RTX 2080 vs Arc A750: Which Graphics Card Should You Buy?

The RTX 2080 is the faster and more consistent card, roughly 10-20% ahead of the Arc A750 on average, so it is the better pick when the two cost about the same, especially on older PCs or for libraries full of DirectX 11 games. The Arc A750 is the right choice only when it is noticeably cheaper, your system supports Resizable BAR, and you mostly play modern DirectX 12 and Vulkan titles where Intel’s drivers are strongest.

The RTX 2080 launched in 2018 as a $699 high-end Turing card, while the Arc A750 launched in 2022 at $289 as Intel’s midrange contender. Both carry 8 GB of GDDR6 on a 256-bit bus, and both now trade in the budget-to-midrange used market, which is why buyers compare them.

As an Amazon Associate we earn from qualifying purchases at no extra cost to you.

#1 Pick

msi Gaming GeForce RTX 2080 8GB GDRR6 256-bit HDMI/DP/USB Ray Tracing Turing Architecture Graphics Card (RTX 2080 Ventus 8G)

Check Price on Amazon →

#2 Pick

ASRock Intel Arc A750 Challenger D 8GB OC 2200 MHz 256-bit GDDR6 DisplayPort 16 Gbps HDMI 7680 x 4320 Graphics Card Overclocked

Check Price on Amazon →

#3 Pick

Intel Arc A750 Limited Edition 8GB PCI Express 4.0 Graphics Card

Check Price on Amazon →

Spec Comparison

The cards are surprisingly similar in their physical requirements: nearly identical power ratings and the same 6-pin plus 8-pin connector layout. The A750 has more memory bandwidth and a much larger L2 cache, but shader counts across NVIDIA and Intel are not directly comparable, and in games the Turing card’s maturity usually wins out.

How They Compare in Games

TechPowerUp’s Arc A750 review placed the A750 at roughly the level of the RTX 2070, RX 5700 XT and RX 6600 in its test suite. The RTX 2080 sits a full step above the RTX 2070 in the same outlet’s data, typically around 15-20% faster. Intel’s driver improvements since launch have helped the A750 in newer DirectX 12 games, so a fair range today is that the RTX 2080 is roughly 10-20% faster on average, with the gap smallest in modern DirectX 12 titles at 1440p and largest in older DirectX 11 games.

Approximate expectations, which vary with settings, CPU and game patches:

• 1080p high settings: the RTX 2080 typically delivers around 70-100 FPS in demanding modern games, with the A750 often 10-15 frames behind.
• 1440p: the RTX 2080 can hold around 50-75 FPS at high settings in many games; the A750 often needs medium settings or upscaling to stay near 60 FPS.
• Older DirectX 9 and 11 games: the RTX 2080’s lead is larger and more consistent, and Arc cards are more prone to frame pacing issues in some older titles.

Resizable BAR is non-negotiable for Arc

Intel lists Resizable BAR as required for the expected Arc experience, and reviewers found large drops in performance and smoothness without it. In practice, that means an Intel 10th-gen or AMD Ryzen 3000-series or newer platform with ReBAR enabled in the BIOS. On older systems, the A750 can fall well behind, which makes the RTX 2080 the obvious choice for upgrading a pre-2019 PC.

Ray Tracing and Upscaling Features

Both cards have hardware ray tracing. Intel’s ray tracing units are relatively capable for the A750’s class, so the gap narrows in ray-traced games compared with raster. Still, neither card has enough performance for heavy ray-traced settings at 1440p, and 8 GB of VRAM limits ray tracing in newer games because it increases memory use.

For upscaling, the RTX 2080 supports DLSS super resolution, including the transformer-based model introduced with DLSS 4, which NVIDIA made available on RTX 20-series cards. It does not support DLSS frame generation, which requires RTX 40 or newer. The A750 supports XeSS running on its XMX engines, which provides better XeSS image quality than on other vendors’ cards, and XeSS 2 frame generation in games that implement it. Both can use AMD FSR 3.1 upscaling and frame generation where available. DLSS remains supported in far more games than XeSS, which is a practical advantage for the RTX 2080.

