Graphics Cards - Page 53

Let's face it: we all know NVIDIA will unleash DLSS 4 upscaling technology with its next-generation GeForce RTX 50 series "Blackwell" gaming GPUs, and now we have some 'proof' from some new leaks posted on X. Check it out:

NVIDIA's new GeForce RTX 50 series GPUs will be announced at CES 2025 and released in the weeks and months later, with the GeForce RTX 5080 launching first, and the new ultra-enthusiast and mega-expensive GeForce RTX 5090 launching in the weeks after. Now we're getting early retailer listings for an ASUS RTX 5080 graphics card which mentions DLSS 4 technology.

In a new post on X, leaker "kopite7kimi" confirmed DLSS 4 is coming by saying "DLSS 4 is true" and that we should expect those monster new 1080p 750Hz, 600Hz, and ultra-high-end 4K 240Hz QD-OLED gaming monitors have enough performance out of the RTX 5090 and RTX 5080 with DLSS 4 enabled to hit those lofty high-performance mega frames.

NVIDIA's upcoming GeForce RTX 50 series Laptop GPUs have been leaked, with the mobile Blackwell GPUs spotted inside of the PCI ID repository: the new RTX 5090, RTX 5080, RTX 5070 Ti, RTX 5070, RTX 5060, and RTX 5050 Max-Q laptop GPUs have been teased.

In the PCI ID repository files we can see "AD108M" listed against multiple RTX 50 Max-Q GPUs, suggesting that NVIDIA could replace the Blackwell GPU dies with Ada Lovelace GPU dies. AD108M doesn't really exist in the ADa family, and can only be found up to the AD107 for the RTX 50 series GPUs.

NVIDIA's Max-Q chips are usually destined for thin-and-light, power-efficiency gaming laptops versus regular Laptop GPUs for high-performance gaming laptops. AD108M seems like a strange choice, acting as an entry-level GPU for rRTX 5070 and other lower-class mobile GPUs.

NVIDIA's next-gen GeForce RTX 5090 was expected to feature a 600W TDP, but new leaks suggest we'll see a 575W TDP on the company's new ultra-enthusiast graphics card.

In a new post on X, leaker "hongxing202" said "next-gen 90 > 575W" to which reputable leaker "kopite7kimi" replied with "and RTX 5080 360W". NVIDIA's current-gen flagship GeForce RTX 4090 has a 450W TDP, which means the new RTX 5090 has 125W more power at 575W TDP, making for some considerable power draw (and performance).

NVIDIA's flagship ultra-enthusiast GeForce RTX 5090 will feature the GB202 GPU with the full 21,760 CUDA cores (and the first GeForce GPU to sport over 20,000 CUDA cores). We'll have 32GB of ultra-fast GDDR7 memory on a 512-bit memory bus and 1.79TB/sec of memory bandwidth, at a 575W TDP.

MSI's new GeForce RTX 5080 GAMING TRIO graphics card has just been leaked ahead of its CES 2025 reveal, confirming some specifications along the way.

The new MSI GeForce RTX 5080 GAMING TRIO was spotted on Chiphell before the post was removed, confirming that this is the OC Edition, and that the GeForce RTX 5080 does indeed feature 16GB of GDDR7 memory on a 256-bit memory bus.

We will have 3 x DisplayPort connectors and 1 x HDMI port (these should be DisplayPort 2.1 connectors, and a HDMI 2.1 port) but full confirmation of that will happen when the unveiling takes place next week at CES 2025.

With NVIDIA set to release a new generation of desktop and laptop GeForce RTX 50 Series graphics cards this year, today we've got an early look at mainstream GeForce RTX 5060 laptop GPU performance via a leaked 3DMark Time Spy benchmark from reviewer Superalloy Skittles at Bilibili (via Huang514613 on X and Videocardz).

The synthetic 3DMark Time Spy benchmark is a DirectX 12 API test that has been around for a few years. The GeForce RTX 5060 laptop GPU's Graphics Score of 13821 indicates that NVIDIA's new mainstream laptop GPU will deliver a 30% performance increase over its predecessor, the GeForce RTX 4060. And a massive 70% increase over the GeForce RTX 3060 laptop GPU.

We're also seeing a 10% improvement compared to the GeForce RTX 4070 laptop GPU, which is nice to see. As a synthetic benchmark, the performance might not translate 1:1 to in-game performance, but it indicates that NVIDIA targets a typical 30% gaming performance improvement for the Blackwell generation.

