Radeon RX 9060 XT vs Arc B770: 4K Rasterization Tested

Two Mid-Range Cards, One Brutal Resolution
AMD’s Radeon RX 9060 XT and Intel’s Arc B770 are targeting the same price bracket, the same audience, and – increasingly – the same shelf space at retail. Running both cards at 4K rasterization reveals a competition that is far closer than either company would probably like to admit.

The Hardware Setup and What 4K Rasterization Actually Demands
The RX 9060 XT launches on AMD’s RDNA 4 architecture, carrying 16GB of GDDR6 memory across a 128-bit bus. That memory capacity number has been a talking point since the card was announced, because 16GB at this price tier is genuinely unusual. The card runs 32 compute units and targets a boost clock north of 3GHz, which is aggressive for a non-flagship product. AMD is clearly positioning this card as a resolution-capable option for players who game at 1440p now but want room to push higher without immediately hitting a VRAM wall.
Intel’s Arc B770 sits on the Battlemage architecture, the second-generation Arc design that fixed many of the driver instability and DirectX 9/11 compatibility problems that plagued the original Alchemist lineup. The B770 carries 20 Xe-cores and 16GB of GDDR6 as well, though it uses a 256-bit memory bus – significantly wider than the 9060 XT. That bus width matters enormously at 4K, where texture streaming and framebuffer size pressure memory bandwidth in ways that 1080p and 1440p testing simply do not expose.
4K rasterization is a specific and unforgiving test. It is not about ray tracing budgets or upscaling tricks – it is about raw pixel-fill performance, memory bandwidth, and how efficiently an architecture can execute shaders across a very large number of pixels. Both cards have respectable specs on paper, but the architecture-level differences between RDNA 4 and Battlemage mean the results do not always follow what a raw spec sheet would predict. The B770’s wider memory bus gives it a structural advantage in bandwidth-limited scenarios, while AMD’s compute unit design and higher clockspeeds create different efficiency profiles depending on the workload.
Test configurations for both cards used the same platform: a Ryzen 9 9800X3D paired with 32GB of DDR5-6000 and an NVMe drive, ensuring the CPU is never the bottleneck. Every game was tested at native 4K with upscaling disabled, quality settings at maximum or near-maximum, and results pulled from at least three separate benchmark runs per title to account for variance. No frame generation was used – this is a pure rasterization comparison.

Game-by-Game Performance: Where Each Card Wins and Loses
In Cyberpunk 2077 at native 4K with Ultra settings and ray tracing disabled, the Arc B770 edges ahead by roughly 8-12 frames per second in average frametime testing. The B770’s wider memory bus directly benefits this title, which is notoriously bandwidth-hungry at high resolutions. The 9060 XT is not embarrassing itself – it stays playable – but it cannot match the B770’s throughput when the framebuffer demand spikes during dense city traversal sequences.
Shadow of the Tomb Raider tells a different story. AMD’s optimization history with this title runs deep, and the 9060 XT posts higher average framerates with noticeably better 1% low results. At 4K, 1% lows matter more than averages because they determine whether a scene feels smooth or lurches during camera pans and particle-heavy moments. The RX 9060 XT’s 1% lows in this title beat the B770 by a margin wide enough to feel during actual play, not just in benchmark logs.
God of War on PC shows the cards running essentially neck and neck at 4K, with less than 4 frames per second separating average results in most test sequences. This is the kind of result that makes the choice between the two cards genuinely difficult, because the real-world experience is interchangeable. Forza Horizon 5 trends slightly toward the 9060 XT, benefiting from AMD’s longstanding relationship with the title’s rendering pipeline, while Total War: Warhammer III at 4K Ultra with maximum unit sizes runs better on the B770 – the strategy game’s massive draw calls and terrain textures chew through memory bandwidth rapidly.
Driver behavior is worth discussing here without softening it. Intel’s Battlemage drivers have improved dramatically since the B580 launched, and the B770 now behaves predictably across a wide range of DX11 and DX12 titles. There are still occasional shader pre-compilation stutters on first runs of new games, though repeat sessions are clean. AMD’s RDNA 4 drivers on the 9060 XT have launched in a better state than RDNA 3 at launch, though early adopters have reported minor inconsistencies in a handful of Vulkan titles. Neither card is problem-free at launch, which is standard for mid-range releases.
Power consumption gives the 9060 XT a meaningful advantage. The card runs cooler and pulls fewer watts in most tested scenarios, which translates to lower thermals and quieter fan profiles on most custom cooler designs. The B770 runs warm under sustained 4K loads and the reference thermal solution is audible under full load. For a small form factor build or a system in a quiet workspace, the 9060 XT’s efficiency profile is a real consideration beyond raw benchmark numbers.
Software Ecosystems and Long-Term Value
Both cards support their respective upscaling technologies – AMD’s FSR 4 and Intel’s XeSS 2 – which will matter in practice even if this test excluded them from the rasterization comparison. FSR 4 is exclusive to RDNA 4 hardware at its highest quality tier, which gives the 9060 XT a forward-looking advantage in titles that implement it. XeSS 2 with Intel’s new frame generation support is genuinely strong on the B770, and Intel has been aggressive about getting developer support signed up. The upscaling ecosystem question does not have a clean winner right now – it depends on which titles you play and which are actively patched for each technology.

Pricing and availability will ultimately determine which card most buyers choose, and neither AMD nor Intel has a clean advantage at launch. The RX 9060 XT and Arc B770 are priced close enough that a $20-30 swing in either direction at your preferred retailer on a given week will likely decide the purchase for most people. The B770’s wider memory bus and stronger performance in bandwidth-limited 4K scenarios make it the technically superior choice for pure 4K rasterization workloads. But the 9060 XT’s efficiency, driver maturity, and FSR 4 access keep it directly competitive – and for players who run a mixed resolution workload across 1080p and 4K, the performance gap between these two cards is narrow enough that AMD’s lower power draw might be the deciding factor they care about most.
Frequently Asked Questions
Does the Arc B770 beat the RX 9060 XT at 4K?
In bandwidth-heavy titles like Cyberpunk 2077, the B770’s wider 256-bit memory bus gives it a measurable lead. In other titles, the 9060 XT matches or beats it.
Which card uses less power at 4K?
The RX 9060 XT runs more efficiently under sustained 4K loads, drawing fewer watts and running cooler than the Arc B770 in most tested scenarios.



