Graphics Cards - Page 194

If there was a custom GeForce RTX 2080 Ti that I'm waiting for, it would be the MSI GeForce RTX 2080 Ti Lightning Z which has now been spotted - at least naked, so we get a good look at the PCB.

We are expecting MSI to unveil the new GeForce RTX 2080 Ti Lightning Z at CES 2019 in a few days time, with new HWBOT submissions teasing the card and its existence. There are two HWBOT entries with GPU clocks that exceed 2450MHz, which will be thanks to the monster RTX 2080 Ti Lightning Z and its 19-phase VRM and 3 x 8-pin PCIe power connectors.

There are multiple fan connectors and voltage connectors on the end of the card, with the entire end of the card cover ed in its own awesome heat plate that should keep that side of the card cool. This is a big deal for overclockers and enthusiasts, as there will be a LOT of power going into these cards when they're pushed to beyond their limits with LN2 cooling and the like.

I reported back in April 2018 that NVIDIA's new GPU would be the Turing architecture followed by Ampere, and now we're hearing more about the next-gen GPU architecture from NVIDIA that will arrive in 2020.

NVIDIA will reportedly be using Samsung's new 7nm extreme ultraviolet lithography (EUV) process that uses a plasma laser to drive silicon material into 7nm transistor structures. The new GeForce RTX and Quadro RTX graphics cards are both made on the 12nm process thanks to TSMC, with NVIDIA no stranger to working with Samsung for GPU production.

AMD will have the world's first GPU on 7nm beating NVIDIA to the punch, with its upcoming Navi GPU launching in July 2019 according to our sources. NVIDIA's new GPU architecture on 7nm should end up being Ampere, which will succeed the current Turing GPU architecture inside of the new RTX-powered cards.

ZOTAC is preparing its new GeForce RTX 2060 family of graphics cards with two members showing up to the party ahead of time, with the new ZOTAC GeForce RTX 2060 AMP and GeForce RTX 2060 Twin Fan graphics cards turning up on the internet.

The new ZOTAC GeForce RTX 2060 graphics cards were discovered by HD Tecnologia, with both models looking pretty much identical. The main difference will come down to the GPU and GDDR6 clock speeds between the cards, with the AMP series usually reserved for factory overclocked while the Twin Fan will most likely be better-than-Founders clocks and temperatures because of the improved thermal design over what the Founders Edition is expected to be like.

Some of the promotional shots see it as ZOTAC marketing the new cards for Mini-ITX gaming rigs, offering real-time ray tracing in a lower price bracket than the GeForce RTX 2070. A single 8-pin PCIe power connector is all that's needed.

NVIDIA was in the headlines over the Christmas break with a new GeForce RTX 2060 that will include not one, not two, but six different models. If this leak is correct, we're looking at an absolute landslide of RTX 2060 graphics cards to choose from, and it's going to get messy.

The latest leak on the GeForce RTX 2060 is from a new entry at the EEC (Eurasian Economic Commission) where three different SKUs are listed with two different memory technologies. There will reportedly be a 6GB, 4GB, and 3GB model of the GeForce RTX 2060 in both GDDR6 and GDDR5 models, making for six cards launching under the GeForce RTX 2060.

If this is true it's going to be very confusing for consumers to choose between the cards as there are six in the price range that the GeFOrce RTX 2060 will launch in. Testing performance between each exact card is going to be harder, as one card should be faster than another in a particular game and resolution, but lose against it in something else.

Intel is still over a year away from launching its discrete graphics card, promised in 2020, but between now and then we have a beautiful new render to take our minds off of the 'what ifs' about Intel Xe.

3D artist Cristiano Siqueira has teased us with other graphics card concept art, namely Radeon graphics cards, but now he has dipped his toes in the Intel Xe. Siqueira's concept art of the Intel Xe sees it taking design inspiration from the Intel Optane storage world, with a blower-style fan, LEDs, and a full cover backplate with slick Xe branding.

In this concept art, the Intel Xe graphics card dons an awesome black/silver look with blue LEDs that really do look great. I don't know why Intel wouldn't be taking this design into consideration when they begin the design process of Xe in 2020. The designs really are kick ass.

NVIDIA announced its new TITAN RTX in a push by influencers who pretended they didn't know they received an NDA, a new unreleased graphics card from the largest GPU company in the world, and then went live with videos in a 'oops, did I do that' push.

Well, that new $2499 graphics card has now reached the hands of people who are pushing them to their limits with PC enthusiast 'Death' posting on Twitter about the new TITAN RTX, and the overclocking adventures of the card when watercooled. The card belongs to 'jbn1010' from South Korea, with a Bykski waterblock and LCD installed onto the TITAN RTX it was then overclocked and benchmarked.

