KurabeDojoThe comparison dojo

Compare · 2 products compared

NVIDIA GeForce RTX 5080vsNVIDIA GeForce RTX 4070 Super

The NVIDIA GeForce RTX 5080 and NVIDIA GeForce RTX 4070 Super are powerful desktop graphics cards built for high-end rendering and gaming. While they share matching PCIe lane counts and display output slot configurations, they differ sharply in architectural generation, raw compute power, and price.

Report a problem

Intro

What is the problem?

eBay prices refresh hourly — check the listing before you buy. We may earn a commission.

Report a problem

Best for

What is the problem?
Which products? Tick all that apply, or none if you are not sure.

Biggest differences

  • Memory bandwidth

    NVIDIA GeForce RTX 5080 leads

    +90%

    higher is better

    • NVIDIA GeForce RTX 5080960 GB/s
    • NVIDIA GeForce RTX 4070 Super504 GB/s

  • Launch price

    NVIDIA GeForce RTX 4070 Super leads

    −40%

    lower is better

    • NVIDIA GeForce RTX 5080999 USD
    • NVIDIA GeForce RTX 4070 Super599 USD

  • Board power

    NVIDIA GeForce RTX 4070 Super leads

    −39%

    lower is better

    • NVIDIA GeForce RTX 5080360 W
    • NVIDIA GeForce RTX 4070 Super220 W

  • Shader cores

    NVIDIA GeForce RTX 5080 leads

    +50%

    higher is better

    • NVIDIA GeForce RTX 508010,752 cores
    • NVIDIA GeForce RTX 4070 Super7,168 cores

Where each one wins

  • NVIDIA GeForce RTX 5080

    • Shader cores
    • Boost clock
    • Base clock
    • FP32 compute
    • Video memory
    • Memory type
    • Memory bus width
    • Memory bandwidth
    • Memory speed
    • Last-level cache
    • PCIe version
    • HDMI version
    • DisplayPort version

  • NVIDIA GeForce RTX 4070 Super

    • Launch price
    • Board power
    • Recommended PSU

What both have

  • Power connectors12V-2x6
  • PCIe lanes16 lanes
  • HDMI outputs1 ports
  • DisplayPort outputs3 ports
  • Max displays4 displays

Every spec

Red marks the better value in each row.

SpecNVIDIA GeForce RTX 5080NVIDIA GeForce RTX 4070 Super
Overview

Report a problem

Overview specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
Release date2025-012024-01
Launch price999 USD599 USD
Graphics chip

Report a problem

Graphics chip specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
ArchitectureBlackwellAda lovelace
Shader cores10,752 cores7,168 cores
Boost clock2,620 MHz2,475 MHz
Base clock2,300 MHz1,980 MHz
FP32 compute56.3 TFLOPS35.5 TFLOPS
Memory

Report a problem

Memory specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
Video memory16 GB12 GB
Memory typeGddr7Gddr6x
Memory bus width256 bit192 bit
Memory bandwidth960 GB/s504 GB/s
Memory speed30 Gbps21 Gbps
Last-level cache64 MB48 MB
Power

Report a problem

Power specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
Board power360 W220 W
Recommended PSU850 W1600 W
Power connectors12V-2x612V-2x62
Interface

Report a problem

Interface specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
PCIe version5.04.0
PCIe lanes16 lanes16 lanes
Display outputs

Report a problem

Display outputs specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
HDMI version2.1b2.1
DisplayPort version2.1b31.4a
HDMI outputs1 ports1 ports
DisplayPort outputs3 ports3 ports
Max displays4 displays4 displays
Video engine

Report a problem

Video engine specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
Hardware encodeH.264, HEVC, AV1AV1
Video encoders2 encodersNot published
Size

