Graphics Cards - Page 114

It looks like the Chinese GPU market is about to get an interesting domestic, and international hopefully, series of graphics cards from Shanghai-based Muxi. Muxi is expected to launch China's first GPU dedicated for gaming by 2025.

Muxi, or Muxi Integrated Circuit Shanghai Co Ltd, is a chipmaker based out of Shanghai in China, was formed in 2020 and has been making high-performance AI and general-purpose chips for the domestic market in China. We've already seen Chinese companies offering locally-made CPUs that are now getting pretty comparable to Intel and AMD in the CPU business, but the GPU business still has some work to do.

We've got Birentech with its BR100 GPU that's made on 7nm with an impressive 77 billion transistors -- for comparison, NVIDIA's new Hopper H100 has 80 billion transistors on 5nm -- that's got performance that competes against some of the best GPUs from NVIDIA and AMD. But when it comes to gaming, well, there's a lot of work to be done... the best China can do for gamers is offer the GPU performance of the GeForce GTX 1050 Ti and GTX 1050. Not good enough.

Uh, so out of nowhere EVGA has terminated its relationship with NVIDIA in what the company describes was "about respect" and exits the graphics card business entirely... even after it had around 20 samples of EVGA GeForce RTX 4090 FTW3 graphics cards made.

EVGA will not be making any GeForce RTX series graphics cards of any form moving forward, nor will be making AMD Radeon GPUs or Intel Arc GPUs as it will not make graphics cards whatsoever. EVGA says that this decision will stay until the CEO is replaced, as current CEO Andrew Han is firm on his decision for the company.

EVGA has terminated its relationship with NVIDIA after it had deteriorated during the years, with EVGA telling GamersNexus that "it's about respect". The decision to sever the relationship with the biggest GPU company in the world was a principal decision, and not a financial one. EVGA is making things clear here: there were major issues between NVIDIA and EVGA (and I'm sure other AIBs).

NVIDIA's next-gen flagship GeForce RTX 4090 is no stranger to the world at this point, and it hasn't even been announced yet... but Lenovo is forging ahead showing off its monster new GeForce RTX 4090 LEGION graphics card.

The new quad-slot Ada Lovelace beast was teased by Wolfstame, the Lenovo China Gaming Desktop Product Planning Manager, with some photos of the new Lenovo GeForce RTX 4090 graphics card.

You can see that when Lenovo's new custom GeForce RTX 4090 is placed next to a 16-inch laptop (Lenovo's own Legion Y9000X laptop) it's virtually the same size in length (35.6cm for the RTX 4090, 35.8cm for the Y9000X laptop).

Intel CEO Pat Gelsinger usually has company-wide addresses on a Friday, and given how crazy the last couple of weeks have been about Intel Arc GPU cancellations... we're sure Gelsinger had a lot of questions thrown at him from Intel staffers regarding the leaks that they're reading and watching.

Now we're hearing some of that Intel inside (heh) goodness from leaker Moore's Law is Dead, who said that when Gelsinger was asked what was happening with AXG (Intel's Accelerated Computing Systems and Graphics Group) he said two things, according to Moore's Law is Dead, that have "a lot of repercussions if you read into them".

Firstly, Gelsinger reportedly said that the Intel board is behind his strategy and his vision for the company, and they're behind his decisions -- MLID adding that "so whatever decisions are about to be announced... Pat's in on them -- while secondly, Gelsinger said he wouldn't speak about specific product lines. But he said: "changes in the business were going to be announced very soon" hinting at next week.

Can you believe we're only hours away from NVIDIA's big reveal of their next-gen GeForce RTX 40 series graphics cards, based on the next-gen Ada Lovelace GPU architecture.

But now that we're under 48 hours from NVIDIA's upcoming "GeForce Beyond" broadcast during GTC 2022 this week, with the company uploading their "GeForce Beyond" keynote, which will feature CEO and founder Jensen Huang once again in his infamous leather jacket.

NVIDIA describes the "GeForce Beyond: A Special Broadcast at GTC" on YouTube with the description: "PC enthusiasts, don't miss the GeForce Beyond special broadcast. Learn about the latest breakthroughs in gaming, creating, and graphics technology -- from NVIDIA founder and CEO Jensen Huang in the opening of his GTC keynote on Sept 20 at 8 a.m. Pacific time (UTC-7)".

