TweakTown NewsRefine News by Category:
NVIDIA's reference GeForce GTX 980 is a monster when it comes to power efficiency and heat output - and not in a bad way. The talk of a watercooled reference GTX 980 using an Asetek unit is interesting, as it'll be thought of in the same light as AMD's Radeon R9 295X2.
The GTX 980 pictured above has the Asetek cooler slapped into it, but if you look closer, there's still the reference fan installed. This will see the GPU itself cooled with the Asetek watercooler, keeping it nice and cool even with some crazy overclocking, while the rest of the components such as the VRM and RAM and cooled by the reference cooler.
It would be interesting to see NVIDIA offer out a new reference GTX 980 with the watercooling used above, as the Maxwell architecture has some massive headroom for crazy overclocks, which watercooling will only amplify.
NVIDIA has had huge success with its new second-generation Maxwell architecture, forming the new power-efficient, but super-fast GeForce GTX 980 and GTX 970 GPUs. AMD were rumored to fight back at GAME24, but nothing materialized, until now. WCCF Tech is reporting from a leak from overclockers.ru, that teases AMD's next generation GPUs.
AMD reportedly talked with 4gamer.net journalists in a round table discussion, teasing details on its upcoming products to compete against NVIDIA's new GTX 900 series of GPUs. The new product that AMD will fight back with is the Radeon R9 380X, and not the 390X that most would've presumed. The new R9 380X will be based on the Pirate Islands architecture, powered by the Fiji GPU. Overclockers.ru is reporting that AMD have three new cards in the pipeline, with the Fiji-based R9 380X which will replace the R9 290X, the Treasure Island-based R9 370X, and another.
The most exciting news is that AMD will not only reportedly fight back with new architecture, but it'll shift onto a smaller process, moving over to TSMC's new 20nm manufacturing process, as well as using 3D stacked HBM memory. The third card we mentioned above is what we're all here for today, with the new R9 390X reference GPU to feature AMD's hybrid "hydra' liquid cooling, which the company used on its dual-GPU R9 295X2 earlier in the year.
AMD is starting to ramp up its rumor machine, with talk of the new Radeon R9 390X, it's next generation flagship GPU. AMD's new R9 390X is expected to be available in the first half of 2015, according to DigiTimes.
The new Radeon R9 390X is based on AMD's Bermuda GPU core, which should kick some serious ass, bringing a slew of new things to AMD's silicon. First off, we should see the R9 390X being the first GPU to be built on TSMC's new 20nm manufacturing process, but the card is rumored to arrive with High Bandwidth Memory, or HBM. HBM is 3D stacked memory technology that will offer an incredible amount of bandwidth on the already-fast GDDR5 tech that is used, with around 100% more bandwidth, all while using less power.
AMD is also rumored to be using hybrid liquid cooling on the new reference R9 390X, similar to what the company used on its dual-GPU Radeon R9 295X2 earlier this year. The problem here is, without a next-gen GPU coming out until at least Q2 2015 or so, what will AMD do to fight off what is an incredible new set of GPUs from NVIDIA? Aggressive price cuts on its R9 290 series, that's what. The only thing AMD can do until its launches its new R9 300 series is drop the prices of what it has now, considerably, to fight off Maxwell from NVIDIA.
If you're rocking an NVIDIA GeForce GTX SLI setup, what better way to make it stand out from the crowd than with the new SLI LED bridges that the company unveiled during its GAME24 event.
The new SLI bridges come in three flavors: 2-way SLI, 2-way spaced SLI, and 3-way SLI. The 2-way SLI LED bridges are priced at $29.99, while the 3-way SLI bridge will set you back $39.99. The SLI bridges are designed to work with all modern GeForce GTX cards that support SLI. NVIDIA specifically mentions the following cards: GeForce GTX 770, GTX 780, GTX 780 Ti, GTX TITAN, GTX TITAN Black, GTX 970 and GTX 980.
There's one big caveat though, NVIDIA has built these beautiful SLI bridges for reference GPUs, so they might not work on the crazy third-party cards. You'll need to make sure that your custom-cooled GeForce GTX card will work with these new bridges before jumping on-board.
NVIDIA has impressed the world with the launch of its second generation Maxwell-based GPUs, the GeForce GTX 970 and GTX 980. The cards sip power compared to their already power conservative predecessors, but introduce a slew of new technologies, features and best of all: a great price point. Where to from here? GM200, that's where.
GM200 is now being teased, which is going to be the big new GPU from NVIDIA. NVIDIA's GM200 will feature a die size of 551mm^2, 20-22 SMMs, 2560-2816 CUDA cores, a 384-bit memory bus, a performance boost of around 50% over the GTX Titan Black Edition, with it launching sometime before the end of the year. What it'll arrive as, in terms of naming, is a mystery. With the power efficiency of the GM204 core, which powers the GTX 970 and GTX 980, we should expect a dual-GPU card offered as the GeForce GTX 990, so that's one name it won't use.
