The GTX 1080 is huge news – 2x the performance of the Titan X for half the price is no small boast. There are, understandably, a few annoyed Titan X owners out there, but it’s great news for everyone else if Nvidia’s claims ring true.
Not much is known about the lower-specification GTX 1070 card, aside from price, so I’ll mostly talk about the GTX 1080 here.
We learned that the GeForce GTX 1080 is coming to market in several shapes:
- GeForce GTX 1080 8GB
- GeForce GTX 1080 Founders Edition
- GeForce GTX 1080 Air Overclocked Edition
- GeForce GTX 1080 Liquid Cooled Edition
Stock GTX 1080 is clocked at 1.66 GHz, with Turbo Boost lifting it to 1.73 GHz. Founders Edition includes overclocking-friendly BIOS to raise the clocks to at least 2 GHz, and the presentation showed the chip running at 2.1 GHz. The main limiting factor for the overclocking beyond 2.2 GHz is 225 Watts, which is how much the board can officially pull from the power circuitry: 75 Watts from the motherboard and 150 W through 8-pin PEG connector. However, there are power supply manufacturers which provide more juice per rail, and we’ve seen single 8-pin connector delivering 225 W on its own. Still, partners such as ASUS, Colorful, EVGA, Galax, GigaByte, MSI are preparing custom boards with 2-3 8-pin connectors. According to our sources, reaching 2.5 GHz using a liquid cooling setup such as Corsair H115i or EK Waterblocks should not be too much of a hassle.
Search for performance lead the company to remove as much legacy options as possible, and you can no longer connect the GTX 1080 with an analog display. D-SUB15 is now firmly in the past, and you cannot make the connection work even if you use a 3rd party adapter. The rest of connectors include a 144Hz-capable DVI, three DisplayPort 1.4 and a single HDMI 2.0B connector.
The Architecture
Nvidia’s 2016 cards are all powered by its new Pascal technology. Pascal replaces Maxwell, which was behind the GTX 9xx series and the Titan X. Nvidia’s headline claim about Pascal as a whole is that it’s three times more efficient than the previous generation, which is done both with computing improvements and a new design and manufacturing process.
Pascal uses a 16nm (nanometer) process, where Maxwell used a 28nm process. This means Nvidia can fit more transistors (the bits that do the computing work) into a smaller space while consuming less power, meaning the whole GPU is a lot more efficient and runs cooler.
In terms of clock speed, the Nvidia GeForce GTX 1080 will sit at 1607MHz but is capable of boosting to 1733MHz, which is a huge figure when you consider the maximum boost speed of the GTX 980 was 1216MHz.
All this is possible in a GPU that consumes just 15W more power than the GTX 980 (180W versus 165W), and operates at a lower maximum temperature (94C versus 98C).
Considering how much more efficient the GTX 1080 is, expect to see the trend of top-end Nvidia cards inside laptops continue.
The Memory
There’s a huge leap here, too. GDDR5 has been repalced by GDDR5X, a new memory technology that allows the GTX 1080’s 8GB of memory to perform more than a third quicker, with a bump in bandwidth up from 7Gbps to 10Gbps. This is despite the memory’s actual width not improving from the old 980, sticking firmly with a 256-bit bus. This leaves room for a GTX 1080 Ti down the road; the GTX 980 Ti offered a wider, 384-bit bus when it was released months after the GTX 980. The GTX 1070 will continue to use GDDR5.
GDDR5X: The Magic Ingredient
When GDDR5 memory came to market, development roadmap showed two generations of product: Single-Ended and Differential GDDR5 memory. Sadly, this memory standard received little development after the launch in 2008, as the memory manufacturers such a SK.Hynix and Samsung Semiconductor refused to bring it to the market. The team behind the GDDR5 standard at AMD focused its attention on developing HBM (High Bandwidth Memory), leaving GDDR5 to manufacturers, slowly building the capacity and that was that. Year and a half ago, after seeing that HBM1 is limited in capacity and that HBM2 memory won’t be available in real volume before 2017, Nvidia started to work with Micron’s team in Germany on building the ultimate performance GDDR5.
Manufactured in 20nm process, GDDR5X memory showed being overclocking-friendly even with the initial silicon. As the roadmap shows, the target was to hit the 10 Gpbs i.e. 2.5 GHz QDR. Given that the memory actually moves four times per cycle, it should be called Quad Data Rate, but the name GDDR SGRAM (Graphics Double Data Rate Synchronous Graphics Random Access Memory) was kept for continuity.
GeForce GTX 1080 has the memory clocked at 2.5 GHz but we do expect some of the samples clocking at 2.75-3.5 GHz (11-14 Gbps). That would raise the available bandwidth from 320GB/s to 352-448 GB/s and we do expect to see extreme overclockers pushing the memory even more. If Micron adopts 10nm process for GDDR5X, we’ll get to 4 GHz clock / 16 Gbps rather sooner than later.
ASYNCHRONOUS COMPUTE
This is a big deal, and something many Nvidia fans were praying the firm would announce. They weren’t disappointed. Asynchronous compute is a rewrite of the way GPUs process graphics rendering and computing tasks, such as where objects are, and what effect on lighting it might have.
In non-asynchronous GPUs, graphics tasks and computing tasks are handled separately, meaning one has to wait for the other to finish. This is inefficient and isn’t taking full advantage of a given GPU’s power at any given moment, with many of the GPU’s cores sitting idle. Async compute lets both tasks run simultaneously (although not starting or finishing at exactly the same time), dramatically increasing efficiency. There’s lots of work to be done in this area, so it’ll be interesting to see how far Nvidia has come.
Async compute is part of DirectX 12, the standard graphics API in Windows 10, and it’s something AMD has been doing very well for quite some time. Nvidia has also been using a form of async compute for a while, but it’s significantly less effective than AMD’s integration.
Nvidia therefore needed proper async compute integration this time around in order to stay competitive, but this will be a very interesting development and when AMD launches its own Polaris-powered hardware we’ll see how far Team Red has come with its own developments.
Will It Outperform Titan X?
Apparently so. Nvidia claims the GTX 1080 will be twice as fast when performing certain VR workloads when compared to the Titan X, that cost more than £1000 (RM 5808.12) at launch. While this will leave Titan X owners weeping, it’s worth mentioning that such a performance difference will not be quite so sharp in other tasks, but because we know so little about the GTX 1080’s real-world performance it’s very hard to say. Nvidia also reckons the GTX 1080 is faster than two GTX 980s running in SLI.
Don’t worry buddy. This beast will be released on 27th May 2016. There’s still time to conserve some money. (Well, i wouldn’t call ‘some’ in my case of financial crisis. Poor me.)