Graphics Cards - Page 129

Falcon Northwest has just announced that it is offering AMD's new Radeon RX 6950 XT reference graphics card, inside of its new Tiki Mini PC.

The new Radeon RX 6950 XT reference aka "Black Edition" GPU isn't a custom model at all, but a reference design from AMD -- rocking the same style cooling design as the Radeon RX 6900 XT reference -- but a sweet all-black look versus the silver look on other reference RX 6000 series GPUs from AMD.

You'll be spending at least $4000+ by configuring a high-end Falcon Northwest Tiki Mini PC, but remember you're getting a great amount of PC hardware inside of a small SFF gaming PC.

AMD just launched its new Radeon RX 6950 XT, Radeon RX 6750 XT, and Radeon RX 6650 XT graphics cards in the last couple of weeks... and now we're hearing rumors of the Radeon RX 6700 non-XT graphics card.

The new Radeon RX 6700 non-XT graphics card was spotted in custom form from SAPPHIRE, in their PULSE family of cards, and an unknown SKU reported on by Cowcotland. Inside, we should expect a Navi 22 GPU with 36 Compute Units, and what should be 2308 Stream Processors.

We're expecting to see 10GB of GDDR6 memory on the Radeon RX 6700 non-XT, on a 160-bit memory bus while the GPU should have a 2330MHz game clock, and a 2495MHz boost clock. SAPPHIRE is using a single 8-pin PCIe power connector on each model, so we should expect a TDP of 200W or less on the Radeon RX 6700 non-XT.

NVIDIA and AMD's next-gen GPUs will be out later this year, with rumors that the highest-end flagship GPUs in the form of the GeForce RTX 4090 Ti... might consume up to 600W of power... well, you've got to cool that 600W of power.

Taiwan's cooling component suppliers foresee fantastic demand for its products in 2H 2022, which is when NVIDIA and AMD will be launching their next-gen GeForce RTX 40 and Radeon RX 7000 series GPUs. According to the latest report from DigiTimes, their sources have said Taiwanese manufacturers like Auras Technology and Sun Max, are excited for 2H 2022 and beyond.

The two Taiwanese cooling component suppliers sat with DigiTimes to talk about the next-gen graphics cards from NVIDIA and AMD, and how their knowledge in cooling designs will be required in the next few years with GPUs going ballistic in performance (and power).

Intel has teased its next-gen Rialto Bridge data center GPU a the International Supercomputing Conference in Germany, for next-gen AI and HPC dominance. Check it out:

The new Intel Rialto Bridge GPU will succeed the upcoming Ponte Vecchio GPU that launches later this year, where Rialto Bridge will pack up to 160 Xe-Cores, and enhanced tiles on the next-gen Intel process node. Intel teases increased density, performance, and efficiency but wasn't clear on the process node: spoiler alert, expect Intel 4.

Intel's upcoming Ponte Vecchio GPU has up to 128 Xe-Cores, with the new Rialto Bridge GPU packing up to 160 Xe-Cores we have a 25% increase in core count. We will see the company most likely using super-fast HBM3 memory on Intel Rialto Bridge, as Intel has promised "increased I/O" bandwidth -- without numbers to back it up.

The GPU rumor mill is spinning back up for NVIDIA's next-gen Ada Lovelace GPU architecture, and upcoming GeForce RTX 40 series GPUs... of which, the GeForce RTX 4090 is launching first.

NVIDIA is expected to unleash its new flagship GeForce RTX 4090 graphics card in August, the new GeForce RTX 4080 a month after in September, followed by the GeForce RTX 4070 in October. AIB partners still have significant numbers of GeForce RTX 30 series GPU stock left, so we'll have to see those gone first before a next-gen GPU family can be launched.

We're expected to see NVIDIA to debut the new GeForce RTX 4090, GeForce RTX 4080, and GeForce RTX 4070 all on the same day... but then stagger the launches in the weeks after.

We might be consumed with CPU news with AMD's upcoming Zen 4-based Ryzen 7000 series CPUs, teasing a 16-core engineering sample at 5.5GHz+ but now we're back to GPU rumors again with NVIDIA reportedly launching the higher-end GeForce RTX 4090 first.

