The NVIDIA GeForce RTX 4070 TI VS RTX 4080: A comparison of GPUs for the game in 1440p

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Choose your graphics card between Nvidia, AMD and Intel

The graphics cards are “customized” by assemblers: the value of the frequency, the quantity of memory and the cooling system differ from one card to another.

If Nvidia often dominates the market in terms of power, AMD remains a good choice on entry and mid -range with almost powerful and more affordable cards on this segment. Intel arrival on medium -range GPUs upset the market.

In addition to the power-price ratio, some details can tip the scales towards a founder rather than another.

AMD technologies

Less prolix in on-board technologies than NVIDIA, AMD graphics cards are betting a lot on their performance-prix report.

Eyefinity

This technology manages up to six high definition screens.

AMD liquidvr

AMD liquidvr improves the compatibility of VR headsets, in particular by reducing latency between the head movement and the display of the image.

Freesync

Freesync technology synchronizes the refresh of your monitor with the flow of the graphics card to avoid the jerks and the tears of the image. Available in standard version, premium (120 Hz min in Full HD) and Premium Pro (low latency, HDR and LFC, compensation for low image frequencies)

AMD VSR

THE AMD VSR consists in increasing the resolution of a game beyond that of the screen in order to gain finesse and detail.

AMD GPU generations and architectures

Generation	Architecture name	Year	Bus Interface	Memory
Radeon RX 5000	Rdna	2019	PCI-Express 3.0 x16 / PCI-Express 4.0	GDDR6
Radeon RX 6000	RDNA 2	2020	PCI-Express 4.0 / PCI-Express 4.0 x16	GDDR6
Radeon RX 7000	RDNA 3	2022	PCI-Express 4.0 x16	GDDR6

New AMD graphics cards see AMD RX graphics cards

Radeon RX 7700 XT, PCI-Express 16x 4.0, 12 GB GDDR6

Radeon RX 7700 XT, PCI-Express 16x 4.0, 12 GB GDDR6

Radeon RX 7700 XT, PCI-Express 16x 4.0, 12 GB GDDR6

Radeon RX 7700 XT, PCI-Express 16x 4.0, 12 GB GDDR6

Radeon RX 7700 XT, PCI-Express 16x 4.0, 12 GB GDDR6

Radeon RX 6800 XT, PCI-Express 16x 4.0, GDDR6

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Radeon RX 7600, PCI-Express 4.0, 8 GB GDDR6

Radeon RX 550, PCI-Express 3.0, 2 GB GDDR5

Radeon RX 7900 XTX, PCI-Express 16x 4.0, 24 GB GDDR6

Radeon RX 7900 XT, PCI-Express 16x 4.0, 20 GB GDDR6

Radeon RX 7900 XT, PCI-Express 16x 4.0, 20 GB GDDR6

Nvidia technologies

Nvidia has woven her canvas made of ads from technologies, each more sympathetic than the other, even if they are not always very widespread. Here is an overview of technos specific to NVIDIA graphics cards:

RTX (Ray Tracing Texel Extreme)

And there was light ! Or rather, with the Raytracing implemented by this technology, lighting and light games in games and applications are calculated in real time. Despite the very energy -consuming side, the rendering is superb. Ray-tracing is however not specific to Nvidia and is included in DirectX 12, nothing prevents AMD from supporting it.

DLSS (Deep Learning Super Sampling)

Second big innovation of the NVIDIA RTX generation, this “super-sampling” technology makes it possible to surpass the usual anti-aliasing solutions thanks to the AI ​​which learns with supercomputer to build the best possible rendering.

V-Sync adaptive

Vertical synchronization (V-Sync) eliminates the defects of superposition of images (called “tearing”), that is when the number of images sent by the graphics card exceeds the display capacity of the screen ( or 60 images in 60 Hz). This technology sharply drops the perfs when the graphics card cannot maintain 60 fps. The V-Sync adaptive comes to activate it only if necessary.

GPU Boost

THE GPU Boost automatically increases the clock frequencies to give a boost to the graphics card when the energy load and the temperature allow it.

Physxx

Physical engine improving the visual rendering of compatible games thanks to the display of many collision or particles effects.

CUDA (Compute Unified Device Architecture)

THE Cuda is a technology specific to Nvidia, the graphics card processor performs operations initially provided for the motherboard processor. The PC thus gains in speed.

Shadowplay

THE Shadowplay Record and share your game sessions on the web.

G-sync

Technology G-sync Synchronizes the refresh of your monitor with the flow of the graphics card to avoid jerks and tears of the image.

