Introduction

Intel today launched the Arc B570 graphics card, its second release from the Arc Battlemage generation, and we have with us the Sparkle Arc B570 Guardian OC. The B570 is a follow-up to the B580, which has had a great response both from reviewers and gamers for the value and future-proofing it offered at its starting price of $250, outclassing segment leaders such as the GeForce RTX 4060—no small feat, AMD has been trying. The B570 comes in at a starting price of $220, although custom design cards seem to all start at $230, including the Sparkle Guardian OC we are reviewing today. This sets the B570 about $30-40 apart from the B580, and since the GPU has fewer physical components on the board, such as one less memory chip and lighter VRM, Intel designed the B570 to fry the competition around the $200-mark, or even below it. NVIDIA hasn't released a successor to the RTX 3050 in the RTX 40-series, and AMD has a Pacific-wide performance gap between the RX 6500 XT and the RX 7600. This segment is what Intel is going after with the slick new B570.



The B570 is based on the Xe2 Battlemage graphics architecture which the B580 brought to discrete gaming GPUs, but which made its debut with the iGPU of Intel's Core Ultra Lunar Lake mobile processor. Intel took many learnings from the Arc Alchemist generation, took a good look at where the market is headed, and where the money is at, before sitting down to design Battlemage. The company claims a massive 70% SIMD performance gain over Alchemist for its newer Xe cores, which allows Intel to give its GPUs fewer numbers of them to achieve a performance-watt target; and anywhere between 50% to 100% gain in the various processes related to ray tracing, allowing Intel to lower the performance cost of enabling ray tracing in your AAA games. The company also claims a 50% generational performance per watt gain, due in part to the new TSMC 5 nm EUV foundry node. We know from our Arc B580 reviews that many of these claims hold up when you compare it with the Arc A580.

The B570 is based on the same 5 nm BMG-G21 silicon as the B580, but is further cut down. We don't know the exact count of Xe cores on this silicon yet, but the B580 has 20 of them enabled, and the B570 reduces this by 10%, to 18 Xe cores. This results in 112 execution units (EU), or 2,240 unified shaders. Other specs include 144 XMX units for AI matrix math acceleration, 18 second generation ray tracing units, 144 TMUs, and 60 ROPs. Intel has given the B570 a respectable 10 GB of memory, across a slightly narrowed 160-bit GDDR6 memory bus. This memory ticks at 19 Gbps, to yield 380 GB/s of memory bandwidth, which is still a superior memory configuration to the ones the RTX 4060 and the RX 7600 come with—both with 8 GB of 18 Gbps 128-bit GDDR6 memory. The B570 retains all the display I/O and media acceleration chops of the B580. On the I/O front, you get DisplayPort 2.1 with up to UHBR13.5 along with HDMI 2.1; and on the media front, you get two sets of hardware-accelerated video encoders and decoders, which can crunch through HEVC and AV1, making the B570 a fairly powerful GPU for video professionals.

Beyond the hardware, Intel brings to the table a surprisingly mature software backbone for its Arc GPUs, including an agile driver update cycle that responds to new game releases, a clean software front-end for its drivers; and a formidable software feature-set, including the new XeSS 2. This combines XeSS super resolution, along with the new XeFG (Xe Frame Generation), and XeLL (Xe Low Latency), which bring Intel up to par with NVIDIA and AMD on similar technologies. XeSS 2 not only makes gameplay at higher settings possible, but also gives the GPU a degree of future-proofing.

We could categorize the Sparkle Arc B570 Guardian OC as a semi custom-design. The cooling solution is custom-designed by Sparkle, including its aluminium fin-stack heatsink and a set of premium axial flow fans; but the underlying PCB is Intel reference (which is a good thing). There is no reference design card for the B570, so the Sparkle Guardian OC is as close as it gets to one; but at a $10 premium. Justifying this is a small factory overclock of 2.66 GHz engine clock, compared to 2.50 GHz reference. The memory is left untouched at 19 Gbps.

Xe2 Battlemage Architecture

The Arc B580 debuted Intel's second discrete gaming GPU architecture, codenamed Xe2 "Battlemage," in December 2024, and the new Arc B570 is the second discrete GPU based on it. A variant of Xe2 is used in the integrated graphics solution of Intel's Core Ultra 200V "Lunar Lake" mobile processors, the one today is its fully-fledged discrete gaming GPU version, with all hardware components enabled. The Arc B570 is a mid-range model based on the BMG-G21 silicon. The B570 is further cut down from the B580, which itself doesn't appear to max out the BMG-G21 silicon it is based on. The BMG-G21 is built on the TSMC N5 (5 nm EUV) foundry node, and packs 19.6 billion transistors across a 272 mm² monolithic die. The 5 nm node is contemporary, given that both NVIDIA "Ada" and AMD RDNA 3 gaming GPUs use it.

