Video Cards & GPUs News - Page 256

We've seen MSI's upcoming GeForce RTX 2080 Ti GAMING X TRIO as well as the upcoming MSI GeForce RTX 2080/2080 Ti DUKE... and now it's time for Palit.

Palit will reportedly, according to the leak from VideoCardz, be offering their new series in the form of the GamingPro and GamingPro OC. Both cards will use custom PCBs and cooling solutions, with a dual-fan cooler on both models.

Both of the Palit offerings in the GeForce RTX 2080 and GeForce RTX 2080 Ti will have factory overclocks applied (clock speeds unknown right now), with 8GB of GDDR6 on the RTX 2080 and 11GB of GDDR6 on the GTX 2080 Ti.

I'm literally sitting at the airport here in Adelaide, South Australia writing all of this news about the new flow of leaks on NVIDIA's new GeForce RTX 20 series and it isn't stopping. Germany, here I come!

Before I went to bed last night I wrote up about MSI's new GeForce RTX 2080 Ti GAMING X TRIO graphics card, something VideoCardz leaked to the world and now they're back with a leak on MSI's other RTX 20 series card.

We now have our first look at the MSI GeForce RTX 2080 and RTX 2080 Ti DUKE graphics cards, which are similar - but not identical to the GAMING X TRIO variant.

It looks like NVIDIA will be unveiling the higher-end GeForce RTX 2080 Ti next week, and not just the expected GeForce RTX 2070 and RTX 2080 graphics cards. VideoCardz is now showing off some photos of MSI's first custom RTX 2080 Ti graphics card.

MSI's new GeForce RTX 2080 Ti GAMING X TRIO graphics card is similar to the current GTX 1080 Ti GAMING X TRIO, but the new card features a few fairly large changes, apart from the new Turing GPU and GDDR6 memory. The new RTX 2080 Ti GAMING X TRIO has a VirtualLink connector, the new USB-C port for next-gen VR headsets but it is the loss of SLI and gain of NVLink which has me happy.

I only just wrote about NVIDIA ditching the SLI connector in favor of NVLink, which could see VRAM doubling instead of being wasted on the other card. This would mean that two RTX 2080 Ti graphics cards in NVLink would have a truly insane 22GB of GDDR6 at its disposal.

NVIDIA is expected to announce its new GeForce RTX range of graphics cards during Gamescom 2018 next week, with all signs pointing to the RTX 20 series unveil taking place on August 20 in Cologne, Germany.

VideoCardz is reporting information from a member of the Baidu forums who claims to "have a lot of information from a board partner". This source said that the Turing-based GeForce RTX 2080 would cost $649, but not only that it would beat the TITAN V... a graphics card that costs a whopping $3000. TITAN V is based on the Volta GPU architecture and packs HBM2 memory technology, compared to the RTX 2080 with its Turing GPU architecture and GDDR6 tech.

The purported GeForce RTX 2080 would feature 3072 CUDA cores from the Turing GPU architecture, 8GB of GDDR6 memory on a 256-bit memory bus. We don't know GPU base/boost clocks and I'm sure we'll see a large change to the way it works with tweaks made to NVIDIA GPU Boost technology. We don't know what speeds the GDDR6 will be clocked at, but we do know Samsung is providing its GDDR6 at up to 14Gbps.

Intel has released a short video at Siggraph 2018 this week in Vancouver teasing about their upcoming discrete GPU in 2020. The newly installed Intel discrete graphics boss, Raja Koduri, has been spotted roaming around Vancouver this week as well. The Twitter post by Intel can be found here.

The image of the card in the video has been confirmed to us by Intel as a rendering and not an actual product image. Intel made no formal presentations this year at Siggraph in Vancouver about their upcoming plans, although we are expecting more information to begin appearing in 2019.

Raja Koduri joined Intel late in 2017 as we reported here. Earlier this year Chris Hook of AMD and ATI fame, as well Tom Forsyth joined the expanding Intel discrete graphics team.

NVIDIA should be announcing its next-gen GeForce RTX series of graphics cards on August 20, with more details, speculation and rumor being thrown onto the internet than ever before.

We're now seeing some detailed specifications on not just the forthcoming GeForce RTX 2070 and RTX 2080 but the higher-end RTX 2080 Ti that should truly blow the doors off of everything else on the market. The current rumors peg the RTX 2080 at beating the TITAN V, and considering that's a Volta-based HBM2-powered $3000 graphics card, those claims are beyond imagination.

