TweakTown NewsRefine News by Category:
This is a super-juicy rumor, but let's get on with it: NVIDIA could be preparing a watercooled GeForce GTX Titan X that could be released as the Titan X Ultra.
This purported Titan X would see a hybrid AIO cooler that would allow for some serious overclocking of the GM200-powered Titan X, and see those performance numbers jump up even more. We could see NVIDIA or AIBs increase the Core Clock by 250-300MHz (or more) but that would see the Titan X getting even hotter, which would require the increased efforts on the cooling front.
This is where we would see the company using an AIO liquid cooling unit that would keep the card nice and cool, even under all of that increased pressure. This would be the perfect counter to the impending release of the Radeon R9 390X from AMD, which should be here any week now.
We heard the rumors, but ASUS has just made its new GTX 980 20th Anniversary Gold Edition official: the world's fastest GeForce GTX 980 video card. The company has used its impressive DirectCU cooler with 0dB-fan technology that allows for games like League of Legends and StarCraft to be played in total silence.
The all-important fact here is that it has a Boost Clock of 1431MHz, which is 215MHz above the reference Boost, providing an additional 15% performance. 4GB of GDDR5 is splashed on top, with three DisplayPort outputs, one HDMI and one dual-link DVI. One of the more unique parts of the VGA card is that it has a beautiful gold and black cooler, which is going to look incredible in gamers' PCs.
ASUS has told us that the GTX 980 20th Anniversary Gold Edition is a "short run, exclusive limited edition card", with its clock speeds beating even the best GTX 980s on the market from EVGA and its Kingpin GTX 980 and GALAX HOF 980 cards. We have reached out to our contact at ASUS to try and secure a sample, so check back for a review here at TweakTown soon.
ASUS is beginning to tease a sure-to-be record-breaking GeForce GTX Matrix Gold Edition video card, where the company wants you to guess "how fast" you think it's going to be.
The company uses the reference GTX 980 and Matrix Platinum GTX 980 as examples, with the stock GTX 980 clocking in at 1126MHz while the Matrix Platinum cranks that up to 1342MHz. For ASUS to be bragging about this, we would have to guess and opt for 1500MHz. We should expect to be introduced to the new 20th anniversary Gold Edition card in the near future.
It seems like a given, but NVIDIA is preparing another VGA card for release in the very near future: the GeForce GTX 980 Ti. The GTX 980 Ti would be a cut down version of the GM200-powered Titan X, with 6GB framebuffer and it'll be around 10% faster than Titan X, too.
According to SweClockers the new GeForce GTX 980 Ti would be the flagship card that NVIDIA will push, versus the Titan X. One of the reason this is, in my opinion, is that the new GTX 980 Ti can have add-in-board (AIB) partners play with the PCB and cooling setup, something that can't be done on the Titan X.
This will allow for some huge overclocks on the GTX 980 Ti, and some very exotic cooling setups. We should expect NVIDIA to unveil the GeForce GTX 980 Ti sometime in the second half of the year, after AMD launch its Radeon R9 390X.
It looks like AMD's Radeon 300 series is beginning to take better form, where we're learning that the new cards will be filled with a bunch of rebrands, something that has been discovered hiding in the Catalyst 15.3 Beta drivers. Here's the run down of what is found in the Catalyst 15.3 Beta drivers:
- AMD665F.1 = "AMD Radeon R9 360″
- AMD6610.2 = "AMD Radeon(TM) R7 350X"
- AMD6610.3 = "AMD Radeon(TM) R5 340X"
- AMD6611.10 = "AMD Radeon R7 340″
- AMD665F.1 = "AMD Radeon R9 360″
- AMD6660.1 = "AMD Radeon(TM) R5 M330″
- AMD6660.2 = "AMD Radeon(TM) R5 M330″
- AMD6660.3 = "AMD Radeon(TM) R5 M330″
- AMD6778.8 = "AMD Radeon R5 310″
- AMD6811.1 = "AMD Radeon R9 370″
An early April Fools joke is appearing online, so we're going to make it clear: this will probably not happen. But, OverclockersUK and BFG are teasing a new GeForce GTX Titan XXX, which has a huge 24GB of VRAM, double the framebuffer of the GeForce GTX Titan X, which has 12GB.