The new TITAN RTX was overclocked to 2070MHz on the GPU and 2025MHz on the memory, with 3DMark Fire Strike the only benchmark shared so far. The card achieved 31,862 points in Fire Strike, and 41,109 for the graphics card. This isn't bad considering a GeForce RTX 2080 Ti in our Core i7-8700K test bed scores 34,910 total and a graphics score of 26,760.

It's funny... when I was in Hawaii covering the Snapdragon Summit that Qualcomm hosted, we noticed that JEDEC was hosting a meeting there. We didn't know what JEDEC were announcing, but now we know: an update to the HBM standard.

The first consumer graphics card to launch with HBM technology was from AMD in the form of the Radeon R9 Fury X, followed by the HBM2-based Radoen RX Vega family of cards. AMD and NVIDIA continue to use HBM2 today with AMD deploying it on their consumer level Radeon RX Vega 56 and 64 cards but also their Radeon Pro family with various Vega GPUs and even the world's first 7nm GPU with Vega 20 rolling out with 32GB of HBM2. NVIDIA's current flock of Tesla V100 graphics cards are powered by up to 32GB of HBM2 memory, while the new Quadro RTX 5000/6000/8000 cards are all using GDDR6 memory.

JEDEC has since updated the HBM standard to 24GB per stack with 307GB/sec memory bandwidth per stack, giving AMD and NVIDIA the option of using 4-Hi and 8-Hi stacks of HBM2 that would pave the way for up to 96GB of memory on a single card. If we had a card with a 4096-bit memory bus (like the Radeon RX Vega 64 for example) and used the updated HBM2 standard, we're looking at 1.2TB/sec of memory bandwidth.

The best benchmark for testing out NVIDIA's new GeForce RTX series graphics cards is nearly here, with the new 3DMark Port Roay benchmark to be one of the first real-time ray tracing benchmarks available.

GALAX has recently hosted its huge GALAX GOC 2018 overclocking event where some of the best overclockers in the world descended to Ho Chi Minh City, Vietnam to play around with the new 3DMark Port Royal benchmark among other things. Some of the benchmark runs have been published which means we get an early look at what performance to expect on the flagship GeForce RTX 2080 Ti graphics card from NVIDIA.

The highest 3DMark Port Royal benchmark score is from Swedish overclocker Tobias 'Rauf' Bergstrom who scored 11069, with an average of 51.25FPS on the GALAX GeForce RTX 2080 Ti that was overclocked to 2640MHz on the GPU and 2088MHz on the GDDR6, all cooled with liquid nitrogen. Inside, Rauf used an overclocked Core i9-9900K and an ASRock Z390 Phantom Gaming-ITX/ac motherboard.

AMD was the first to kinda strike 8K video editing but couldn't quite get there, something we exposed during SIGGRAPH 2017 in the days after the Radeon RX Vega unveiling in LA. But now, NVIDIA is handling real-time 8K video editing on a single Quadro RTX 6000 graphics card with some truly impressive results.

During an event in the US recently, RED Digital Cinema announced a partnership with NVIDIA and announced NVIDIA CUDA-accelerated REDCODE RAW decode SDK that provides software developers and studios new ways to work with 8K video content. RED and NVIDIA announced the news in front of some of the largest names in the industry including Adobe, Colorfront, HP, and others.

RED Digital Cinema president Jarred Land said: "Our mission is to bring cinema-grade images and performance to content creators everywhere. RED, NVIDIA and our industry partners are leveling the playing field, making the technology for high-resolution processing and image quality accessible to everyone".

AMD announced the world's first 7nm GPU just last month during its Next Horizon event in San Francisco, where it also unveiled the next-gen EPYC 'Rome' CPU with 64C/128T which will also be made on 7nm. But now... a new trademark for Radeon Vega II has been filed by AMD, something you can scope out below.

VideoCardz has noticed that AMD registered the Radeon Vega II trademark two weeks ago now, which should see the company using this slick new logo for the upcoming 7nm Vega 20 GPU that will power the next-gen Radeon Instinct graphics cards.

The new Radeon Instinct MI60 is coming soon, built on the 7nm node and rocking a huge 32GB of HBM2 with built-in ECC and 1TB/sec of memory bandwidth. The new Radeon Instinct MI60 will also be the first to be on PCIe 4.0, offering 64GB/sec of bi-direction CPU-to-GPU bandwidth.

Intel has just peeled back the veil of its future generation CPU architecture and plans for the next couple of years, but one of the areas the company touched on during its Architecture Day was its new scalable graphics architecture known as Xe.

