Introduction

ASRock Arc A580 Challenger is the company’s custom-design graphics card that debuts Intel’s latest and most interesting addition to the desktop GPU space. The new A580 Alchemist comes in at a disruptive starting price of just $180, offering 1080p gaming with high through enthusiast settings, the full DirectX 12 Ultimate feature-set, including real time ray tracing, and XeSS performance enhancement. For these reasons, and the fact that it’s based on the contemporary 6 nm foundry node, the A580 is every bit a current-generation GPU, designed to appeal to the class of people entering the PC gaming segment, and could do with a trusty brand like Intel.

Why Intel decided to launch the A580 now, nearly a year after its Arc 7-series debut, can be explained with two distinct theories: first, prices of graphics cards are finally cooling down, which brought 1080p-class graphics cards down to around $250, presenting Intel with an opportunity to undercut products such as the GeForce RTX 3050 and Radeon RX 6600 at $180; Intel has taken huge strides over the past year towards upgrading the software backbone of its Arc GPUs, which effectively push each product a segment above. This has been done with dozens of game-specific improvements, as well as API-specific uplifts for DirectX 11 and DirectX 9, which are still relevant for popular e-sports titles.

The Arc A580 is based on the Xe-HPG Alchemist graphics architecture, which is a thorough, grounds-up attempt by Intel to engineer a contemporary discrete GPU that isn’t missing out on any kind of feature compared to the latest GeForce RTX and Radeon RX GPUs. The GPUs also feature a real time ray tracing engine that’s technically more advanced than the one AMD uses, with hardware acceleration for even more stages of the ray tracing, full asynchrony with ray tracing, and shader re-ordering. The GPU also features XMX a dedicated matrix math accelerator that makes these GPUs AI processing powerhouses, something Intel leverages for its XeSS performance enhancement that’s technologically at par with DLSS 2.

Intel carved the Arc A580 out of the same 6 nm ACM-G10 silicon powering the company’s Arc 7 series. The new SKU is endowed with 24 out of the 32 Xe Cores physically present on the silicon, across 6 out of 8 Render Slices. This works out to 384 execution units, which are worth 3,072 unified shaders, 384 XMX cores, 24 Ray Tracing Engines, 192 TMUs, and 96 ROPs.

The most interesting aspect of the A580 is that it gets the same exact memory sub-system as the A750. There’s 8 GB of GDDR6 memory on offer, across a wide 256-bit memory bus, which paired with its 16 Gbps memory speed, yields 512 GB/s of memory bandwidth, the highest in the market segment. That’s not all, the A580 has the full PCI-Express 4.0 x16 host interface, compared to its rivals, which usually come with truncated Gen 4 x8 interfaces. What this means is that you get uncompromised performance on older PCIe Gen 3 platforms such as the 10th Gen Core “Comet Lake” or Ryzen 2000, which support Resizable BAR (something that’s required for Arc GPUs to reach their advertised performance).

We’ve known the Challenger line of graphics cards from ASRock with a several lines of AMD Radeon graphics cards. These cards represent the company’s value custom-designs, with compact product designs, a cooler focused on low-noise and easy serviceability; and a price that’s usually close to the GPU’s baseline, which in this case, is $180. The ASRock Arc A580 Challenger is for those that just want an A580 that they can install and forget about.