GALAX has just teased its new "Serious Gaming" graphics card in a mysterious tweet, where it tells gamers to "not panic" and that we could see a "New Serious" soon.

NVIDIA is about to announce its next-gen Ada Lovelace GPU architecture and new GeForce RTX 4090, and GeForce RTX 4080 16GB and RTX 4080 12GB graphics cards... so it makes sense that GALAX is teasing a "New Serious" which should refer to the GALAX GeForce RTX 4090 SG graphics card in the coming weeks and months.

GALAX has its GeForce RTX 3090 SG graphics card in some markets, which is interesting -- and damn beasty -- with an optional fourth fan that can be attached to the back of the card.

NVIDIA has just pumped out yet another teaser for its upcoming "GeForce Beyond" special broadcast next week during GTC 2022, where we should be introduced to the next-gen Ada Lovelace GPU architecture.

The new teaser is pretty brief, with a 12-second video that slides past a GeForce RTX-powered desktop PC system and then zooms down into "Diagram for the computation by the Engine of the Numbers of Bernoulli". This chart in particular was created by Ada Lovelace back in 1843, made for Charles Babbage's Analytical Engine which is considered the first "non-trivial" computer program.

NVIDIA's ANZ (Australia and New Zealand) Twitter account posted the GeForce Beyond teaser, which also included another hidden message in the form of a sticky note attached to the monitor. The sticky note has what seems to be a phone number (208) 629-7538 which some think "7538" refers to 75.38 billion transistors (recent rumors peg the flagship AD102 GPU with 75 billion+ transistors).

AMD's upcoming Radeon RX 7900 XT reference graphics card might be a monster in terms of performance, but it could be a tame kitten when it comes to power consumption compared to NVIDIA.

The new RDNA 3-powered Navi 31-based Radeon RX 7900 XT should battle NVIDIA's upcoming Ada Lovelace-powered GeForce RTX 4080 16GB graphics card at the very least, and could possibly battle it out with the flagship GeForce RTX 4090. But... here's the kicker: at much, much less power.

NVIDIA's new GeForce RTX 4090 is expected to use 450W+ power by default, and 660W+ in OC mode with custom AIB versions of the RTX 4090 to be fun (on your power bill, and heating your room/house up considerably). The RTX 4080 16GB is expected to have a default 340W TGP and up to 516W for OC models, so the RX 7900 XT falls right on the RTX 4080 16GB TGP numbers.

NVIDIA's next-gen GeForce RTX 40 series "Ada Lovelace" GPUs launch next week, and while we're hearing a mountain of info on Ada, RDNA 3 isn't hiding too far away.

AMD's next-gen RDNA 3-based Radeon RX 7900 XT flagship GPU has been teased, based on the upcoming Navi 31 GPU and now posing for the camera in a new way: a leaked PCB diagram from Igor's Lab shows off the Radeon RX 7900 XT in all its next-gen GPU glory.

The new leaked PCB on the Radeon RX 7900 XT shows us what should be very close to the reference design of the RX 7900 XT, where Igor has placed 3 x 8-pin PCIe power connectors on the PCB diagram. With each of the 8-pin connectors capable of 150W, the three here will chew down up to 450W (and 75W from the PCIe 4.0 x16 slot, if AMD uses Gen4).

NVIDIA's new GeForce RTX 4090 "Ada Lovelace" graphics card is nearly here, and now we're finding out that a couple of the top-end Ada GPUs will be split into two variants.

The first of them will be the new GeForce RTX 4090 graphics cards which will arrive in two forms: 450W default TGP and 600W+ OC mode for custom RTX 4090 variants from AIB partners. Moore's Law is Dead has pushed out a new video updating his audience with the latest from his sources.

The first source told him that the GeForce RTX 4090 will launch in the "first half of October" while the GeForce RTX 4080 launches in the "first half of November (as of now)". His source does state what I was reporting on earlier today: that NVIDIA will launch the RTX 4090 first (October) then the RTX 4080 16GB and RTX 4080 12GB in November, followed by the RTX 4060 series GPUs in January at CES 2023.

As we march into the launch of NVIDIA's next-gen GeForce RTX 40 series "Ada Lovelace" GPUs, we're hearing more about the Ada launch lineup: GeForce RTX 4090, GeForce RTX 4080 16GB, and GeForce RTX 4080 12GB according to the latest leaks.