According to the latest from leaker "kopite7kimi", NVIDIA will reportedly launch the GeForce RTX 4090 first, then the GeForce RTX 4080 and GeForce RTX 4070 after. This would break tradition, as NVIDIA normally launches the x080 and x070 series GPUs first, followed by the x090 series GPU... but the RTX 4090 launching first makes sense.

AMD is gearing up its new RDNA 3-powered Radeon RX 7000 series GPUs wiht the Navi 31-based purported Radeon RX 7970 XT graphics card set to go toe-to-toe and evne possibly beat NVIDIA's new flagship GPU. This makes sense for NVIDIA to launch this new GeForce RTX 4090 first if the purported Radeon RX 7970 XT is going to dominate. NVIDIA will want to strike first with the best, while the Ada Lovelace GPU iron (well, silicon) is hot.

MSI must have a big stock of GeForce RTX 3080 graphics cards as it is getting rid of a bunch of them with a new branded version, for a game that I'm sure barely anyone plays anymore.

The "new" MSI GeForce RTX 3080 SUPRIM X Assassin's Creed Valhalla Special Edition is nothing really special at all, just a re-painted version of the GeForce RTX 3080 SUPRIM X. Ubisoft's Assassin's Creed: Valhalla is the star here: with some new colors on the shroud and backplate and some logos on the side.

I mean it doesn't look totally bad, but there's just nothing new here apart from a re-painting of MSI's GeForce RTX 3080 SUPRIM X -- down to the identical clock speeds and power settings -- with 1920MHz GPU boost, and up to 370W of power with 3 x 8-pin PCIe power connectors.

AMD's next-gen RDNA 3 GPU architecture will debut later this year, with major testing going on right now: one of those tests involves the confirmation of the next-gen DisplayPort 2.0 standard on Radeon RX 7000 series graphics cards of the future.

In a new tweet discovered by "Kepler," we see that AMD has kicked off putting patches for DisplayPort 2.0 UHBR (Ultra-High-Bit-Rate) certification. DisplayPort 2.0 UHBR comes in UHBR10, UHBR13.5, and UHBR20 standards with varying bandwidth, resolution, and refresh rate support.

Why do we need DP2.0 UHBR20? Well, 16K at 60Hz (with DSC) is pretty damn impressive... the same goes for 10K @ 60Hz, and what will be my favorite: 4K @ 240Hz. All over a single cable.

TCL has been busy leaking out Sony and Microsoft's upgraded current-gen consoles, which we've already heard rumors about, but now we're finding out where the GPU performance might land when compared to the PC.

Polish website PPE attended a press event that TCL held recently, where it talked about its predictions for the upgraded consoles. TCL talked about the "new Xbox Series S/X" and "PS5 Pro" consoles, predicting that we'll see them push out 4K @ 60-120FPS "render" while they'll have 4K to 8K @ 60-120FPS "output".

More interestingly, TCL predicts that the upgraded consoles will have GPU power comparable to AMD's upcoming Navi 33-based Radeon RX 7700 XT graphics card. Out of context, this sounds amazing... but we did hear last month that the Radeon RX 7700 XT would beat the Radeon RX 6900 XT. So, what does this mean?

NVIDIA has just pushed out its latest GeForce Game Ready 512.95 WHQL drivers, which have the usual best day-0 gaming experience for a couple of the latest games. You can grab them right here (786MB download).

Hitman 3 is getting its huge "Year 2" update which packs both NVIDIA DLSS support and ray tracing, as well as Sniper Elite 5 which isn't out yet -- but you've got the best day-0 gaming experience with these new GeForce Game Ready 512.95 WHQL drivers.

NVIDIA didn't have too much to celebrate at Computex 2022, but the company did show off its hectic new ASUS ROG Swift 500Hz gaming monitor: a new 1080p @ 500Hz (yes, 500FPS) with a brand new "E-TN" (Esports TN) panel. We also heard about 10 other games being announced NVIDIA DLSS, including the aforementioned Hitman 3 + Sniper Elite 5 as well as F1 22, and many more super-powered with DLSS tech.