Moving on, we could expect it to be the GTX 980 Ti, but if it's offering a huge performance increase over the GTX Titan Black Edition, we might see something like the GTX 985 Ti, or GTX Titan X, something I like the sound of better. Then we have the shift to 20nm, which is expected next year, so if the new GM200 is built on the current 28nm node, the shift to 20nm is going to be an entire new world for NVIDIA. The company has positioned themselves incredibly well after the mess that was Fermi, so much so, that AMD really has no firepower right now - so it'll be interesting to see if NVIDIA holds back, waiting for AMD to release something, to quickly smack it down to reality with its new GM200-based card.
NVIDIA Editor's Day 2014 - NVIDIA had some truly impressive technology to showcase during its Editor's Day in the beautiful Monterey Bay in California, where the technology giant proved that the Moon landings weren't faked, thanks to its new Voxel Global Illumination technology.
Voxel Global Illumination, or VGXI, is dynamic, with 'no baking required' according to NVIDIA. It works on Epic Games' Unreal Engine 4, and other major engines (in Q4 of this year) and uses one-bounce indirect diffuse, specular, reflections, area lights and much more.
NVIDIA were bashing conspiracy theorists, as many of them say the Moon landings were faked. Due to the power found in its Maxwell architecture, VXGI comes alive, and is able to do some truly incredible real-time lighting to prove how light was reflecting off of the lunar surface, lighting up various objects, Neil Armstrong and Buzz Aldrin, as well as the lunar module itself.
This shot, has the real image side-by-side with the rendered image. I'll let you try and work out which one is which, so let us know in the comments section below.
NVIDIA Editor's Day 2014 - NVIDIA's G-SYNC technology is something I've truly embraced, sitting in front of the ASUS ROG Swift monitor right now as I type this, but 4K G-SYNC? Sign. Me. Up.
In Monterey Bay, NVIDIA was showing of Absolute Systems' GeForce GTX 980 SLI-powered system, running the aforementioned Acer 4K G-SYNC monitor, running Crysis 3.
A different look at the Absolute Systems PC.
Close up, the Acer 4K G-SYNC monitor is truly gorgeous. Look out for a review here on TweakTown of it in the near future.
The bad boys in question, two NVIDIA GeForce GTX 980 GPUs in SLI.
NVIDIA has officially launched its Maxwell-based GeForce GTX 900 series, with ZOTAC out of the gate with a bunch of new GTX 980 and GTX 970 GPUs. First off, we have three versions of each: Standard, AMP! Omega, and AMP! Extreme.
Starting with the ZOTAC GeForce GTX 980 Standard, we have the stock Core and Boost Clocks of 1126MHz and 1216MHz, respectively. The 4GB of GDDR5 RAM is clocked at 7010MHz, with the reference cooler, and the usual three DisplayPort outputs, one HDMI 2.0 and dual-link DVI port. Moving onto the AMP! Omega edition, which cranks the Core Clock to 1202MHz, and Boost to 1304MHz.
We have the Memory clock up to 7046MHz, the same 4GB of GDDR5, identical display outputs, but the "Dual-fan IceStorm" cooler. Finally, the AMP! Extreme which doesn't have Core, Boost or Memory Clock speeds revealed yet, but we do know it comes with a sexy-sounding "Triple-fan IceStorm" cooler, so expect the clocks on this GPU to be much higher. All three cards have the same number of CUDA cores, sitting at 2048.
Manli has announced its new Maxwell-based NVIDIA GeForce GTX 980 and GTX 970 GPUs, both rocking a reference cooler, with a third offering a cooling design by Manli for its GTX 970.
Starting with the Manli GeForce GTX 980 which is powered by NVIDIA's GM204 Maxwell-based GPU, which has a TDP of 178W, up from the stock 165W on NVIDIA's reference GTX 980. We have a 2048 CUDA cores, 4GB of GDDR5 RAM on a 256-bit memory bus, a Core Clock of 1126MHz, Base Boost Clock of up to 1216MHz, and 7010MHz Memory Clock. For display connectivity, the Manli GTX 980 has three DisplayPort outputs, a single HDMI 2.0 out, and two DVI outputs.
Moving onto the GTX 970, we have two models: the reference GTX 970 and then the aftermarket design. The reference card features 1164 CUDA cores, 4GB of GDDR5 RAM on a 256-bit memory bus, a Core Clock of 1050MHz, Boost Clock of 1178MHz, and 7010MHz on the memory, the same as the GTX 980. When it comes to the display connectivity, we have the same ports as the GTX 980: three DisplayPort outputs, one HDMI 2.0 and two DVI ports.
Manli has also provided a slightly modified GTX 970 with an aftermarket cooler, something that comes with slightly increased Base and Boost Clocks. We have 1076MHz, up from 1050MHz for Base, and 1216MHz up from 1178MHz for the Boost Clock. The Memory Clock remains the same: 7010MHz. These cards should be available right now in select markets.
NVIDIA Editor's Day 2014 - Another interesting technology unveiled by NVIDIA in Monterey Bay was MFAA, a new anti-aliasing technique that isn't as harsh on your hardware.
MFAA stands for Multi-Frame Sampled AA, where NVIDIA finds more information inside of every pixel on your screen, anti-aliasing the scene much better. It has the performance hit of 2x MSAA, but the look of 4x MSAA, which will provide gamers with a better looking image, without the performance hit they're used to with traditional AA methods, like MSAA.