Graphics Cards - Page 182

CES 2020 - AMD was swinging big at CES 2020 this year between unveiling its new Ryzen 4000 series mobile CPUs, crazy 64-core/128-thread Ryzen Threadripper 3990X processor and the new mid-range Radeon RX 5600 XT. But what about Big Navi, aka Navi 21 and the high-end enthusiast graphics card market?

Thousands of gamers and Radeon fans were thirsty for something -- anything -- to be said about AMD's return to the high-end graphics card market. AMD didn't say anything officially during its CES 2020 keynote, but AMD CEO Dr. Lisu Su sat down for a media Q&A round table and was asked the heavy questions.

AnandTech has it laid all out for your viewing pleasure, with a question from Gorgon Ung from PC World asking: "Do you think that AMD has to have a high-end competitor in the discrete graphics market?". Lisa replied: "[laughs] I know those on Reddit want a high-end Navi! You should expect that we will have a high-end Navi, and that it is important to have it. The discrete graphics market, especially at the high-end, is very important to us. So you should expect that we will have a high-end Navi, although I don't usually comment on unannounced products".

It must be the day to show off the semi-custom APUs for Microsoft's next-gen Xbox Series X console, with Senior Director of Hardware at Xbox, David Prien, teasing some gorgeous Xbox Series X APUs.

Xbox boss Phil Spencer updated his Twitter profile picture in the last 24 hours with a photo of the Xbox Scarlett APU, but the image wasn't perfect. Prien trumped him, throwing up a picture of 7 different Xbox Series X APUs in a tray, and I don't think I've seen anything more beautiful.

The new Xbox Series X console will be powered by a semi-custom chip using Zen 2-based CPU cores and a new RDNA2-based GPU from AMD with hardware ray tracing abilities. The new Xbox Scarlett APU will be made on the 7nm node and will power the next-gen Xbox Series X console when it launches in Q4 2020.

CES 2020 - AMD expanded on its FreeSync display technology at CES 2020, adding a new tier system to the range that sees the introduction of FreeSync Premium and FreeSync Premium Pro.

The new FreeSync Premium and FreeSync Premium Pro displays will be two new tiers above the regular FreeSync tier, which will be a new baseline for laptops and desktop monitors. What AMD will be adding with FreeSync Premium is that it will be a minimum 1080p resolution, 120Hz refresh rate, and low framerate compensation (LFC).

AMD's flagship FreeSync Premium Pro monitors will have the same minimum 1080p resolution, 120Hz refresh rate and LFC -- but will add in HDR technology. FreeSync Premium Pro is what AMD used to call FreeSync 2 HDR -- but a new dab of paint has been applied. FreeSync Premium is the new entry here, while AMD is just getting their FreeSync branding into a better firing line starting 2020 off with a bang.

CES 2020 - Intel teased its DG1 discrete graphics card at CES 2020, running Destiny 2 during its CES 2020 keynote presentation.

We don't know if it was a real-time demo running at the show, or a pre-recorded clip just yet -- but I'm going with the latter. Intel teased that its DG1 discrete GPU is running, showing off a quick less-than-30-second video of a game running on Intel's new DG1 discrete GPU. Intel showed off the DG1 demo right after talking about its new Tiger Lake architecture, coming later in 2020 on 10nm.

Intel showed off the new DG1 graphics card (I'm guessing in prototype form) running Destiny 2 -- but at an unknown resolution, detail level, or frame rate. Literally nothing to go by at all, just a demo from a mysterious 'DG1'. Did Intel just have another water chiller moment at CES 2020?

CES 2020 - AMD didn't really make its new Radeon RX 5600 XT graphics card a big surprise, but now it's official and the latest member of the Navi graphics card family is here and priced at $279.

AMD's new Radeon RX 5600 XT features a Navi 10 GPU with 2304 cores, 6GB of GDDR6 at 12Gbps on a 192-bit memory bus. The Navi 10 GPU in reference form has its GPU clocked at 1130/1375/1560MHz for base, game, and boost clocks respectively. The GDDR6 will have 288GB/sec bandwidth at its disposal.

The new Radeon RX 5600 XT will require just a single 8-pin PCIe power connector with a 150W TBP (total board power), with AMD tapping its RDNA1 architecture, PCIe 4.0 interface, and made on the fresh 7nm node. There are plenty of custom Radeon RX 5600 XT graphics cards on the way from SAPPHIRE, MSI, XFX, ASUS, and others.