NVIDIA is reportedly gearing up to unveil a new version of its A100 Tensor Core GPUs, check it out:

The new NVIDIA A100 PCIe "Ampere" accelerator has been updated with a liquid-cooled variant, and man it looks so damn good. NVIDIA is making this card, using a beautiful design that sees the tubing connectors on the rear, sitting directly to the left of the single 8-pin PCIe power connector.

A liquid-cooled A100 accelerator isn't exactly new, as you could buy the passively cooled dual-slot version and manually replace the cooler... but that is a lot of work. Professionals aren't going to do that, they're going to want to buy it made like this... and install it into their system. This is a work of art.

NVIDIA's next-gen GeForce RTX 40 series GPUs will be released this year, with the mosnter new flagship GeForce RTX 4090 / RTX 4090 Ti shaping up to be quite the gigantic beast.

In recent leaks, we've seen the gargantuan cooling system that the purported GeForce RTX 4090 Ti Founders Edition would use: what looks like a 4-slot cooler. But in his latest video, Tom @ Moore's Law is Dead said that he's been talking with AIBs who said:

"AIBs are preparing 4.5-slot coolers right now for some undisclosed NVIDIA product. Already case-makers are adjusting their designs to support these monsters".

AMD's next-gen RDNA 3 graphics architecture will be led by the new Navi 31 GPU, which has changed specifications according to the very, very latest leaks.

Navi 31 is now reportedly packing a wider 384-bit memory bus, up from the 256-bit memory bus from previous rumors... while also packing 24GB of GDDR6. Navi 31 has been mentioned with a 256-bit memory bus + 32GB GDDR6 before, so this new SKU is interesting.

The newer Navi 31 rumored GPU specs would see the 384-bit memory bus not just increasing overall memory bandwidth over a 256-bit memory bus, but AMD would require fewer memory modules at the same time. This means that AMD and its AIB partners can make their Navi 31-based offerings cheaper, and require less cooling (which also helps with costs).

ASUS has revealed some interesting information in their recent earnings report, with ASUS Co-CEO Sy Hsu explaining GPU demand for crypto miners is dying, while prices could stabilize soon, but maybe not.

ASUS's Co-CEO, Sy Hsu explained: "Because the demand for cryptocurrency mining on GPU shipments has been slowly coming down, the demand for graphics cards across the market is normalizing. Because the demand from cryptocurrency is disappearing, it's made us wonder if the pricing for GPUs will also normalize. In reality, the demand for gaming is still strong, so we still don't think we can necessarily meet all of the demand".

We've seen GPU prices skyrocketing over the last year or so, where prices were hitting 150-300% of their regular MSRP... but now, now they're floating back down to reality. This isn't just for one company, but rather both -- AMD and NVIDIA graphics cards are seeing most retailers and e-tailers pricing Radeon and GeForce GPUs at 5-10% above MSRP on average.

AMD's next-gen RDNA 3 architecture will debut later this year, with the introduction of the flagship Navi 31 GPU that we're now hearing is codenamed "Plum Bonito".

The new codename Plum Bonito GPU is the Navi 31 ASIC board according to leaker "Kepler", who recently said that the Navi 31-based purported Radeon RX 7950 XT graphics card will arrive on the next-gen PCIe 5.0 interface. We can also see that a Navi 31 board is reportedly called "Gemini" which could be a dual-GPU version of Navi 31 (workstation-class GPU possibly).

But if you look close, the next-next-gen Navi 41 GPU was teased in very early form as "gfx1200". AMD's current-gen Radeon RX 6000 series graphics cards are based on the RDNA 2 GPU architecture, while next-gen Radeon RX 7000 series graphics cards will be powered by the RDNA 3 GPU architecture. After that, we have the Radeon RX 8000 series powered by RDNA 4... of which we're now being teased.

We have been hearing plenty about NVIDIA's next-gen Ada Lovelace GPU-powered GeForce RTX 4090, but now it's time for the GeForce RTX 4090 Ti -- and more specifically -- NVIDIA's upcoming GeForce RTX 4090 Ti Founders Edition graphics card.

