Intel Sparkle Arc B580 OC review: One of the best budget 1440p cards we’ve seen
A refreshing option for 1440p gaming
Share this article...
WePC is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
The Intel Arc B580 is the latest addition to Intel’s expanding Arc lineup. It promises mid-tier performance for gamers and creators alike. Built on Intel’s Xe2 architecture, the B580 aims to provide a competitive option in the crowded $300 GPU market. Positioned to compete directly with AMD’s RX 7600 and Nvidia’s RTX 4060, the B580 has a lot riding on its shoulders as Intel continues to carve out its place as one of the best GPUs in the budget category. And we can confidently say this launch has been far better than Alchemist.
This review examines the Intel Arc B580’s design, gaming and synthetic benchmark performance, and overall value for money. We hope it will help you decide if this is the GPU upgrade you’ve been waiting for—if you can find one in stock.
- Core clock speed: 2670 MHz – 2670 MHz (boost)
- Shading units: 2560
- Memory: 12 GB
- Dimensions: 222 x 101mm, 2.2slot
- PSU required: 450 W
- TBP: 190 W
- Core clock speed: 2670 MHz – 2670 MHz (boost)
- Shading units: 2560
- Memory: 12 GB
- Dimensions: 222 x 101mm, 2.2slot
- PSU required: 450 W
- TBP: 190 W
- Memory clock: 2375 MHz
- Memory bus: 192 bit
- Card bus: PCIe 4.0 x8
- Output: 1x HDMI 2.1a 3x DisplayPort 2.1
- Power connectors: 1x 8-pin
The Intel Arc B580 is one of the best mid-range 1440p cards we have tested, Intel have come a long way since Alchemist and they are here to inject some value into the more budget end of the scale. We saw some fantastic performance in 1440p, and some significant triumphs over the RTX 4060 in most cases. We highly recommend this card if you can find it close to the MSRP of $249. (AIB prices will vary)
- Strong performance in 1440p
- Very good ray tracing performance even compared to the 4060
- New architecture improves efficiency
- Three fan design keeps high temps at bay
- The 50 series will likely supersede this card
- Availability is thin (as of the time of writing)
Full Specifications
Core clock speed
2670 MHz – 2670 MHz (boost)
Shading units
2560
Memory
12 GB
Dimensions
222 x 101mm, 2.2slot
PSU required
450 W
TBP
190 W
Core clock speed
2670 MHz – 2670 MHz (boost)
Shading units
2560
Memory
12 GB
Dimensions
222 x 101mm, 2.2slot
PSU required
450 W
TBP
190 W
Memory clock
2375 MHz
Memory bus
192 bit
Card bus
PCIe 4.0 x8
Output
1x HDMI 2.1a 3x DisplayPort 2.1
Power connectors
1x 8-pin
As you can see, Intel has been hard at work improving the B series to be bigger and better than the A series could ever hope. It is a welcome fact that Intel is pushing to give people on the more budget end of the scale more options. We think it’s an area of the market that hasn’t received some love for a while. AMD mentioned that it would also stop competing at the top end of the market and focus more on the mid-to-budget end. Leaving Nvidia to run a mock.
Design and Quality
The Intel Arc B580 Sparkle Titan is the GPU we have for testing; it is a three-fan card painted in a lovely muted blue with pale grey fans. The internal heatsink spans the card’s entire width, allowing plenty of surface area to keep itself cool. Because this GPU isn’t an absolute monster, it only takes up two expansion slots in your case, meaning you can comfortably fit the Sparkle Titan into a smaller form-factor case.
Power comes in the form of a single 8-pin fan connector around the middle of the GPU, which is easily accessible for those building in tight spaces. However, it does look a mess if you don’t manage the cable correctly. While powered on, there’s no discernable coil whine whatsoever, as is present with some of the more powerful AMD cards (the very loud 7900 XT).
How We Tested the Intel Arc B580
As with all our GPU reviews, the Intel Arc B580 underwent rigorous testing in real-world gaming scenarios and synthetic benchmarks.
Following our ‘how we test’ methodology, we compared the Arc B580 to its main competitors, focusing on key metrics like average FPS, 99th percentile performance, and power efficiency across multiple resolutions. We even included some synthetic benchmarks to see how the B580 performed computationally. Here are our results:
Performance
Here at WePC, we keep testing as fairly as possible, even across platforms and generations. That’s why ensuring all testing environments are as close as possible is important. This may involve using the same speed memory kits, motherboards with similar specifications, and CPUs that won’t limit the performance of the GPU (bottleneck). Here are the components that we chose to test the Arc B580.
| Component | Name |
|---|---|
| Motherboard | ASUS ProArt X670E-CREATOR WIFI |
| CPU | Ryzen 7 7800X3D |
| Memory | Corsair Dominator Titanium RGB (64GB, 6600MT/s, CL32) |
| CPU cooler | ROG RYUJIN III 360 |
| PSU | ASUS ROG Thor 1000W |
| Case | Coolermaster Master Frame |
Outlining components is important as it also gives us an idea of how well a certain GPU performs and what you could expect from a similar system at home. On the flip side, if you don’t get the same numbers we do with this GPU, it’s likely the difference in hardware.
