Ti and GeForce RTX 4090 graphics cards -- in Founders Edition form -- have been teased in some new renders. Just remember, this could be completely fake, so I've got a link here to make sure this is safe.
The new renders are courtesy of Moore's Law is Dead, with the alleged NVIDIA GeForce RTX 4090 Ti Founders Edition, and NVIDIA GeForce RTX 4090 Founders Edition both shown off in their BFGPU goodness. NVIDIA will reportedly using a 3-slot, or 3.5-slot cooling design with more fins and more heat sink surface area with its new Ada Lovelace GPUs in the high-end RTX 4090 Ti and RTX 4090 designs.
In a new tweet about his video, Tom @ Moore's Law is Dead explains: "Tonight I share renders of the leaked #NVIDIA RTX 4090 Ti cooler, and break down what there is to learn about AD102 from said leak! 💚🧵 I also provide information about when you can expect decent Lovelace Volume this fall".
AMD is celebrating the first anniversary of the first-gen FidelityFX Super Resolution 1.0 (FSR 1.0) technology, by making its next-gen FSR 2.0 technology open source, and available to download right now.
AMD has now published the source code of FSR 2.0 on GitHub, which should see many more FSR 2.0-capable titles in the future. AMD explains that with the introduction of FSR 2 "our temporal upscaling solution earlier this year, there are now over 110 games that support FSR. The rate of uptake has been very impressive - FSR is AMD's fastest adopted software gaming technology to date".
With over 110 games supporting FSR 1.0, there is now a dedicated FSR 2.0 page that also includes the FSR 2.0 API, the full C++, and HLSL source code behind it, with AMD adding that you'll "also discover our Cauldron-based sample and comprehensive API documentation to help you with your integration".
NVIDIA looks to be making some changes to its upcoming GeForce RTX 40 series GPUs, with leaker "kopite7kimi" revealing "some updates" on the new Ada Lovelace graphics cards on Twitter.
In a new tweet, the leaker has provided some updated specs on NVIDIA's three first Ada GPUs: the second-flagship GeForce RTX 4090, the new GeForce RTX 4080, and the GeForce RTX 4070. Let's tackle the elephant in the room: the GeForce RTX 4090.
The updated specs for NVIDIA's flagship GeForce RTX 4090 see its CUDA core configuration tweaked, with kopite7kimi noting that the AD102-300 GPU will have 16384 CUDA cores (up from 16128 CUDA cores). The SMs (Streaming Multiprocessors) have changed slightly, with the leaker saying NVIDIA is using 128 SMs on the GeForce RTX 4090 (up from 126).