Between now and then we've got Gen 9 graphics, Gen 11 is around the corner and then in 2020 we'll begin to see Xe. The new Xe scalable graphics architecture will be split into four markets: integrated/entry-level markets, mid-range, enthusiast and datacenter/AI. The one I'm most interested in here would have to be the mid-range and enthusiast Xe products, as they'll be the ones that will compete with NVIDIA and AMD in 2020 and beyond.

Kicking things off we have Gen 11 graphics detailed, with Intel teasing continued efficiency and improved performance. There's also advanced 3D, media, and display capabilities with Gen 11, as well as better gaming experiences. I need this in my hands before I sign off on better gaming experiences, but Intel has been continuing to make improvements in gaming performance over the years and things are really kicking into another gear now that most of RTG's best talents work for Intel now.

ZOTAC has done an accidental oops that was quickly removed, teasing for a moment a purported GeForce GTX 1070 that features GDDR5X memory like the bigger brother in the GTX 1080.

The leaked model continues to pack NVIDIA's GP104 GPU which features 1920 CUDA cores, but the big change here is the 8GB of GDDR5X memory. VideoCardz has noticed that the GDDR5X-based GTX 1070 from ZOTAC has a TDP of 250W, which means this is just a GTX 1080 with artificially disabled CUDA cores. This would be strange, but NVIDIA could have an excess of GTX 1080s that they want to get rid of, so they could disable CUDA cores and release a GeForce GTX 1070 with GDDR5X.

NVIDIA could also have a crap load of GDDR5X memory left that it can't use, since GDDR6 is the way forward with the GeForce RTX series graphics cards. Whatever is happening, we could expect a GDDR5X-based GeForce GTX 1070 to be unveiled in the very near future.

So, I've been in Hawaii for Qualcomm's annual Snapdragon Summit and saw these rumors break while I was enjoying the lifestyle that Maui has to offer, but wanted to wait until I was home and in front of my workstation to get all of my thoughts ready for this rumor.

A few days ago various sources including AdoredTV and Chiphell reported that AMD's first Navi GPU will be the Navi 10 and that it will have the performance of NVIDIA's latest GeForce RTX 2080. This seems like a huge stretch, as I've had personal sources tell me earlier this year that Navi will be "just as bad" as Vega but weren't specific in how bad it would be. I took it as the performance will be similar to Radeon RX Vega 64, but now we're hearing about the price and performance that will be "surprising".

It was only back in April that rumors floated around that Navi would offer GeForce GTX 1080 performance for $250... and now the latest rumors are saying that it will offer GeForce RTX 2080 performance for $249. This rumor corroborates the previous rumors, but they're just that... rumors. The rumors state that the 7nm node will help AMD offer some out-of-this-world power efficiency, but even if we're looking at GTX/RTX 2080 performance... it won't be here until the second half of 2019, so we're at least 6-7 months a way from its release.

NVIDIA already has a slew of Turing-based graphics cards on the market in the form of GeForce RTX and Quadro RTX cards, but now it seems it's nearly time for the TITAN RTX to shine.

In an "accidental" leak Linus from LinusTechTips showed off the TITAN RTX box during the recent WAN show, as well as The Slow Mo Guys showing off the card itself with "TITAN" branding on it in place of the "GeForce RTX" branding we've seen on the RTX 2080 Ti, RTX 2080 and RTX 2070 graphics cards.

We don't know anything more than this, but this means that we can expect TITAN RTX graphics cards to be seeded out to reviewers in the coming weeks, especially if cards and boxes are in the hands of influencers now.

If you own one of NVIDIA's new GeForce RTX graphics cards there's not much you can do with it with real-time ray tracing right now apart from Battlefield V (if you can even get it to work with RTX that is), but that'll soon change with 3DMark's new ray tracing benchmark.

The new 3DMark Real-Time Ray Tracing Benchmark Port Royal will be teased ahead of its January 2019 launch at the GALAX GOC Grand Final overcloccking event in Ho Chi Minh City, Vietnam on December 8. What makes this special is that UL Benchmarks will be the first to market with a dedicated real-time ray tracing benchmark for gamers, where they can (get this) test real-time ray tracing performance of any graphics card that supports Microsoft's DirectX Raytracing standard, DXR.

UL Benchmarks explains: "Real-time ray tracing promises to bring new levels of realism to in-game graphics. Port Royal uses DirectX Raytracing to enhance reflections, shadows, and other effects that are difficult to achieve with traditional rendering techniques. As well as benchmarking performance, 3DMark Port Royal is a realistic and practical example of what to expect from ray tracing in upcoming games-ray tracing effects running in real-time at reasonable frame rates at 2560 x 1440 resolution".

