Geforce GTX 1080 8GB GDDR5X and GTX 1070 8GB GDDR5 (Paper) Launching On 6th May at Nvidia Editor’s Event – Market Availability by Computex 2016 in June

Usman Pirzada
Posted Apr 30, 2016
2.9kShares
Share Tweet Submit

The exact date of Nvidia Editor’s Event has finally leaked out. The event will be held on the 6th of May 2016 and will be closed to the general public, with a vetted list of editors invited. This appears to be a good time to round up everything we know about the next generation lineup so far. Pretty much everything has leaked out and only a few bits and pieces remain to be seen. One of the major things, that is still not confirmed at this point is the public nomenclature Nvidia will be using for its new lineup. Whether it adopts a new nomenclature or christens the new lineup, the Geforce GTX 1000 series, remains to be seen.

Nvidia’s next generation Pascal lineup launching on 6th May, will hit the shelves in June – beats AMD to market

Nvidia will be following an interesting release schedule for the information this time around. The Editor’s Event that Nvidia is hosting will be closed to the public eye and will basically be a briefing about the upcoming Nvidia lineup. The time line, according to the information we currently have, is as follows:

  • Our sources tell us that the Editor’s Event will only be for information dissemination. Nvidia has at least two SKUs ready for the event (Source) .

  • Reviews are expected to go live within a few weeks after that (Mid-May)

  • The cards will have a hard launch at Computex 2016 (Market Availability in June).

The new lineup of Nvidia GPUs will feature a mix of GDDR5X and GDDR5 memory (HBM2 yields are still not ideal and will debut with the full powered GP100 chip in Q1 2017 next year). All cards presented at this event are expected to exceed the minimum VR spec. We have also seen the reference and cooler shroud design that might end up being on the actual cards part of the GTX 1000 series. Without any further ado, here is what we have learned about the lineup so far:

The Geforce GTX 1080 8GB GDDR5X Graphics Card with the GP104-400-A1 GPU

Status: Confirmed
Reveal: Editor’s Event 6th May
Paper Launch: Computex 2016
Availability: June 2016

The flagship graphic card of this year, and the first GPU to utilize the brand new GDDR5X standard from Micron, the Nvidia GTX 1080 will be one impressive card. Rocking the GP104-400-A1 GPU flavor the card is expected to feature 2560 CUDA Cores (given a few assumptions about GPCs and the brand new Pascal Architecture). Given that the die size is around the 300mm² mark, we are looking at performance levels roughly faster than the GTX 980 Ti (depending on how it is clocked).

Although there have been multiple leaks about the Geforce GTX 1080 in the last few weeks, this close to the reveal, its pointless to speculate further regarding the specifications of Nvidia’s upcoming GPU. Here is what we do know for sure: Nvidia’s Pascal architecture will be manufactured on the 16nm FinFET process (which inherently brings a 2x Perf/Watt increase). GP100 focused on increasing gimped FP64 performance (that Maxwell suffered from) so the GP104 is expected to focus primarily on single precision performance (good news for games).

NVIDIA Officially Launches GeForce GTX 1060 Graphics Card at $249 US - Review Round Up of The First Card To Utilize GP106 GPU

The price range of the cards depends entirely on how Nvidia is positioning its lineup. If the GTX 1080 is officially positioned (performance wise) to replace the GTX 980 and not the GTX 980 Ti, then it should be priced somewhere around its $549 price point. On the other hand, if it is positioned (officially) to replace the 980 Ti, the price range can be similar to the MSRP of the Ti: $649 (or even above). In my opinion however, it is more likely that Nvidia will officially position the card against the GTX 980.

The Geforce GTX 1070 8GB GDDR5with the GP104-200-A1 GPU

Status: Confirmed
Reveal: Editor’s Event 6th May
Paper Launch: Computex 2016
Availability: June 2016

This particular piece of graphics technology will feature a cut down GP104 chip. Depending on the amount of Streaming Multiprocessors axed (SMs) you are looking at CUDA Cores in the range of 2048. Considering that the GTX 980 itself had 2048 CUDA Cores, you are looking at a beast of a graphics card (and probably an insane overclocker). Depending on how Nvidia plays the pricing game, this could constitute the sweet spot for value seekers in Nvidia’s 16nm FinFET lineup. We got some beautiful die shots already so we will definitely be seeing this card soon. In my opinion it will probably be priced around the $449 to $549 mark.