Generation of NVIDIA graphics cards

Generation	Architecture name	Year	Bus Interface	Memory	Associated technologies
GeForce 16 Series	Turing	2019	PCIe 3.0 x16	GDDR5 / GDDR6	–
GeForce 20 Series	Turing	2018	PCIe 3.0 x16	GDDR6	–
GeForce 30 Series	Ampere	2020	PCIe 4.0 x8 / PCIe 4.0 x16	GDDR6 / GDDR6X	Tensor Core 3rd Gen / RT Core 2e Gen
GeForce 40 Series	Ada Lovelace	2022	PCIe 4.0 x16	GDDR6X	Tensor Core 4th Gen / RT Core 3rd Gen

Assemblers overclock AMD and NVIDIA GPUs, thus improving performance. Know that if you decide to overclock yourself, and although it is easy and not very dangerous on certain models, you lose the warranty on your product. They also work on cooling and make it the main sale argument. In this case, no fear to have, your card will not heat more than reason.

New Nvidia graphics cards see NVIDIA RTX graphics cards

GeForce GTX 1650, PCI-Express 16x, 4 GB GDDR6

The NVIDIA GeForce RTX 4070 TI VS RTX 4080: A comparison of GPUs for the game in 1440p

The NVIDIA GeForce RTX 4070 TI VS RTX 4080: A comparison of GPUs for the game in 1440p

• by
• In News
• On August 12, 2023

The NVIDIA GeForce RTX 4070 TI and the RTX 4080 are highly sought -after GPUs for the game in 1440p and 4K. With a respective price of $ 799 and $ 1,199, they are considered to belong to the average price category. This article aims to compare the performance of these two GPUs through various high resolution game tests, including Ray-Tracing tests.

The specifications of these GPUs are as follows:
– GPU die: AD104 for RTX 4070 Ti WindForce OC, AD103 for RTX 4080 Fe, AD104 for RTX 4070 Fe
– SM: 607 for the RTX 4070 TI, 646 for the RTX 4080 and 646 for the RTX 4070 Fe
– Cores Cuda: 7,680 for the RTX 4070 Ti, 9,728 for the RTX 4080 and 5,888 for the RTX 4070 Fe
– Cores RT: 607 for the RTX 4070 TI, 646 for the RTX 4080 and 646 for the RTX 4070 Fe
– Cores Tensor: 240 for the RTX 4070 TI, 304 for the RTX 4080 and 184 for the RTX 4070 Fe
– Basic frequency: 2,310 MHz for RTX 4070 TI, 2,205 MHz for RTX 4080 and 1,920 MHz for RTX 4070 Fe
– Boost frequency: 2,640 MHz for RTX 4070 Ti, 2,508 MHz for RTX 4080 and 2,475 MHz for RTX 4070 Fe
– Memory: 12 GB GDDR6X for RTX 4070 TI, 16 GB GDDR6X for RTX 4080 and 12 GB GDDR6X for RTX 4070 Fe
– TBP (Thermal Design Power): 285W for the RTX 4070 TI, 320W for the RTX 4080 and 200W for the RTX 4070 Fe
– Cache L2: 48 MB for RTX 4070 Ti, 64 MB for RTX 4080 and 36 MB for RTX 4070 Fe
– Recommended sale price: $ 799 for RTX 4070 TI, $ 1,199 for RTX 4080 and $ 599 for RTX 4070 Fe

Benchmarks were made using the following specifications:
– Motherboard: Asus Z790 Maximus Hero
– cooler: Lian Li GLD 360
-CPU: Intel Core i9-13900k
– Memory: 16 GB X2 DDR5 6 000 mt/s cl 38

Benchmarks focused on medium FPS, percentile of 1% and the lowest frequency average. The average FPS represent overall performance, while the percentile of 1% and the average frequency average provides information on the lower frequencies of 1%.

Regarding the game in 1440p, the GeForce RTX 4080 generally offers better performance than RTX 4070 TI on different titles. For example, in games like “A Plague Tale” and “Cyberpunk 2077”, the RTX 4080 has an improvement of almost 30% compared to the RTX 4070 Ti. However, the performance gap is reduced in titles like “Assassins” Creed Valhalla “and” Shadow of the Tomb Raider “, with differences ranging from 10% to 18%.

Overall, the RTX 4080 offers better game performances in 1440p compared to the RTX 4070 Ti. However, users must take into account their individual needs and their budget before making a purchasing decision.