Welcome Stranger to OCC!Login | Register
Reviews » Video Cards

NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Review

   -   
Price: RTX 2080 FE $799, RTX 2080 Ti FE $1199
Buy Now

NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Introduction:

Technology stops for no man or woman. It steadily marches on dragging us along sometime willingly and well sometimes kicking and screaming. One of the biggest improvements you can make to your system is to upgrade your video card. It alone can make games more enjoyable and ultimately drives us down a rabbit hole as we get more and more engaged in our game of choice. Be it first person shooters or the newest iteration of your favorite RPG. The graphics card is the means with which we enjoy or curse the game. The rush to give gamers the best visual quality comes at an ever rising hardware expense. With Real time Ray tracing becoming the next big thing NVIDIA is delivering hardware that is capable of moving the needle further to a more realistic experience.

Now we welcome the next architecture from NVIDIA named Turing. Today we will look at the first two cards in the product stack from NVIDIA the TU102 based  RTX 2080 Ti Founders Edition and the TU104 based  RTX 2080 Founders Edition.  Equipped with all new with new tech including high speed GDDR6 memory, Ray Tracing and Tensor cores along with a new look and cooling solutions the perfomence can only scale upwards As of today these Founders Edition cards are available from NVIDIA;s own website for $799 and $1199 respectively. A  pretty strong bump in price over the Pascal Based GTX 1080 Ti. Lets see if the performance and feature set make these cards worth the price.

 

NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Closer Look:

This new generation is based on an all new core architecture called Turing. Pushing for a future with real time Ray Tracing the hardware had to built to put the parts in consumer hands to create the demand. Build it and they shall come as the catch phrase goes, This new architecture delivers an up to 50% boost in delivered performance per CUDA core. To gain this kind of performance required significant architectural changes.

The first change is the Turing SM adds a new "independent integer datapath that can execute instructions concurrently with the floating-point math datapath. In previous generations, executing these instructions would have blocked floating-point instructions from issuing. Second, the SM memory path has been redesigned to unify shared memory, texture caching, and memory load caching into one unit."   NVIDIA's "TU102 GPU includes six Graphics Processing Clusters, 36 Texture Processing Clusters, and 72 Streaming Multiprocessors. Each GPC includes a dedicated raster engine and six TPCs, with each TPC including two SMs. Each SM contains 64 CUDA Cores, eight Tensor Cores, a 256 KB register file, four texture units, and 96 KB of L1/shared memory which can be configured for various capacities depending on the compute or graphics workloads. Ray tracing acceleration is performed by a new RT Core processing engine within each SM. Each memory controller includes eight ROP units and 512 KB of L2 cache.

The full TU102 GPU consists of 96 ROP units and 6144 KB of L2 cache." The TU104 GPU includes six Graphics Processing Clusters, 23 Texture Processing Clusters, and  46 Streaming Multiprocessors. Each GPC includes a dedicated raster engine and six TPCs, with each TPC including two SMs. Each SM contains 64 CUDA Cores, eight Tensor Cores, a 256 KB register file, four texture units, and 96 KB of L1/shared memory like the Tu102 core.  On the TU102 based RTX 2080 Ti FE this equates to 4352 Cuda Cores, 544 Tensor Cores, 68 RT cores.and 11GB of all new high speed GDDR6 memory. On the TU104 based RTX 2080  FE this equates to 2944 Cuda Cores, 368 Tensor Cores, 46 RT cores.and 11GB of all new high speed GDDR6 memory.

From the factory the RTX 2080 Ti FE runs at a factory Boost clock of 1635 Mhz although based on the cooling performance of the card it may run even higher. The TU104 based RTX 2080 FE comes from the factory with a boost clock of 1800Mhz and like the RTX 2080 Ti FE it will run higher as long as the thermal and power commitments are not over run.