CES 2020 - NVIDIA has unleashed a special new CES 2020 labeled driver, with the new CES 2020 Game Ready Driver including a bunch of new features and adding performance to the latest AAA games.

The new driver includes a slew of new features that see NVIDIA starting off 2020 with a bang, helping out RTX ray tracing and LSS support for Wolfenstein: Youngblood, as well as a new VRSS feature that taps Turing's Variable Rate Shading technology to improve image quality in VR games -- something I'll be testing next week. The full feature list of NVIDIA's new CES 2020 Game Ready Driver:

NVIDIA has added in RTX support for both Wolfenstein: Youngblood and Deliver Us The Moon, with RT Shadows and RT Reflections as well as DLSS dropping in Deliver Us The Moon while Wolfenstein: Youngblood gets RT Reflections and DLSS in the new CES 2020 drivers.

ASRock is coming out swinging with its own leaks on their custom Radeon RX 5600 XT graphics card, that the company will be calling their RX 5600 XT Phantom Gaming D3.

The new ASRock RX 5600 XT Phantom Gaming D3 is a factory-overclocked Radeon RX 5600 XT, offering up the Navi 10 at its fastest GPU clocks yet (before it's even announced) at 1355, 1560, and 1620MHz boost GPU clocks respectively. It still has 6GB of GDDR6 at 12Gbps on a 192-bit memory bus, all powered with a single 8-pin PCIe power connector.

ASRock is using a smaller PCB here with its RX 5600 XT Phantom Gaming D3, at just 29cm -- but the card has a larger cooling system that runs right over the PCB. I'd love to see a custom version of this with the cooler running up to the edge of the PCB for smaller, mITX gaming PCs powered by the ASRock RX 5600 XT Phantom Gaming D3 graphics card.

Today is the day of custom Radeon RX 5600 XT graphics cards leaking, with our first look at the custom MSI Radeon RX 5600 XT MECH (and MECH OC) models but now we're getting a look at XFX's own custom RX 5600 XT.

XFX's upcoming Radeon RX 5600 XT THICC II PRO STAGING has a strange name, and I seriously do not know why it's called this. The original THICC branding was kinda cool as a tongue-in-cheek thing, but PRO STAGING? Yeah, I don't know. The new XFX Radeon RX 5600 XT THICC II PRO STAGING has factory overclocking at up to 1460MHz in game mode, but we don't know anything more about its boost clocks.

We should see the XFX Radeon RX 5600 XT THICC II PRO STAGING coming in faster than MSI's second-best RX 5600 XT MECH OC (the GAMING series will beat the MECH series). But XFX's new card will also include the same 6GB GDDR6 at 12Gbps on a 192-bit memory bus.

AMD is expected to unveil its new Navi 10-based Radeon RX 5600 XT grapphics card at CES 2020 in the next couple of days, but now we have some leaks from MSI's new MECH range of cards.

MSI's upcoming Radeon RX 5600 XT MECH graphics cards have been leaked, with teh MECH series cards finding their place just under the flagship GAMING series. MSI's upcoming cards will come in two SKUs with an OC and non-OC model in the upcoming Radeon RX 5600 XT MECH graphics cards.

The overclocked RX 5600 XT MECH OC graphics card has 1185, 1420, 1600MHz GPU clocks for base, game, and boost mode respectively. The non-OC version of the RX 5600 XT will have its Navi 10 GPU clocked at 1130, 1375, and 1570MHz for base, game, and boost clocks respectively. Both models include 6GB of GDDR6 memory at 12Gbps on a 192-bit memory interface and a single 8-pin PCIe power connector, with a TDP that should find itself at 150W or so.

AMD was the first to get to 7nm with consumer GPUs and CPUs and while we know 5nm is on the way, Bitcoin ASIC design houses in Bitmain and Canaan will be the first with 5nm chips from TSMC to play with.

TSMC gave its freshly-made 5nm products to Bitmain in the last weeks of December 2019, while Canaan will receive some 5nm ASICs in the coming months of Q1 2020. Canaan is well on its way to become the second-largest cryptocurrency mining manufacturer, with the company being the only mainstream mining company that has support from both Samsung and TSMC.