In some of the latest rounds of rumors, we've been hearing whispers that NVIDIA would unleash the GeForce RTX 40 series GPUs sooner rather than later -- a possibly mid-July release -- meaning we're just months away. NVIDIA looks to be providing some rather impressive cooling array to the RTX 40 series Founders Edition GPUs so that they can handle the 450W to 600W of power we're also hearing about.

NVIDIA's new GeForce RTX 4090 Ti (this name could change) Founders Edition graphics card has three different parts: first up we have the silver-themed metal skeleton that goes around the heat sink for that beautiful finish. You can see the nice little "RTX 4090 Ti" logo there, too.

Intel's upcoming Arctic Sound-M data center GPU has been spotted in the flesh by ServeThe Home, packing the new DG2 GPU -- the same GPU that will eventually find its way into the Intel Arc A-series GPUs, hopefully, maybe, later this year.

Intel will have two models of its Arctic Sound-M GPUs with 150W and 75W power, with the Arctic Sound-M in 150W form packing a single GPU, the Arctic Sound-M in dual-GPU form using 75W. The dual-GPU model has 75W TDP and is a smaller half-height, low-profile form factor design.

The new GPU photos that STH snapped are from the Intel Vision conference for developers, investors, and media. Intel had multiple SKUs of its Arctic Sound GPUs on the floor, including models that didn't have an EPS power connector, with STH also noting that many components were missing on some of the cards.

COLORFUL has just announced its new GeForce RTX 3060/3070Ti BattleAx NARAKA Bladepoint Edition, now available in RTX 3060 + RTX 3070 Ti form.

The new GeForce RTX 3060/3070Ti BattleAx NARAKA Bladepoint Edition graphics cards join competitor ZOTAC and their GeForce RTX 3080 Ti X-Gaming Naraka Bladepoint Edition graphics card. COLORFUL's new GeForce RTX 3060/3070Ti BattleAx NARAKA Bladepoint Edition has a plain-ish design, with Naraka aethestics through some stickers.

Inside, we have the GeForce RTX 3070 Ti with a faster set of 12GB GDDR6X memory at 19Gbps, while the RTX 3060 version of the card uses 8GB of GDDR6 at 15Gbps.

ASUS has just announced its new GeForce RTX 3080 Noctua Edition graphics card, the second in the ASUS x Noctua family of GPUs.

The new ASUS GeForce RTX 3080 Noctua Edition moves to a dual-fan cooling design, away from the 3 x 92mm fans on the original ASUS TUF Gaming GeForce RTX 3080. In their place are 2 x Noctua NF-A12x25 120mm fans that are much quieter, and a heat sink that keeps the GPU, GDDR6X memory, and VRMs much cooler.

Kent Chien (Corporate Vice President and General Manager of the ASUS Multimedia Business Unit) explains: "The synthesis of our custom-engineered unified heat sink and Noctua's famously whisper-quiet fan technology forges a creation that is more than the sum of its parts. This inspiring partnership with a world-class leader in DIY PC cooling establishes a noise level breakthrough for the powerhouse GeForce RTX 3080 Noctua Edition graphics card - and a visual style that you won't find anywhere else".

The rumor mill for next-gen GPUs runs harder than ever before in silicon history, with fresh rumors on NVIDIA's upcoming flagship GeForce RTX 4090 graphics card.

The new GeForce RTX 4090 with an Ada Lovelace-based AD102-300 GPU rocking 16128 CUDA cores, 24GB of GDDR6X memory at 24Gbps, and a 450W TDP will reportedly pack twice the performance of the GeForce RTX 3090 graphics card.

This is according to leaker "kopite7kimi" on Twitter, who said: "OK, let's do a new summary. RTX 4090, AD102-300, 16128FP32, 21Gbps 24G GDDR6X, 450W, ~2x3090. I am disappointed with RDNA3. That's all". This lower 450W power news flies in the face of previous 600W power claims... as the RTX 4090 (450W) is now rumored to be twice as fast as the RTX 3090 (350W). That's a huge achievement for only 100W increase in TDP.