As the first cards to utilize GDDR6 memory NVIDIA built a robust memory subsystem that is 20% more efficient than the design used on Pascal. The 11GB of GDDR6 high speed 14Gbps memory on the RTX 2080 Ti FE goes through 11 32 bit(352 bit)  memory contrllers to deliver the increased memory bandwidth. On the RTX 2080 FE a 256 bit ( 8 x 32 bit)  memory subsystem is used to manage the 8GB of 14Gbps GDDR6 memory.

 

 

The NVIDIA GeForce RTX 2080 Ti Founders Edition and RTX 2080 Founders Edition cards come in the ready for retail package. The NVIDIA green lets you know that you are going to be playing for the green team. There are subtle differences in the graphics on the foam lined box between the two cards but the messaging is clear that this is an RTX card. A dense foam is used inside the robust box to ensure the card makes it to you undamaged. The cards sit vertically in the box with the accessories in a small bocx behind the card.

 

 

 

The slim accessory bundle is in a box that features a pull tab to assist with opening the box due to its shape. Inside are a support guide and a quick start guide in case you have any troubles getting the cards installed or running. A DisplayPort to DVI adapter is included since the DVI port has gone the way of the Dodo for this generation.

 

 

The RTX 2080 and RTX 2080 Ti Founders Edition cards sport a much different look when compared to the previous generation cards. First and foremost the industrial look that has been in place since the GTX 690 is still going strong. Taking a page out of the aftermarkets book NVIDIA has ditched the radial fan pushing airflow through the heatsink shroud and out the chassis in favor of a dual axial fan design that pushes the airflow both out the chassis and into the chassis airstream. As cases have advanced with more than enough airflow flowing through them pumping heated air into the chassis has become less of a concern.

The front view of each card is dominated by the pair of 100mm 13 blade fans powered by a three-phase motor. The center of the aluminum wraparound shroud is black plastic for effect. Each card is identified on this part of the shroud. Each card uses a thin aluminum backplate to handle any heat on from the components on the back side of the card.

Measuring 10.5 inches in length this pair of RTX cards use the same dual slot form factor than previous generation cards have been using so concerns about chassis fitment are not a concern. This card is built to be used in motherboards equipped with at least a single 16x PCIe slot. The backplates on each card allow for quick identification of which model card you are running.

 

 

 

The side view of the cards show that the two slot vapor chamber based cooling solution covers the entirety of the PCB. Much like past cards from NVIDIA the GEFORCE logo is built to light up when the system is powered on. The tried and true GTX logo gets retired now that this series is built for Real time Ray Tracing and is replaced with the namesake RTX logo.

 

 

Display connectivity sees some changes on the RTX cards. Gone are the DVI ports now replaced with a more modern combination that includes a trio of DisplayPort 1.4a outputs that can support displays up to 8K resolutions at 60 Hz whioe using a single cable (with DSC 1.2), A single HDMI 2.0b connector with HDCP 2.2 support, and last but not least a single VirtualLink (USB Type-C™) connector built to drive the next generation of Virtual Reality headsets. Gone also is the shiny I/O plate now replaced with a matte black vented piece that better fits the themeing of the cards. Of course NVIDIA has the logo posted prominently right above the full length vents. The back side is closed off with the wrqp around shroud over the cooling solution. It is a multi piece desgn that looks seemless .

 

 