NVIDIA will be launching the GeForce RTX 4090 in all its next-gen GPU glory in October, but in November we'll be greeted by not one, but two new GeForce RTX 4080 graphics cards. The first is the higher-end GeForce RTX 4080 with 16GB of GDDR6X memory, clocked at a higher 23Gbps bandwidth for up to 768GB/sec of memory bandwidth.

Inside, the GeForce RTX 4080 16GB will use the new AD103-300 GPU with 9728 CUDA cores, with a GPU boost of up to 2505MHz (nearly the same as the RTX 4090, with a boost clock of up to 2520MHz). NVIDIA's new GeForce RTX 4080 has a default TGP of 340W, while it can be modified up to its maximum power limit of 516W. Remember, the 340W + 516W power numbers are for the RTX 4080 16GB graphics card model.

Lenovo is joining the custom GeForce RTX 4090 family of graphics cards with GIGABYTE and ZOTAC this week, with its new quad-slot GeForce RTX 4090 LEGION graphics card... and man, it's a beast. Check it out:

The new leaked photos of Lenovo's upcoming GeForce RTX 4090 LEGION graphics card show us that it is an absolute, utter, case-crushing, and PCIe port-crushing quad-slot, right-over-your-motherboard crazy size. It's a quad-slot GPU with an updated RTX 40 logo design (the new font NVIDIA is using) and a single 12VHPWR power connector.

Lenovo's new custom RTX 4090 LEGION graphics card barely fits into their in-house LEGION gaming case, which the company does for its LEGION gaming brand of pre-built PC systems. It's actually a really interesting-looking design, something I actually dig -- just like I did with Lenovo's custom Radeon RX 6900 XT LEGION graphics card.

NVIDIA's next-gen GeForce RTX 4090 "Ada Lovelace" flagship GPU is getting more and more detailed as we get closer to launch, where we're hearing about maximum TGP and GPU boost clocks.

The new flagship GeForce RTX 4090 graphics card will use the AD102-300 GPU with 16384 CUDA cores with GPU boost clocks of up to 2520MHz according to VideoCardz, who combined the specs for the RTX 4090 from multiple sources and reports from leaker "kopite7kimi" and from the Chiphell forums, keep in mind that NVIDIA has AD102 in full-fat form, where it will reportedly have up to 18342 CUDA cores.

We also find out that the default TGP will be 450W while the maximum configurable TGP is 660W, with the maximum TGP being tweaked through the BIOS and might not be available to all of the custom GeForce RTX 4090 models that get released.

Exclusive: As we get closer to the launch of NVIDIA's next-gen GeForce RTX 40 series "Ada Lovelace" GPUs, we're seeing more and more leaked on the cards, but now I've just heard from an anonymous source of mine that NVIDIA has been stockpiling next-gen graphics cards for a couple of months now.

I've been speaking with a few industry sources of mine, with one of them well-connected in Taiwan saying that NVIDIA has been stockpiling GeForce RTX 4090, GeForce RTX 4080 16GB, and GeForce RTX 4080 12GB graphics cards since August. Yeah: next-gen GPUs have been waiting for us... sitting in warehouses... dormant.

Towards the end of July 2022, I reported that Chinese GPU suppliers said they were getting their first batches of GeForce RTX 40 series graphics cards arriving in August. We didn't hear much after that, but we have been hearing rumbles from the leaks, and then we've recently seen the retail packaging for the custom GIGABYTE GeForce RTX 4090 GAMING OC and ZOTAC GeForce RTX 4090 AMP Extreme AIRO graphics cards.

NVIDIA's next-gen GeForce RTX 4090 graphics card is back in the headlines, this time it's turned up at a Vietnamese retailer where the RTX 4090 is up for pre-order.

The new NVIDIA GeForce RTX 4090 graphics card is up for pre-order with Vietnamese retailer AZPC, which has the "basic RTX 4090 version" listed for 50 million VND ($2100 USD or so) while the "premium RTX 4090 version" is listed for 60 million VND ($2520 USD or so).

We don't know if the retailer has the GeForce RTX 4090 contracted so they know the price, or if they're just guessing, but "I_Leak_VN" on Twitter noted that they think AZPC workstation stores price on the RTX 4090 pre-order is the "most realistic price because Vietnam also had this price when the new RTX 3090 was released".

NVIDIA will be unleashing its new monster GeForce RTX 4090 graphics card next month, but the other Ada Lovelace GPU releases in the GeForce RTX 4060 will use 8 PCIe lanes.