The Geforce GTX 1060 (Ti) 6GB GDDR5 with the GP104-150-A1 GPU

Status: Rumored
Reveal: Editor’s Event 6th May (TBC)
Availability: Fall 2016 (TBC)

Various entries for potential Pascal based graphic cards were spotted at 3D Mark a while ago. And if they are anything to go by, Pascal architecture has incredible potential. Which means the GTX 1060 Ti will be nothing to sneeze at. If clocked high enough, the card could even be powerful enough to near the VR minimum spec (GTX 970 performance levels) which will allow Nvidia to market an additional VR ready GPU. The fact that the card has 6GB worth of GDDR5 means that it features a 192 bit bus.

There has existed a gap in Nvidia’s lineup of GPUs which this card could very well fill. VR is the next frontier for GPU manufacturers to concur. With a high attach rate, and a relatively small Total Addressable Market, the race is on to see who can put the most VR capable GPUs out there. The Total Available Market for VR is currently limited by the amount of high end cards on the market right now (GTX 970/R9 290 and above).

If entry-level graphic cards like the GTX 1060 (Ti) can reach the VR minimum spec (roughly 3.5 TFLOPS) then it will expand the market potential by several factors – driving the growth of VR on PC and matching the value proposition of products like the PSVR. AMD has recently vowed to lower the price of the “minimum VR spec” and this could be Nvidia’s attempt to do the same, which would put the price point anywhere between $249 to $329.

Nvidia's Geforce GTX 1070, Pascal Based Graphics Card Has Arrived - Beats Out The TITAN-X At Just $449 For The Founder's Edition

The Geforce GTX 1060 4GB GDDR5 with the GP106 GPU

Status: Rumored
Reveal: Editor’s Event 6th May (TBC)
Availability: Fall 2016 (TBC)

NVIDIA Pascal GP106 GPUImage courtesy of VideoCardz.com

We learned about the GP106 powered GTX 1060 a few weeks ago and according to the rumor mill, the card will be launching in the fall.. The die size appears to be around 200mm² and if the GPU has two Global Processing Clusters then we can expect around 1280 CUDA cores (256 CUDA Cores per Streaming Multi Processor and 5 SMPs per GPC). The Nvidia GTX 1060 should work only on power provided by the PCIe by design but will probably have an optional 6 pin connector added by AIBs to ensure overclocking headroom. Even though the GP106 is near the bottom of the barrel of the Pascal lineup, depending on the clocks, it could still pack a lot of punch thanks to the use of 16nm FinFETs. The time has come when entry level actually means entry level and not a downright horrible gaming experience.

The product placement of this particular GPU is for budget gamers (as opposed to HTPC builds). Keep in mind that there is a further cut down version of the chip, known as the GP107/GP108. These are the variants which will only be useful for HTPC purposes and not any real gaming. Because of this, we can consider the GP106 the last piece of the Pascal lineup that is of any interest to the gaming public. While we don’t know the exact price of this particular GPU, since we are looking at the general price range of GTX 960, the GTX 1060 should be priced anywhere between $179 to $229.

NVIDIA GeForce GTX 1080, GTX 1070 and GTX 1060 "Rumored" Specs:

Graphics Card NameNVIDIA GeForce GTX 1080NVIDIA GeForce GTX 980 TiNVIDIA GeForce GTX 1070NVIDIA GeForce GTX 980NVIDIA GeForce GTX 1060 Ti
Graphics CoreGP104GM200GP104GM204GP104/106
Process Node16nm FinFET28nm16nm FinFET28nm16nm FinFET
Transistors~8 Billion8 Billion~6.4 Billion (TBC)5.2 Billion~4.8 Billion (TBC)
CUDA Cores2560 CUDA Cores (TBC)2816 CUDA Cores2048 CUDA Cores (TBC)2048 CUDA Cores1536 CUDA Cores(TBC)
VRAM8 GB GDDR5X6 GB GDDR58 GB GDDR54 GB GDDR56 GB GDDR5
Bus Interface256-bit bus384-bit bus256-bit bus256-bit bus192-bit bus
Power ConnectorSingle 8-Pin Power?8+6 Pin PowerSingle 8-Pin Power?6+6 Pin PowerSingle 8-Pin Power?
TDPTBC250WTBC165WTBC
Display Outputs3x Display Port
1x HDMI 2.0
1x DVI
3x Display Port
1x HDMI 2.0
1x DVI
3x Display Port
1x HDMI 2.0
1x DVI
3x Display Port
1x HDMI 2.0
1x DVI
3x Display Port
1x HDMI 2.0
1x DVI
Launch DateJune 2016May 2015June 2016September 2014June 2016/Fall 2016 (TBC)
*TBC = To be confirmed

Share Tweet Submit