We should expect NVIDIA to release a beefed up GeForce RTX 2080 Ti if the naming scheme stays the same as the GTX 10 series (GTX 1070, 1080, 1080 Ti) and from the current crop of information, it will be an utter beast. If the GeForce RTX 2080 is beating the TITAN V, then the RTX 2080 Ti would be truly insane. It should have 11GB of GDDR6 on a 352-bit memory bus (the GTX 1080 Ti has 11GB of GDDR5X on a 352-bit memory bus).

NVIDIA unveiled its next-gen Turing GPU architecture during SIGGRAPH 2018 with the new Quadro RTX range of cards featuring GDDR6 memory from Samsung with transfer speeds of up to 14Gbps.

Samsung is providing its new GDDR6 technology for NVIDIA's latest cards, with the company using the 10nm node for its GDDR6. Jim Ellott, Corporate Senior Vice President at Samsung Semiconductor explained: "It's a tremendous privilege to have been selected by NVIDIA to launch Samsung's 16Gb GDDR6, and to have enjoyed the full confidence of their design team in making our key contribution to the NVIDIA Quadro RTX GPUs".

NVIDIA's new Quadro RTX 8000 has an astounding 48GB of GDDR6 and when used in two-way NVLink mode that VRAM framebuffer doubles to a huge 96GB of super-fast GDDR6 memory. Samsung's new 10nm-built GDDR6 tech will eventually run at up to 18Gbps (up from 14Gbps) and that each die can transfer data at 72Gbps, which is a two-fold increase over the same GDDR5 chips with 8Gbps bandwidth.

SIGGRAPH 2018 - NVIDIA revealed the next-gen Turing GPU architecture today during SIGGRAPH 2018, with some new Quadro RTX graphics cards that are powering all sorts of new demos.

One of those includes an unveiling of "The Speed of Light", a collaboration between Porsche, NVIDIA, and Epic Games. The Speed of Light was created with Unreal Engine 4 using real-time ray tracing, NVIDIA's RTX technology, and their new Quadro RTX graphics cards. The demo was powered by two of NVIDIA's new Quarto RTX graphics cards, all running in real-time.

Manager of Virtual Design at Porsche, Christian Braun, said: "Porsche's collaboration with Epic and NVIDIA has exceeded all expectations from both a creative and technological perspective. The achieved results are proof that real-time technology is revolutionizing how we design and market our vehicles".

NVIDIA just unveiled its next-generation Turing GPU architecture during SIGGRAPH 2018, with three new Quadro cards unveiled: Quadro RTX 5000, RTX 6000, and the flagship Quadro RTX 8000.

Well, just a couple of hours ago NVIDIA uploaded a video to their YouTube account dubbed 'Be For The Game'. The video shows off a bunch of high-quality shots of an unknown graphics card - could be a GTX 10 series card, or it could be a next-generation GeForce graphics card based on the just-announced Turing architecture. There are a few teases of '20' and '2080' in the video, with one particular part standing out.

At the end of the video, NVIDIA has edited it perfectly to display the date '20.AUG.2018' except that the '20' and '08' come up last... which could be a tease of '2080'. It could be some sly trick to tease us of the GeForce GTX 2080, or I could be looking into it too much. But for NVIDIA to rearrange those numbers: 2, 0, 8, 0 and have the other numbers still placed there - they're highlighting '2080' for a reason.

SIGGRAPH 2018 - NVIDIA has just kicked off its SIGGRAPH 2018 press conference with CEO Jensen Huang unveiling their new Turing GPU architecture and three new Quadro RTX graphics cards.

The new Turing GPU is the largest leap in GPU technology from NVIDIA since the CUDA GPU was unveiled back in 2006, with a huge 18.6 billion transistors on a 12nm process. There's also a Tensor Core inside with 125 TFLOPs of FP16 performance, and an RT Core with 10 Giga Rays/sec of performance. The Turing GPU in all of its glory uses 754mm2, compared to 471mm2 of Pascal (which had 11.8 billion transistors).

NVIDIA revealed three new Quadro RTX branded cards with the Quadro RTX 5000, Quadro RTX 6000, and Quadro RTX 8000. The flagship Quadro RTX 8000 features 4608 CUDA cores, an insane 48GB of GDDR6 memory and an equally as huge $10,000 price tag. The new Quadro RTX 6000 has its VRAM cut in half to 24GB with the rest of the specs at the same level for $6300 while the RTX 5000 has its CUDA cores cut down to 3072, and its GDDR6 down to 16GB with a price of $2300.