The news of AD106 and AD107 GPUs will have 8 PCIe lanes is coming from leaker "kopite7kimi" with a new tweet which said AD106 "is not very strong" and that its 3DMark TimeSpy Extreme score is under 7000. The leaker also noted that AD106 and AD107 "are using PCIe x8".

NVIDIA could roll out its fleet of GeForce RTX 40 series GPUs with PCIe 5.0 support, giving the higher-end GeForce RTX 4090, GeForce RTX 4080, and GeForce RTX 4070 all the same PCIe 5.0 x16 lanes. These cards will be powered by AD102, AD103, and AD104: AD106 and AD107 -- if these rumors are true -- won't need 16 lanes of PCIe 5.0, so 8 lanes of PCIe 5.0 will be enough (which has the same bandwidth as PCIe 4.0 x16 for current-gen motherboards).

Taiwan Semiconductor Manufacturing Company (TSMC) is forging ahead with its 2nm process node technology, now that it's nailed the 4nm node for NVIDIA and its new Hopper H100 GPUs, 3nm is coming and 2nm... begins soon.

TSMC is currently nailing down its 3nm process node technology, which Apple will be using for its next-gen SoC inside of the iPhone 15 in 2023. TSMC is currently making wafers of Hopper GPUs for NVIDIA with the Hopper H100 being made on TSMC's new 4N process technology, with the 4nm process node by TSMC optimized and designed exclusively for Hopper H100.

After that comes 3nm for various designs including Apple, which we'll see in iPhones and their upcoming AR/VR headset in 2023 and beyond. Once 3nm is in products and being used across the world, TSMC has its future-tech 2nm process node technology, which the Taiwanese giant says will enter mass production in 2025.

GIGABYTE's new custom GeForce RTX 4090 GAMING OC graphics card has been spotted, mere hours after the list of 11 custom GeForce RTX 4090 graphics cards that the company is working on.

It must be the week for factory leaks because we just saw the ZOTAC GeForce RTX 4090 AMP Extreme AIRO graphics card smiling for the camera. We saw the retail packaging and the actual graphics card itself -- from the back at least -- but the new GIGABYTE leaks on the custom GeForce RTX 4090 GAMING OC don't include photos of the card itself.

What we do have is the front of the GIGABYTE GeForce RTX 4090 GAMING OC retail packaging, where we see a slick-looking box with a robot and a triangle -- GIGABYTE style -- but we do have the new NVIDIA font used here, and 24GB of GDDR6X memory confirmed once again.

NVIDIA's next-gen Ada Lovelace GPU architecture and its big reveal at GTC 2022 next week is just so close, we can smell the silicon... which the flagship AD102 GPU silicon will have over 75 billion transistors.

The new AD102 GPU with 75 billion transistors means that Ada Lovelace is skipping up real close to Hopper H100, with NVIDIA cramming in 5 billion more transistors for a total of 80 billion transistors. We don't know how the GPU architectures compare just yet, but both of them have huge L2 cache (Ada Lovelace has up to 96MB of cache, and Hopper with up to 50MB of cache).

All that cache takes up space, so the huge 75 billion transistor count on AD102 could be because of that, and why it has close to the same transistor count as NVIDIA's monster Hopper H100 GPU. The news is coming from leaker "kopite7kimi" who posted a vote on Twitter asking followers how many transistors they thought NVIDIA's flagship AD102 GPU had.

GIGABYTE is cooking up quite the family of next-gen GeForce RTX 4090 graphics cards, with the company filing 11 new RTX 4090 graphics cards with the Eurasian Economic Commission (EEC).

The new listings are for 11 custom models of GIGABYTE's extensive family of GeForce RTX 4090 graphics cards, from the flagship GIGABYTE GeForce RTX 4090 AORUS EXTREME WATERFORCE, down to the GIGABYTE GeForce RTX 4090 WINDFORCE 3X graphics card.

GIGABYTE will be offering its RTX 4090 WINDFORCE 3X, RTX 4090 EAGLE, and RTX 4090 GAMING graphics cards will be offered in "OC" models as well. Above that, there's the GIGABYTE RTX 4090 AORUS MASTER, RTX 4090 AORUS EXTREME, and then the water-cooled RTX 4090 AORUS XTREME WATERBLOCK and RTX 4090 AORUS XTREME WATERFORCE graphics cards. Quite a stack of flagship Ada there, GIGABYTE.