Nvidia GTX 1080 Benchmarks & Review Roundup - 25% Faster Than GTX 980 Ti, Launching May 27

Independent reviews and comprehensive performance benchmarks of Nvidia'due south GeForce GTX 1080 graphics card have just hit the web. That's right folks, the twenty-four hours that every prospective GTX 1080 buyer has been waiting for has arrived. Nvidia's embargo on GTX 1080 reviews has just lifted and we finally have the opportunity to closely examine and get a good look at what Nvidia'south new graphics bill of fare can really do.

We know that some reviewers received their GTX 1080 samples nearly x days ago, and most take had more a week to run the card through the paces. This gave room to a lot of very interesting and some even creative testing. Although we should not neglect to remind everyone that while Nvidia is allowing reviews to go up today, the Founder'southward Edition - reference blueprint - GTX 1080 volition non be available for buy today only you'll be able to grab it for $699 on the 27th of May. It should exist mentioned that the May 27 launch date that Nvidia has prepare is strictly for the Founder's Edition GTX 1080.

Custom GTX 1080 graphics cards from Nvidia's partners will be available before long after and those volition carry an official MSRP of $599, exactly $100 less than the Founder'south Edition. So you'll definitely want to hang on for a few days before pulling the trigger. Specially because that Nvidia's partners are rolling out GTX 1080 cards that characteristic improve cooling, better ability delivery and offer improve operation via higher clock speeds right out of the box.

Nvidia GeForce GTX 1080 Benchmarked And Reviewed - Up To 25% Faster Than GTX 980 Ti At 2560x1440

Now without whatever further delay, allow's become to the juicy bits! Below you lot'll find a list of all GTX 1080 reviews that have been published to date. We'll update the listing as more reviews become available.

Anandtech
ComputerBase
PCPer
PCGamesHardware
HardOCP
KitGuru
Hexus
TechPowerUp
Guru3D
PCWorld
HotHardware
OverclockersClub
ArsTechnica

GTX 1080 ComputerBase 4K Credit : ComputerBase - GTX 1080 Max = 100% Fan Speed

Anandtech'south take :

Ryan Smith - Anandtech
Translating this into numbers, at 4K we're looking at 30% performance gain versus the GTX 980 and a 70% performance gain over the GTX 980, then we're looking at a very significant jump in efficiency and performance over the Maxwell generation.

Power consumption, temperatures & racket via Anandtech

Official Geforce GTX 1080 and Geforce GTX 1070 Specifications

WCCFTech	Nvidia Geforce GTX 1080	Nvidia Geforce GTX 1070
Compages	Pascal	Pascal
Transistors	seven.2 Billion	7.two Billion
CUDA Cores	2560	1920
Core Clock	1607 Mhz	TBA
Heave Clock	1733 Mhz	1683 Mhz
Memory Blazon	G5X (GDDR5X)	GDDR5
Retention Speed	10 Gbps	viii Gbps
Memory Configuration	8GB	8GB
Bus Width	256-chip	256-bit
Memory Bandwidth	320 GB/southward	256 GB/s
Multi Project	Yep	Yep
HB SLI Bridge Back up	Yep	Yeah
Nvidia GPU Heave	iii.0	3.0
DirectX 12 Feature Level	12_1	12_1
OpenGL	4.5	4.five
Vulkan API	Yes	Yes
Maximum Digital Resolution	7680x4320@60Hz	7680x4320@60Hz
Brandish Connectors	DP 1.42, HDMI 2.0b, DL-DVI	DP 1.42, HDMI 2.0b, DL-DVI
HDCP	ii.2	2.two
Power Draw	180W	150W
Power Connector	Unmarried 8-Pin	Single 8-Pivot
Maximum Operating Temp	94 C	94 C
Partner Price (MSRP)	$599	$379
FE Price (MSRP)	$699	$449

The GTX 1080 Dissected - Diving Into Nvidia's GP104 Pascal Architecture

Historically speaking both graphics vendors Nvidia and AMD take successfully maintained a cadence that insured significant performance and ability efficiency advances are delivered with every new compages and Pascal is no exception. The very basic edifice block of Nvidia's Pascal architecture has been stripped apart and redesigned. This building block which Nvidia dubs Streaming Multiprocessor or SM for short is the engine that drives the graphics and compute horsepower of every Pascal flake.

With Maxwell, Pascal's predecessor powering the GTX 900 series, Nvidia introduced the Streaming Maxwell Multiprocessor. The SMM congenital on the strengths of Nvidia's Kepler SM - introduced with the GTX 600 and 700 series - which Nvidia dubs the SMX. It as well done away with many unnecessary complexities which enabled the engine to evangelize more than throughput and higher clock speeds. The Pascal SM in its own right is an development of the Maxwell SM, a smarter, more streamlined engine.

Nvidia GTX 1080 GP104 Block Nvidia GTX 1080 - GP104 Block DIagram

Inside Nvidia's full GP104 powering the GTX 1080 - a cut back version of the GP104 GPU volition be leveraged in the GTX 1070 - there are four Graphics Processing Clusters or GPCs for short. Each GPC consists of five Streaming Multiprocessors or SMs for brusk - each SM contains 128 CUDA cores - and sixteen Texture Mapping Units or TMUs for short. Each SM includes eight Render Output Units, ROPs. In turn GP104 houses 2560 CUDA cores, 160 TMUs and 64 ROPs. Finally the engine is connected via eight 32-bit retentivity segments - 256bit memory controller - to 8GB of GDDR5X retentivity.

As mentioned, each GP104 streaming multiprocessor includes 128 FP32 CUDA cores, the same as Maxwell. Within each GP104 streaming multiprocessor there are four 32 CUDA core partitions, four dispatch units, two warp schedulers and a fairly large instruction buffer. Twice as large compared to Maxwell.

gp100_SM_diagram The GP100 Pascal Streaming Multiprocessor

This arrangement is almost identical to what we've seen with the much larger 3840 CUDA cadre GP100 GPU that'south powering the Tesla P100 and very likely to power Nvidia'due south next Titan graphics card. Only inside GP100 each SM contains exactly one-half the number of CUDA cores, dispatch units and warp schedulers vs GP104. Just in turn there are twice every bit many SMs per GPC.
And so the merely difference with GP104 is that Nvidia is group 64 CUDA core SMs in pairs made of 128 CUDA cores each and in turn naming the larger 128 unit of measurement an SM instead. This is all while maintaining the exact same ratio of dispatch units, warp schedulers and instruction buffers per CUDA core that we've seen with GP100.

So recall of information technology as Nvidia just pairing 64 CUDA core groups together in a single SM. This decision is probable influenced by the significant reduction of FP64, double precision, CUDA cores per SM inside GP104 vs GP100. GP104 simply contains ane FP64 CUDA core for every 32 FP32 CUDA cores. While GP100 has 1 FP64 CUDA core for every 2 FP32 CUDA cores, 16 times more than GP104.

Additionally because each GP104 SM has twice the number of registers as Maxwell, this in turn means that non simply can Pascal accommodate more threads than Maxwell, but each thread has access to more registers and thus a lot more than throughput. Finally, each warp scheduler tin can dispatch ii instructions per clock.

Nvidia'due south Senior Architect, Lars Nyland admitted that the 16nm FinFET process played an of import role in realizing the squad's power efficiency goals for Pascal, but maintains that numerous architectural improvements aided in further reducing the energy footprint of the architecture. Including the apply of new on-chip voltage signaling technologies.