Don't go thinking that 5nm is going mainstream right now, as the current average yield of TSMC's new 5nm process is at around 80% -- a 7-15% increase over the 7nm process. We should expect another year or so to pass before 5nm chips reach volume production, and while Bitmain and Canaan might be enjoying, or to soon be enjoying 5nm, it doesn't mean CPUs and GPUs will be on 5nm this year, or even in 2021.

A juicy new rumor on NVIDIA's next-gen Ampere GPU (which I revealed back in April 2018) has arrived to kick start the New Year for Team Green, with the Taipei Times reporting that NVIDIA's new Ampere GPU architecture could be up to 50% faster, at half the power consumption of Turing.

NVIDIA will want to fight back big time against AMD's upcoming 'Big Navi' that will come in the form of the currently rumored Navi 21, with NVIDIA hitting the new 7nm node just like AMD is currently on with its Navi-based GPU and the Radeon RX 5500 and RX 5700 series graphics cards. But this news of a 50% hike in performance at the same power consumption is big on its own.

Taipei Times is reporting from a client note by Yuanta Securities Investment Consulting Co that NVIDIA's new Ampere GPU would have a 50% increase in graphics performance while "halving power consumption". NVIDIA has been king of efficiency for a while now and that is on the 16nm, 14nm, and 12nm nodes -- but it looks like it has struck lightning on 7nm if these rumors are true. We could well see the '50% faster performance' in the ray tracing (RT) side of things, and not a general 50% performance improvement.

Samsung broke ground on its new EUV fabrication in Hwaseong, South Korea back in early-2018 and was recently hit with a power outage that saw production grinding to a halt.

Seeking Alpha reports that its source made a statement to Bloomberg, where it seems that DRAM, NAND, and system chip production seems to have been affected. We don't know which chips were affected by the production stopping.

The site reports that the pause was due to a "brief, one-minute outage" and that Samsung is "checking on the extent of the damage". Just a single minute can cause major issues, so let's hope nothing more comes from Samsung's troubles at its EUV plant in 2020.

AMD isn't shy about celebrating 2020 starting off with a bang, with the company teasing through its official Twitter that it will be pushing the envelope yet again in 2020 when it comes to high-performance computing.

We don't know what AMD will be unveiling at CES 2020, but it could be a world-first tease of the next-gen Zen 3 architecture, or it could be an even bigger tease of the new RDNA2 architecture that will power next-gen GPUs in 2020 on Radeon and power Microsoft's new Xbox Series X and Sony's new PlayStation 5 consoles.

But it's very specific to the high-performance computing (HPC) market -- so let's hope we're talking about an RDNA2-based tease which would be really huge. AMD stepping back into the high-performance graphics card market is a big deal, as NVIDIA has been alone there for quite a number of years now with its previous-gen Pascal and current-gen Turing GPU offerings.

AMD has been cooking up a kick-ass successor to Navi with a 2020 flagship GPU that is being referred to as Navi 21, with the second-generation RDNA family coming on the new 7nm+ process node.

Navi 21 will be a large GPU with new rumors pointing to it packing somewhere between 15-16 billion transistors, which makes it bigger than Vega 20 (13.2 billion) and Navi 10 (10.3 billion). But the bigger news here is that AMD's new flagship Navi 20-based graphics card would feature 12-16GB of GDDR6 memory on a much wider memory bus.

The new rumors have AMD possibly using a wider 384-bit or 512-bit memory interface, something that would enable much more memory bandwidth with GDDR6 memory. Previous rumors suggested AMD would tap higher-end HBM2E memory, packing between 16-32GB of framebuffer.

ASRock let slip that AMD is about to launch its new Radeon RX 5600 XT graphics card, with a product listing on ASRock's own website that had the full specs for their upcoming Radeon RX 5600 XT 6GB Challenger graphics card.

The full specs on AMD's new Radeon RX 5600 XT graphics card is now here, so we know that it will use a decent Navi GPU with 36 compute units and 2304 stream processors -- the same core count as the higher-end Radeon RX 5700. Where the new RX 5600 XT is different, is in its GDDR6 configuration.

AMD will be offering its new Radeon RX 5600 XT with 6GB of GDDR6 at 12Gbps on a 192-bit memory bus which will result in 288GB/sec of memory bandwidth. This is a big difference to the 14Gbps modules used on the higher-end Radeon RX 5700/XT (as well as more of it with 8GB versus 6GB) which results in 448GB/sec of memory bandwidth.