PSU, Power and Case Considerations

Power needs are close. Using GPU power plus CPU power plus about 150 W of headroom with a 105 W CPU:

• RTX 2080: 215 + 105 + 150 = about 470 W.
• Arc A750: 225 + 105 + 150 = about 480 W.

A quality 550-600 W power supply with a 6-pin and an 8-pin PCIe connector covers either card. One difference is idle power: Arc A-series cards can draw considerably more at idle unless ASPM power management is enabled in the BIOS and Windows, which matters for a PC left on for long periods.

The RTX 2080 reference design is 267 mm long and two slots thick, while A750 length depends on the model. Both are standard dual-slot cards, so fit in mid-tower cases is rarely a problem, but compact cases should be measured. Display output choice slightly favors the A750, with HDMI 2.1 and DisplayPort 2.0, while the RTX 2080 includes a USB Type-C port intended for VirtualLink headsets.

Drivers, Software and Everyday Use

Driver maturity is the other half of this comparison. The RTX 2080 benefits from years of NVIDIA driver work, so older games, emulators, mods and less popular titles generally run without surprises. It remains in NVIDIA’s supported lineup, and features such as NVENC streaming, NVIDIA Broadcast and wide game-profile support all work on Turing.

Intel’s Arc drivers improved substantially after launch, particularly for DirectX 9 and DirectX 11 games, and reviewers who revisited the A750 generally found better results than at release. Even so, Arc remains more likely to show occasional frame pacing problems or launch-day issues in newer or niche games. The A750 does have a strong media engine with hardware AV1 encoding, which the RTX 2080 lacks, making it attractive for streamers who use AV1-capable platforms or record gameplay at high quality.

For a plug-in-and-play experience with the fewest compatibility questions, the RTX 2080 is the safer card. For a buyer comfortable updating drivers and checking per-game behavior, the A750 is a capable budget option.

Cost per Frame

Using launch MSRPs, with the A750 as 100 performance points and the RTX 2080 as about 115:

• RTX 2080: $699 / 115 = about $6.08 per point.
• Arc A750: $289 / 100 = about $2.89 per point.

Those prices are four years apart, so the more useful comparison is the used market. As a broad guide, RTX 2080 cards usually sell for about $160-$220 used and Arc A750 cards for roughly $130-$180. At a $190 RTX 2080 and a $150 A750, that is about $1.65 versus $1.50 per point, close enough that the RTX 2080’s consistency and DLSS support usually justify the difference. An eight-year-old RTX 2080 deserves a careful inspection for fan wear and high memory temperatures.

Recommendations by Situation

The RTX 2080 is the stronger all-rounder, and the A750 is a budget alternative that performs best on modern platforms with modern games. Neither is a long-term solution for high-texture 1440p gaming because of their 8 GB buffers.

Frequently Asked Questions

Is the RTX 2080 faster than the Arc A750?

Yes, by roughly 10-20% on average. TechPowerUp placed the A750 near RTX 2070 level, and the RTX 2080 is a tier above that. The gap is smaller in modern DirectX 12 games and larger in older DirectX 11 titles.

Does the Arc A750 work in older PCs?

It runs, but Intel requires Resizable BAR for the expected performance, and without it frame rates and smoothness drop significantly. Platforms older than Intel 10th gen or AMD Ryzen 3000 usually lack ReBAR. The RTX 2080 is a better fit for those systems.

Can the RTX 2080 use DLSS frame generation?

No. DLSS frame generation requires RTX 40 or RTX 50 hardware. The RTX 2080 supports DLSS super resolution, including the newer transformer model, and can use AMD FSR 3.1 frame generation in supported games.

What PSU do I need for either card?

A quality 550-600 W unit with one 6-pin and one 8-pin PCIe connector suits a typical gaming build. Both cards draw a similar amount of power, 215 W and 225 W.

Is 8 GB of VRAM still enough on these cards?

For 1080p at medium to high settings, mostly yes. Some recent games exceed 8 GB at high textures, so reducing texture quality is often needed. Neither card has an advantage since both have 8 GB.