Report a problem

Size specs

What is the problem?
Which specs? Tick all that apply, or none if you are not sure.
Which products? Tick all that apply, or none if you are not sure.
Card length304 mmNot published
Card height137 mmNot published
Slot width2 slots4Not published
Notes on 4 values — test conditions and caveats
  1. 1NVIDIA GeForce RTX 5080, Recommended PSU: NVIDIA's minimum for a PC with a Ryzen 9 9950X; it recommends a PCIe CEM 5.1 compliant PSU.
  2. 2NVIDIA GeForce RTX 4070 Super, Power connectors: Nvidia's 16-pin header is the 12V-2x6 variant; an adapter cable for non-ATX 3.0 power supplies ships with the card.
  3. 3NVIDIA GeForce RTX 5080, DisplayPort version: DisplayPort 2.1b at UHBR20 on all three DP outputs.
  4. 4NVIDIA GeForce RTX 5080, Slot width: The card is 2-slot (40 mm thick), but NVIDIA's install guide asks for 3-slot (61 mm) case clearance.

Performance and Architecture

The RTX 5080 relies on the Blackwell architecture to deliver 56.3 TFLOPS of FP32 compute capability. By comparison, the RTX 4070 Super utilizes the older Ada Lovelace architecture to reach 35.5 TFLOPS.

Driving that generation gap are 10,752 shader cores and higher base and boost clocks on the RTX 5080, dwarfing the 7,168 cores found on the RTX 4070 Super. For heavy rendering pipelines and demanding frame rates, the RTX 5080 pushes significantly more raw data.

Report a problem

Performance and Architecture

What is the problem?

Memory and Bandwidth

Equipped with 16 GB of GDDR7 video memory on a 256-bit bus, the RTX 5080 achieves a memory bandwidth of 960 GB/s at 30 Gbps. It also features a larger 64 MB last-level cache to streamline data delivery.

Meanwhile, the RTX 4070 Super relies on 12 GB of GDDR6X memory across a 192-bit bus, delivering 504 GB/s at 21 Gbps alongside a 48 MB cache. The faster memory type and wider bus on the newer card ensure high-resolution textures load with minimal bottlenecking.

Report a problem

Memory and Bandwidth

What is the problem?

Power and System Requirements

Raw performance requires a heavier electrical investment, with the RTX 5080 drawing 360 W of board power. The RTX 4070 Super operates much more efficiently, consuming only 220 W.

To support the flagship card, builders will need an 850 W power supply featuring a PCIe CEM 5.1 compliant unit, whereas the RTX 4070 Super requires a modest 600 W power supply. Both options utilize the 12V-2x6 power connector format.

Report a problem

Power and System Requirements

What is the problem?

Connectivity and Video Engine

For display outputs, the RTX 5080 upgrades to PCIe version 5.0 and includes HDMI 2.1b alongside DisplayPort 2.1b supporting UHBR20 across all three outputs. The RTX 4070 Super is constrained to PCIe version 4.0, HDMI 2.1, and DisplayPort 1.4a.

Creators working with video codecs will benefit from the RTX 5080 hardware encode capabilities supporting H.264, HEVC, and AV1 across two dedicated video encoders. The RTX 4070 Super restricts its published hardware encode support exclusively to AV1.

Report a problem

Connectivity and Video Engine

What is the problem?

The Verdict

Buy the NVIDIA GeForce RTX 5080 if you want bleeding-edge Blackwell architecture, GDDR7 memory speeds, and maximum rendering performance. Buy the NVIDIA GeForce RTX 4070 Super if you want a more affordable entry point with lower power consumption.

Report a problem

Verdict

What is the problem?

Where to buy

FAQ

What are the launch prices for these graphics cards?

The RTX 5080 launched at $999, while the RTX 4070 Super arrived at a lower launch price of $599.

What physical dimensions does the RTX 5080 require for installation?

The RTX 5080 measures 304 mm in length, 137 mm in height, and is built as a 2-slot card, though NVIDIA's installation guide asks for 3-slot case clearance.

When were these graphics cards released?

The RTX 5080 was released in January 2025, preceding the RTX 4070 Super which debuted in January 2024.

Report a problem

FAQ

What is the problem?

Spotted something wrong? Tap the next to any section or spec group to let us know. Every report is read.