With NVIDIA days away from formally announcing its next-generation GeForce RTX 50 Series of PC gaming GPUs, the leaks and rumors are coming in thick and fast. The latest arrives courtesy of Hong Kong outlet HKEPC (via Videocardz) and points to a January 21, 2025 launch for the GeForce RTX 5080.

This new information is in line with previous reports stating that NVIDIA will launch the new Blackwell generation of GPUs with the GeForce RTX 5080, with the flagship GeForce RTX 5090 and mid-range GeForce RTX 5070 and RTX 5070 Ti following in late January and February, respectively. There's no word on the specific release date for these models, but January 21 for the GeForce RTX 5080 makes sense, with the reveal happening on January 6 at CES 2025.

NVIDIA's latest architecture will power the GeForce RTX 5080 and move to a more advanced process. Like other models in the 50 Series, it's set to adopt cutting-edge GDDR7 memory for a significant increase in overall memory bandwidth. However, it will retain the same 16GB capacity as its predecessor.

NVIDIA's purported next-gen graphics cards, the RTX 5080 and 5070 Ti GPUs, have been listed by a PC maker, ahead of what should be their imminent reveal at CES 2025 (at Team Green's big keynote on January 6).

The product listings for gaming PCs from iBuyPower (picked up by @momomo_us on X, as flagged by VideoCardz) have now been taken down, but they showed two of NVIDIA's Blackwell GeForce graphics cards, the RTX 5080 and 5070 Ti as mentioned. Both were listed with 16GB of video RAM as previous chatter from the grapevine has suggested.

These are the rumored models that will in theory be first to arrive, alongside the flagship RTX 5090.

NVIDIA's next-gen ultra-enthusiast GeForce RTX 5090 32GB is in the headlines again, this time teasing its 16+6+7 phase power design, 14-layer PCB, and up to 600W TGP on custom RTX 5080 graphics cards. Oh boy.

In a new post from Benchlife and their AIB contacts, we're hearing that the flagship GeForce RTX 5090 will use a 16+6+7 phase design, compared to the 20+3 phase design used on the RTX 4090, and the 18+3 phase design on the RTX 3090 Ti. The outlet also reports that we'll see a 600W TGP on the RTX 5090, but we should see numbers lower than that on the RTX 5090 Founders Edition.

Benchlife also reports that the GeForce RTX 5090 will use a 14-layer PCB, which is most likely destined for the reference board of the RTX 5090, and NVIDIA's in-house GeForce RTX 5090 Founders Edition graphics card.

ASUS is ramping up into the launches of countless new products at CES 2025, but now we're hearing about the new ROG ASTRAL brand making its grand appearance with custom versions of the GeForce RRTX 5080 16GB graphics card on the way.

In a new post on X by leaker "momomo_us" we're learning that ASUS is preparing 7 new graphics cards (at least on this leak) with 3 custom GeForce RTX 5080 graphics cards. The first two are the ROG ASTRAL GeForce RTX 5080 16GB (two different SKUs) as well as the PRIME GeForce RTX 5080 16GB.

ASUS is also working on a bunch of custom RDNA 4-based Radeon RX 9000 series GPUs, with TUF Gaming Radeon RX 9070 XT and RX 9070 graphics cards, as well as the PRIME Radeon RX 9070 XT and RX 9070 graphics cards.

GIGABYTE has just teased a next-generation AORUS graphics card that will be revealed at CES 2025, which looks like it'll be a new AORUS GeForce RTX 50 series GPU with an AIO cooler. Check out the (very brief) teaser below:

We aren't far away from CES 2025 at all now, with just one week to go, and now GIGABYTE is teasing a new AORUS graphics card, right as all the leaks on NVIDIA and AMD's next-gen gaming GPUs are spreading like wildfire over the Christmas holidays. We can see some text on the fans that reads "TUFO" which stands for "Team Up, Fight On", the usual GIGABYTE flair here.

The folks over at VideoCardz think that we're looking at the new AORUS GeForce RTX 5090 Xtreme Waterforce graphics card, with some images shared reminding us what that card looks like... but from a brief few-second teaser with just the fans, that's not enough for me.

Launching earlier this month, the Intel Arc B580, the first Battlemage desktop GPU, has proven itself as a real mainstream PC gaming contender. With a lower price point than the GeForce RTX 4060 and better performance in several key titles, including games with ray-tracing, 12GB of VRAM, and mature drivers, it's a big win for Intel and the Arc Graphics team.