It looks like AMD Radeon and NVIDIA GeForce graphics cards could get a little more expensive in 2020, with a new report from DigiTimes suggesting graphics card prices will rise because of the current DRAM shortages.

DigiTimes reports that GDDR memory is expected to rise in price sometime during Q1 2020, somewhere over 5% according to DigiTimes' source in DRAMeXchange. DRAMeXchange explains: "Graphics DRAM is more sensitive to demand change than other types of memory products, so its price fluctuations can be dramatic as well. With OEM clients raising their stock-up demand, Graphics DRAM contract prices are projected to increase by over 5% QoQ, the highest among all memory products".

The move by both NVIDIA first with its new Turing-based GeForce RTX graphics cards, and then continued with the Navi-based Radeon RX 5700 series prompted this as both GPU makers moved on to GDDR6 memory. The upgrade from GDDR5 memory to GDDR6 is causing supply issues as less and less GDDR5 is ordered, while GDDR6 memory is in huge demand.

NVIDIA has some fighting words towards its competitor in AMD (and in the process, Microsoft and Sony) during GTC China 2019 with the GPU giant saying its current-gen Turing-based GeForce RTX series graphics cards are more powerful than the next-gen Sony PlayStation 5 and Microsoft Xbox Series X consoles.

During GTC China 2019, NVIDIA founder and CEO Jensen Huang showed off some Max-Q gaming laptops with discrete GPUs powered by GeForce RTX series parts. One of NVIDIA's slides included a comparison between the Lenovo Y900 Gaming laptop which is powered by a previous-gen GeForce GTX 980M to the new Lenovo Y740 Gaming laptop that packs a new GeForce RTX 2080 Max-Q graphics card.

NVIDIA said that the GeForce RTX 2080 Max-Q gaming notebook is superior to the unreleased, and RDNA 2-based Xbox Series X and PlayStation 5 consoles that will drop this time next year. Right now, NVIDIA's GeForce RTX 2080 Max-Q is the best mobile graphics card in a notebook, period.

NVIDIA's next-generation Ampere GPU will be unveiled in 2020, powering a slew of new graphics cards that will span from the GeForce GTX/RTX series through to HPC/datacenter parts. We know Taiwan Semiconductor Manufacturing Company (TSMC) and Samsung are involved, but now we know to what capacity.

During GTC 2019 in China, NVIDIA founder and CEO Jensen Huang addressed questions from the tech press over who will be building out most ogf the new 7nm GPUs in 2020 and beyond. Huang said that most of the 7nm GPU production will be done by TSMC while Samsung will only handling a small portion of 7nm GPU production for NVIDIA.

Huang also said that without TSMC and their advanced process technologies, NVIDIA would not have been so effective with power efficiency and performance of its 12nm-based GeForce RTX and Quadro RTX graphics cards, especially when put up against the latest 7nm node that AMD is using for its Navi GPU architecture and Radeon RX 5500 and RX 5700 series graphics cards.

Intel has secured yet another ex-AMD staffer, with Masooma Bhaiwala leaving AMD after over 15 years to help Intel and its discrete GPU products in 2020 and beyond.

Bhaiwala had a big hand in some of the semi-custom chips that AMD has made over the years, with these SoCs becoming a very important part of AMD's portfolio. Between Microsoft and Sony we have around 150 million Xbox One and PS4 sales, and then the upcoming (and very kick-ass) next-gen Xbox Series X and PlayStation 5 consoles will be continued success for AMD's semi-custom business.

Bhaiwala will be working inside of Intel's Graphics and Throughput Computing Hardware Engineering organization, which is a part of the Intel Architecture, Graphics and Software team that is run by ex-RTG boss Raja Koduri. Bhaiwala explained: "After 15+ amazing years at AMD, I have decided to take on a different opportunity... It was a truly fun ride, with an incredible team, during which we built some truly cool chips".

NVIDIA has just announced it will be releasing PhysX 5.0 "soon in 2020" with the updated physics middleware SDK offering up support for a "unified particle simulation framework".

In NVIDIA's own PhysX 5.0 announcement video you can see the physics at play here, where we can see the soft cloth of the foldable chair influenced beautifully. It reacts to being moved around, and when laid completely flat the cloth physics are displayed in their best light. Even when a plastic chair is thrown into the air and lands on top of the chair, the fabric moves how it would in real life.