The BMG-G21 GPU features a PCI-Express 4.0 x8 host interface on the Arc B580 and B570. It is configured with a 192-bit GDDR6 memory bus on the B580, and a 160-bit bus on the B570. The GPU is organized in a very similar manner to modern GPUs from NVIDIA and AMD—a Global Dispatch processor distributes work among the five Render Slices, which talk to each other over the GPU's fabric and memory sub-system. The GPU's internal last-level cache is 18 MB in size. Besides the five Render Slices, there is the Media Engine, consisting of two MFX multi-format x-coders (encoders/decoders); and there are two sets of hardware encoders and decoders. Then there's the GDDR6 memory controller and the Display Engine, with four display interfaces.


Intel is claiming a 70% generational increase in performance per Xe Core, the indivisible number-crunching subunit of the GPU; and a 50% generational performance-per-watt increase. The above graphs illustrate the contribution of the individual sub-systems of the Xe2 Battlemage architecture toward these improvements; and how this plays out in a frametime analysis example of a real-world use case.


The Render Slice diagram (above) highlights the biggest chunk of the generational performance increase by Intel. It's thanks to increased IPC from the Xe Core, a more specialized and capable Ray Tracing Unit, a 300% faster Geometry engine, faster Sampler, 50% increase in HiZ, Z, stencil caches, and increases in performance of the pixel backends. Intel's engineering goal has been to reduce latency wherever it can, and reduce software (CPU) overhead as much as it can. The new 2nd Gen Xe Core features eight 512-bit vector engines, with SIMD16-native ALUs, and many more data formats. Rather than two sets of FP and INT units per vector engine, there is just one set of each per vector engine in Xe2, with larger numbers of ALUs.


Intel introduced its second gen Ray Tracing Unit, with massive generational improvements in performance and capability. It introduces a third Traversal Pipeline, which yields a 50% increase in box intersection performance. A second triangle intersection unit has been added to double the performance of triangle intersections. The BVH cache has doubled in size to 16 KB.

XeSS 2, Frame Generation, and Low Latency

Intel has codified the original XeSS as XeSS Super Resolution (XeSS-SR), as that's what it originally was—a performance enhancement that relies on super-resolution technology. The XeSS-SR SDK gets a new compute dispatcher backend for popular APIs—DirectX 11, DirectX 12, and Vulkan. There are two XeSS-SR models, the regular one, and a XeSS-SR Lite model for GPUs that lack XMX matrix acceleration capability.

XeSS 2 isn't a single technology, or an improvement over XeSS-SR, but a collection of three technologies—the existing XeSS-SR, which deals with performance; the new XeSS Frame Generation (XeSS-FG) technology, which nearly doubles frame rates based on intelligent frame doubling; and the new Xe Low Latency (XeLL) technology, which works to reduce the latency cost of SR and FG, but is something that can be used as a standalone whole-system latency technology, too.


XeSS-FG can either be implemented at native resolution, or in conjunction with XeSS-SR, where it is located right after the XeSS-SR step in the rendering queue. It relies on motion vectors, depth data, temporal frame data, and optical flow reprojection, to create interpolated frames that are then interleaved with the output frames, to effectively double the framerate. The interpolated image is then passed along to the next stage, where the HUD/UI is added at native resolution, and pushed to the frame buffer for output.


The SR + FG passes contribute to frame latency, and so, just as NVIDIA uses Reflex to counteract this latency, Intel innovated XeLL. The technology intelligently compacts the rendering queue to reduce the time it takes for an input to register as motion on-screen. XeLL remains enabled in all workloads the use XeSS-FG, but it can be used as a standalone feature, too. There's also an implicit driver-based low-latency mode that does this without a game having an explicit XeSS 2 or XeLL implementation.


Intel has updated its software package significantly. The new "Intel Graphics Software" replaces the "Arc Control" utility, and gives you a cleaner user interface. There are many new settings related to the display, including display scaling model/method/quantization range; 3D graphics settings, including a driver-based FPS limiter, the driver-based low-latency mode; and the exhaustive new Performance and Overclocking controls, which include the ability to set frequency offsets, tinker with the V/F curve, power limits, and GPU and memory clocks. It also integrates Intel's PresentMon metrics.

Packaging

The Card

Sparkle's B570 Guardian OC comes with a fantastic-looking blue color theme—I really love how the various shades of blue come together. On the back you get a high-quality metal backplate with a cutout for air to flow through.


Dimensions of the card are 24.0 x 14.0 cm, and it weighs 788 g.


Installation requires three slots in your system. We measured the card's width to be 45 mm.


Display connectivity includes three standard DisplayPort 2.1 and one HDMI 2.1a. On the Intel reference card one port is marked with a black outline, but not on this card. That port supports higher bitrates (UHBR13.5), i.e. 4K up to 360 Hz, while the other two ports support up to 240 Hz, so I guess you have to test them to find the right one (if you have a monitor that requires such high transfer rates).

In terms of codecs, you get full support for H.264, H.265, VP9 and AV1, both encode and decode. Worth highlighting is HEVC 4:2:2 10-bit encoding and decoding, which is a unique capability, and AV1 Screen Content encoding, which improves the quality of text in movies—fantastic for screen recordings or screen sharing.


The card uses a single 8-pin connector, plus PCIe slot power, allowing a maximum power draw of 225 W.

Teardown

The cooler uses four heatpipes and provides cooling for the GPU, memory and VRM circuitry.


On the back, you'll find a metal backplate with a cutout to improve airflow.