According to a new report at Quantum Bits, spotted by @harukaze5719 on X, Intel plans to launch a desktop Battlemage GPU in 2025 with 24GB of memory - double what's found on the Intel Arc B580. The catch? This will be a GPU for the AI, 'Pro,' and workstation market, so it's not a card built for gaming.

This will be a successor to the company's first generation of Intel Arc Pro GPUs and will probably sport the same workstation-friendly slimline design with a blower fan for cooling. With the growing demand for AI workloads and dealing with complex Large Language Models (LLMs), this move makes sense.

In the leadup to CES 2025, where NVIDIA CEO Jensen Huang is set to unveil the next generation of GeForce RTX gaming hardware as part of the company's live keynote presentation, the company is celebrating 25 years of GeForce. As part of a #GeForceGreats campaign on social media, celebrating PC gaming and GeForce history, NVIDIA is offering up some pretty amazing 'Mystery Box' prizes.

One of the latest includes an excellent nod to one of the all-time great GPUs, the GeForce GTX 1080 Ti. In 2025, the GTX 1080 Ti is still a capable GPU, but it's a card that does not support the current GeForce RTX 40 Series line-up's game-changing AI-powered technology like DLSS and Frame Generation.

Created in partnership with Bethesda, the latest #GeForceGreats 'Mystery Box' includes collector's edition copies of DOOM, DOOM Eternal, and The Elder Scrolls Online: Imperial Edition. It also has an impressive GeForce RTX 4080 SUPER Founders Edition, a custom Bethesda backplate, and a very cool bonus.

AMD's next-gen RDNA 4 flagship GPU will be the mid-range Radeon RX 9070 XT. The company is also expected to announce the lower-tier Radeon RX 9070 at CES 2025. Both cards are aimed at the mid-range GPU market, which is currently dominated by the GeForce RTX 4070 and RTX 4070 SUPER.

The Radeon RX 9070 XT is effectively the successor to the Radeon RX 7800 XT from the RDNA 3 generation, with AMD opting to forego calling it the Radeon RX 8800 XT. The new naming puts desktop Radeon model numbers on par with the current Ryzen 9000 CPU Series while sharing similarities with its competition - the RTX 4070 and the upcoming RTX 5070.

Recent rumors and leaks have confirmed some interesting specs, including boost clock speeds of up to 3.1 GHz and the revelation that the Radeon RX 9070 XT reference design will feature a TBP (Total Board Power) of around 260W, with partner cards reportedly pushing this as high as 330W.

We're a week away from NVIDIA finally unveiling its upcoming Blackwell-powered GeForce RTX 50 Series at CES 2025. Several credible sources and rumors suggest the company will debut the new line-up with the GeForce RTX 5080 first, followed by the flagship GeForce RTX 5090 a few weeks later. Both GPUs are expected to hit shelves by the end of January 2025, with NVIDIA expected to keep the momentum going with more GPU releases in February.

Specifically, the GeForce RTX 5070 and GeForce RTX 5070 Ti. These two cards will greatly interest PC gamers as they target the mid-range or more affordable enthusiast market - NVIDIA's X70 class GeForce GPUs often become one of the best-selling options for each generation. Thanks to known GeForce RTX and NVIDIA insider @kopite7kimi over on X, we now have full details on the hardware configuration of these GPUs.

The GeForce RTX 5070 will feature 12GB of fast GDDR7 memory on a 192-bit bus, matching the capacity of the GeForce RTX 4070. It will also have a CUDA Core count of 6,144, higher than the GeForce RTX 4070's 5,888 CUDA Cores but lower than the RTX 4070 SUPER's 7,168 CUDA Cores.

NVIDIA's next-generation ultra-enthusiast GeForce RTX 5090 graphics card with 32GB of ultra-fast GDDR7 memory is expected to sell for around $2600, according to the latest rumors.

If the new GeForce RTX 5090 is priced at $2600, that is a hefty $1000 hike on the RTX 4090 that launched with an MSRP of $1699. NVIDIA's next-gen GeForce RTX 5090 is rumored to pump between 50-70% more performance at 4K over the RTX 4090, and utterly annihilate AMD's upcoming RDNA 4-powered flagship Radeon RX 9070 XT graphics card.