Along the top ogf the cards are the PCIe PEG connections that supply additional power to the PCB in excess of the 75 watts from the PCIe slot. The RTX 2080 Founders Edition has a rated TDP of 225 watts or 45 watts more than the GTX 1080 it is supplanting in the product stack. This added power draw means the card sees the power supply configuration change from a single 8 pin PEG connection to 6 pin plus 8 pin PEG connection to supply up to 300 watts to the card when you include the 75 watts from the motherboard. The RTX 2080 Ti FE gets bump in the TDP from 250 watts on the GTX 1080 Ti to a dual 8 pin PEG Connection configuration that can supply up to 375 watts to the card. SLI is supported on this generation with up to two cards supported. As we saw with the Pascal architecture two way configurations were the preffered path going forward. 3 and 4 card configurations were supported but required some work with NVIDIA to get the support. Hidden under the spiffy cover are the NVLink connections that transfer data between cards when running an SLI setup. TU102 based cards( RTX 2080 Ti) use a pair of x8 second-generation NVLink links while TU104 cards use just a sinlgle X8 NVLink link. Each link provides 25 GB/sec peak bandwidth per direction between each of the GPUs (50 GB/sec bidirectional bandwidth). The two links in TU102 provides 50 GB/sec in each direction, or 100 GB/sec bidirectionally. This configuration when used with the NVLink caple bridges support display configurations up to 4K Surround at 144hz on a pair of RTX 2080 cards and up to 8K surround when sing the TU102 based GTX 2080 Ti.

 

 

 

Pulling the shroud and cooling solution off this generation is a matter of removing a ton of fasteners so I'll let NVIDIAs images paint the picture here. For the RTX cards we have a ton of hardware on the die so the cooling solution had to be more than capable of delivering not only good cooling but great cooling to allow the cards to maximize the performance curve of the Turing cores. To do so NVIDIA used a full length Vapor Chamber based cooling solution on the RTX 2080 FE and RTX 2080 Ti FE. The cooler itself is coated in black to stick with the industrial theme. NVIDIA built a more power efficient card with an excellent power circuit on the Pascal cards and upped their game on the RTX series of cards. The TU102 RTX 2080 TI FE gets an all-new 13-phase iMON DrMOS power supply to supply the power needed for this massive chip. This iMON power supply gets an improved dynamic power management system that utilizes sub-millisecond current monitoring and control to allow for a more rigid control of the flow of power feeding the GPU. By more tightly controlling the power grid you get more power headroom to boost your overclocking adventures. By using the IMON power supply NVIDIA gives us an extremely efficient clean power circuit that wastes less power that usually shows up as added thermal load. The innovative feature found on the iMON power supply is its ability to dynamically switch to fewer power phases when runing at idle or low load situations and manage the switch to running at the full 13 phases under graphically intensive loads. The TU104 based RTX 2080 Founders Edition has less hardware under the hood so consequently uses lower phase count than the RTX 2080 Ti FE with an 8 phase iMON DrMOS power circuit.The 14 Gbps GDDR6 memory circuits are powered by a 3Phase circuit on the RTX 2080 Ti and a 2 phase circuit on the GYX 2080. If this circuitry runs as good as the circuits on the GTX 1080 and GTX 1080 Ti FE cards then we should see some serious overclocking margins.

 

 

 

Right out of the gate this is the opening salvo of the next technological wars to deliver a better visual experience. While there are not games today that take advantage of these technologies there is a list of 25 games that will support DLSS or Deep Learning Super Sampling. The short version from NVIDIA on this tech is as follows,"DLSS leverages a deep neural network to extract multidimensional features of the rendered scene and intelligently combine details from multiple frames to construct a high-quality final image. This allows Turing GPUs to use half the samples for rendering and use AI to fill in information to create the final image. The result is a clear, crisp image with similar quality as traditional rendering (which typically relies on TAA in most of today’s latest games), but with higher performance. DLSS is an easy integration for developers, so far developers have announced that 25 games will have DLSS support.

Buy Now from Amazon


  1. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition: Introduction & Closer Look
  2. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition: Specifications & Features
  3. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: Setup & Overclocking
  4. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: Shadows of the Tomb Raider, Battlefield 1
  5. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: Ghost Recon Wildlands, Far Cry 5
  6. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: The Witcher III, Ashes of the Singularity Escalation
  7. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: 3Dmark, VR Mark
  8. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition Testing: Temperature & Power Consumption
  9. NVIDIA GeForce RTX 2080 Ti & RTX 2080 Founders Edition: Conclusion
Related Products
Random Pic
© 2001-2018 Overclockers Club ® Privacy Policy
Elapsed: 0.2008359432   (xlweb1)