Most of the specs are identical to the reference GTX Titan X, except it has 24GB of VRAM and higher clock speeds. We have two 8-pin PCIe power connectors which use up to 300W of power in "300W OC mode" or 250W at its stock settings. The RAM has been overclocked to 8600MHz, up from 7000MHz on the reference card.
We have a Base Clock of 1268MHz and Boost Clock of 1493MHz, versus the 1002MHz and 1072MHz of the reference GTX Titan X. These overclocks are mammoth, but it would be very interesting to see this become a real card. The price? A huge $2400+, compared to the $999 pricing of the GeForce GTX Titan X.
The release of the card? Exclusive to OverclockersUK on April 1.
GTC 2015 - We didn't expect to see HBM 2 shown off in person here at NVIDIA's GPU Technology Conference, but here we are - SK Hynix has a HBM 2 wafer on the show floor, and it's beautiful.
Here it is, in all of it's glory.
A closer shot of the wafer itself.
SK Hynix's HBM memory promises increased bandwidth, as well as lower power consumption compared to conventional DRAM found on current cards, such as GDDR5.
We haven't even got HBM1 out yet, but HBM2 is promising so much more. HBM1 will have 128GB/sec per 1Gbit chip, while HBM2 will offer double that, with 256GB/sec with 2Gbit chips. We are also going to see lower power consumption, with 1.2V for HBM1 and HBM2, compared to GDDR5 which uses 1.35 to 1.5V.
GTC 2015 - One of the things we didn't quite have the time to talk about yesterday was the 3D Memory and HBM 2.0 side of the upcoming Pascal architecture form NVIDIA. This architecture should find itself onto a new VGA card sometime in 2016, rocking a huge 32GB of VRAM.
NVIDIA briefed the crowd at its GPU Technology Conference here in San Jose, California where they displayed a slide with the amount of VRAM per GPU architecture. We can see that Kepler in 2012, which the last flagship card based on Kepler was the GeForce GTX 780 and GTX Titan Black featuring 3GB and 6GB of RAM respectively, while the Maxwell architecture provides between 4GB (on the GTX 980) and 12GB (on the GTX Titan X) respectively. The slide teases that Pascal will feature 32GB of RAM, and Volta will rock up to 72GB of RAM in 2018.
The way NVIDIA will do this is thanks to SK Hynix's High Bandwidth Memory (HBM) which allows for four-layer stacks, which is also known as 4-Hi. This will come in 1GB and 2GB varieties, but eight-layer stacks will eventually arrive, which should see a huge increase in the amount of framebuffer on the next generation of GPUs.
Not only will Pascal deliver more VRAM on the card, but it will have magnitudes more memory bandwidth. The Maxwell-based GeForce GTX Titan X has 336GB/sec of memory bandwidth from its 384-bit memory bus on its GDDR5 RAM, but the Pascal architecture will be capable of a huge 750GB/sec or more. This will be using a variety of technologies to achieve this lofty height of memory bandwidth, including mixed precision, 3D Memory and NVLink.
We should expect NVIDIA to talk more about Pascal later in the year, or GTC 2016 this time next year.
GTC 2015 - NVIDIA has just unveiled the new Linux-based Digits Devbox which is not a mass production platform, but a PC that is built as the "world's fastest deskside deep learning system" priced at a huge $15,000.
Inside, you'll find four of the new Maxwell-powered GeForce GTX Titan X cards, which will super accelerate deep learning with its insane GPU capabilities. The system is built-to-order, where you have a direct contact at NVIDIA once you purchase the system to help you out. This isn't a gaming PC, but we're sure this system could probably run Crysis.
GTC 2015 - NVIDIA has just started providing many more details on its next generation GPU architecture, known as Pascal. Pascal is going to be a giant leap from Maxwell, where it will be "10x faster than Maxwell" according to co-founder and CEO of NVIDIA, Jen-Hsun Huang.
Pascal will feature NVLink and 3D Memory which will provide some impressive performance and bandwidth, where we can expect 3x the memory bandwidth of Maxwell, which is quite the leak. On top of that, we have 400% more mixed precision, but the Pascal architecture itself is a huge leap for NVIDIA, even compared to the already impressive Maxwell architecture.
We have NVLink which has 5x the bandwidth of PCI Express, which is quite the jump in terms of available bandwidth. NVLink will provide the ability of having 8 GPUs, up from the 4-way limit of SLI right now.