In a new post by @wxnod on X, we're to expect the retail cost of the NVIDIA GeForce RTX 5090 at 18,999 yuan, while the GeForce RTX 5080 costs 9999 yuan, which works out to around $2600 USD for the RRTX 5090 32GB and around $1400 USD for the RTX 5080 16GB.

AMD's next-gen Radeon RX 9070 XT graphics card rumors are swirling, with the flagship RDNA 4 chip said to boost at up to 3.1GHz and have a 260W to 330W TBP depending on the AIB model.

In a new post on Chiphell, we're learning that the flagship Radeon RX 9070 XT graphics card will have a reference board design with a TBP (Total Board Power) of around 260W, while custom Radeon RX 9070 XT graphics cards will chow down on up to 330W of power. This is a 27% difference, allowing for some heavy overclocking on the better-built custom 9070 XT graphics cards.

AMD's new Radeon RX 9070 XT is expected to have a 2.8GHz base GPU clock, with boost clocks of the flagship RDNA 4 chip hitting 3.0GHz to 3.1GHz depending on the model.

NVIDIA and AMD are reportedly shipping their new GeForce RTX 50 series and Radeon RX 9000 series GPUs from China to warehouses in the US before President Trump's new tariffs kick in on January 20, 2025.

Both companies are expected to unveil their new Blackwell and RDNA 4 gaming GPUs at CES 2025 in early 2025, with NVIDIA expected to release 6 new GeForce RTX 50 series GPUs across Q1 2025, with new graphics cards launched in the RTX 50 series GPU family between January and March, lining up with previous rumors.

In a new report from Ctee, we're hearing that in response to an unclear implementation schedule of the new tariff increase after President Trump takes office on January 20, 2025, graphics card supply chain industry partners have said that both major GPU manufacturers (NVIDIA and AMD) have asked their partners to ship custom AIB cards produced in mainland factories in China in advance, before January 20, to warehouses in the United States.

We've got another look at the purported PCB of NVIDIA's next-generation GeForce RTX 5090, showing off the massive GB202 GPU and this time, we see Samsung 28Gbps GDDR7 memory modules. Check it out:

In a Christmas Day leak, we saw the PCB of a custom GeForce RTX 5090 graphics card in a compact, but crowded PCB that looks insanely powerful with its array of components, but now we've got a closer look at the PCB and those juicy GDDR7 memory modules.

NVIDIA's new GeForce RTX 5090 PCB leak is coming from the Baidu forums, with the PCB fully decked out for the holidays with the GPU, VRAM, and VRMs. The GPU is labeled as the GB202-300-A1 GPU which lines up with previous leaks, with this most likely being a qualification sample that should be close to (but not the) final design.

NVIDIA's next-generation GeForce RTX 5090 is just weeks away and while we've had plenty of leaks, now it's time for a gander at the PCB of one of the custom RTX 5090 graphics cards.

In a new post on Chiphell, we're being introduced to a beautifully large PCB with over 40 capacitors and 16 GDDR7 memory modules. The PCB we're looking at is for a custom AIB model of the GeForce RTX 5090, with no visible NVIDIA branding on the PCB itself, so this isn't a Founders Edition or NVIDIA reference model, but a custom AIB card.

We've seen leaks of the GDDR7 memory layout and this new PCB leak lines up with that, with 5 modules on each side (4 on top, 2 below the GPU) with 16 modules in total which would allow for some delicious configurations: 32GB, 48GB, or even 64GB depending on whether 2GB, 3GB, or 4GB GDDR7 memory modules are used.

Intel launched its first 'Battlemage' desktop graphics card earlier this month with the new Intel Arc B580, which has a competitive $250 price point. With 12GB of VRAM and performance that outclasses the GeForce RTX 4060 and Radeon RX 7600, it presents a compelling mainstream PC gaming alternative, backed up with hardware and software that is leaps and bounds better than the first-gen Arc.

For an in-depth look at the new B580, check out our full reviews of the Intel Arc B580 Limited Edition and ASRock Intel Arc B580 Steel Legend GPUs. With reports of the new Arc B580 selling out in multiple regions and Intel promising weekly shipments to retailers to try and keep up with the demand, the Arc team has just released a new Game On Graphics Driver for its Intel Arc line-up.

Intel Graphics Driver 32.0.101.6325 and 32.0.101.101.6253 (which is weirdly a single package) fixes a few issues the Intel Arc B580 has shown with games like Elden Ring and F1 24 while outlining 'Known Issues' that the Arc Graphics team is working to resolve for the B580 - including software compatibility.