Intel has more hype around its upcoming discrete GPU codenamed 'Arctic Sound' with fresh rumors hot off the press from DigiTimes saying that Intel will be hosting a conference in December to unveil its new GPU. Yeah, well, that's not happening.

I reached out to some industry sources to poke around and see what's going on, and it seems it will not be happening next month at all. Intel has previously said it will be launching its discrete GPU in 2020, so why would they be ready for a full detailing of the Arctic Sound architecture in late-2018? Previous rumors teased a reveal at CES 2019 but seemed to have fizzed out, as Intel is sticking to its 2020 release window.

Intel has secured most of the talent from AMD's own Radeon Technologies Group in the last year or so, with GPU architect Raja Koduri joining Intel as well as Radeon marketing legend Chris Hook just to name a few. These talented individuals have vast resources at their hands under Intel compared to the playing cards they had at AMD to fling around the room.

NVIDIA lost a gigantic $23 billion in market cap in a single 24-hour period on Friday, with NVIDIA stock peaking on October 1 at $289.36 and at the time of writing was down to a much weaker $144.70.

Why? Well, there are many reasons - with some claiming it was the cryptocurrency mining bubble popping, but it seems the various issues with the newly launched Turing-based GeForce RTX series being center focus for NVIDIA right now. NVIDIA stock dropped heavily after the company posted their Q3 2018 earnings report which saw the company missing revenue estimates.

NVIDIA shares reached a 52-week low a couple of days ago when they hit $161.61 but the continued spiral down to $144.70 is worrying, with this now being the worst 24-hour percentage drop in 10 years. The last time this happened was back in July 2008 when NVIDIA shares dropped 30% over disappointing sales forecasts. Jensen Huang, CEO and founder of NVIDIA told MarketWatch in an interview on Thursday afternoon that this was all over a "crypto hangover", where Huang explained: "The crypto hangover lasted longer than we expected and we were surprised by that, but it will pass".

Years ago the marketing slogan for Intel was 'Intel Inside' but Intel aren't inside of as many products as AMD these days, with Team Red inside of all variants of the Xbox and PlayStation consoles, and interesting devices like the upcoming SMACH Z.

SMACH Z is a new handheld PC device that is powered by AMD Ryzen Embedded and features Radeon Vega 8 graphics, with the SMACH Z team announcing that they will be launching their new product in Q1 2019. The SMACH Z team explained: "The SMACH Z team confirms that providers are already working on components production and delivery. Mass production of SMACH Z, the device that promises to revolutionize the world of PC gaming, is scheduled for Q1 of 2019. After two years of development, the device is ready to go to market with a 6-inch high-definition touchscreen and hardware based on AMD Ryzen Embedded and AMD Radeon Vega 8 Graphics".

What will the SMACH Z be powered with exactly? Well, there are there different variants to choose from: SMACH Z, SMACH Z PRO, and SMACH Z ULTRA. Each of these models feature varying specs, with the SMACH Z including 4GB of RAM, 64GB of storage and NO camera while the other models bump up the RAM and storage with each tier, and both include a camera.

NVIDIA has given MSI the great ability of making a bunch of awesome custom GeForce RTX graphics cards, but then MSI just upped the ante with its new GeForce RTX 2070 AERO ITX graphics card... the first ITX-based RTX 2070.

The new MSI GeForce RTX 2070 AERO ITX is the first small form factor ITX-based RTX series graphics card, packing everything Turing has to offer except for the inclusion of VirtualLink connectivity for next-gen VR headsets. Connectivity wise MSI still includes 3 x DP1.4 ports and 1 x HDMI 2.0, less than the other RTX series cards, but then this is the one and only ITX style RTX series card.

MSI doesn't overclock the GeForce RTX 2070 AERO ITX, with a single-fan cooler and 175W TDP. It should run quite warm with stock fan speeds, but I'd like to see the cooling on this card when it's under stressful gaming loads, and/or overclocked.

NVIDIA has released a new set of GeForce drivers with the GeForce Game Ready 416.94 WHQL release, which include optimizations for the latest games that have dropped with Battlefield V, Fallout 76, and Hitman 2. You can download the new GeForce Game Ready 416.94 WHQL drivers here.

The new driver should be of importance to GeForce RTX series owners with Microsoft resuming the rollout of its troubled Windows 10 1809 update that was broken from day one. This new update adds in DirectX ray tracing (DXR) abilities, which in turn enables RTX features for Battlefield V for GeForce RTX 2080 Ti, RTX 2080, and RTX 2070 graphics card owners.

NVIDIA notes on the official driver download site: "Game Ready Drivers provide the best possible gaming experience for all major new releases, including Virtual Reality games. Prior to a new title launching, our driver team is working up until the last minute to ensure every performance tweak and bug fix is included for the best gameplay on day-1".