Gaming benchmarks
When comparing the Intel Arc B580 and Nvidia RTX 4060 across various games and benchmarks, we see a battle where each GPU excels in different areas. At 1080p, the RTX 4060 generally maintains a performance advantage in titles like CS2; for instance, in CS2, the RTX 4060 delivers an average FPS of 188, compared to the Arc B580’s 167.
However, the Arc B580 flips it around in games like Doom Eternal (222 FPS) and SOTTR (137 FPS), consistently posting higher average and percentile FPS values – though it is pretty close in SOTTR. This trend suggests that Arc B580’s optimization may favour specific engines and workloads, particularly those more VRAM-intensive.
At 1440p, the performance gap widens further. While the Arc B580 holds a slight edge in more demanding scenarios like SOTTR (98 FPS vs the RTX 4060’s 84), the RTX 4060 struggles in memory-intensive situations. Cyberpunk 2077 shows noticeable limitations due to its lower VRAM capacity. This is a recurring theme at higher resolutions, where the Arc B580’s 12GB VRAM enables smoother performance, especially in titles with high texture demands or ray tracing workloads.
In ray tracing, the results are a mixed bag. At 1080p with RT enabled, the Arc B580 leads significantly in titles like Doom Eternal, achieving 162 FPS compared to the RTX 4060’s 125 FPS. However, as the resolution scales to 1440p and 4K, both GPUs struggle with intensive RT workloads, though the Arc B580’s larger VRAM capacity often allows it to maintain better frame rates.
The RTX 4060 slows down in 1440p and 4K RT scenarios, sometimes failing them completely (as we see in Doom Eternal), frequently running out of memory and dropping frames, particularly in Cyberpunk 2077 and SOTTR. However, it has to be mentioned that these GPUs are not meant to play games in 4k.
Synthetic benchmarks
Synthetic benchmarks further highlight the strengths and weaknesses. The B580 scores significantly higher in all 3Dmark tasks, indicating superior architectural efficiency for traditional rasterized performance (in most cases). Video encoding workloads, such as H.265 encoding in Handbrake, also go to the Arc, as it easily handles the 4K TOS test. Scoring 319 average FPS compared to the 157 FPS of the 4060.
Productivity performance provides an interesting twist. In Blender, the RTX 4060 consistently outperforms the Arc B580, demonstrating its strength in 3D rendering tasks. Conversely, the Arc B580’s robust H.265 hardware acceleration allows it to encode 4K video nearly twice as fast as the RTX 4060, making it a compelling choice for content creators focused on video workflows.
The Arc B580 is a capable GPU, particularly in memory-heavy scenarios, ray tracing, and video encoding tasks. However, the RTX 4060 holds its ground with better support in 3D workloads like Blender.
The Arc B580’s larger 12GB VRAM may be more appealing when prioritising higher resolutions or ray tracing. At the same time, if you’re looking for smoother performance in well-optimized titles or 3D rendering, you may gravitate toward the RTX 4060. Both cards offer unique strengths, leaving the final choice dependent on individual use cases and software priorities.
A closer look at the Xe2 Battlemage architecture
Intel’s Xe2 Battlemage architecture powers the Arc B580, marking the company’s second generation of discrete gaming GPUs. Positioned as the mid-range successor to the Arc A580 “Alchemist,” the B580 is built on the new BMG-G21 silicon and represents a refined leap in gaming GPU technology.
The Arc B580 leverages the TSMC N5 (5 nm EUV) process node, packing an impressive 19.6 billion transistors into a 272 mm² die. While its configuration doesn’t maximise the silicon’s capabilities, it perfectly aligns with the mid-range gaming GPU market, balancing power and efficiency. The chip features a PCIe 4.0 x8 host interface and a 192-bit GDDR6 memory bus, paired with an 18 MB last-level cache to support its five Render Slices.
Each Render Slice processes geometry, rasterization, and rendering tasks, connected via an internal fabric and memory subsystem. This modernized layout competes with Nvidia’s Ada Lovelace and AMD’s RDNA 3 architectures, all of which use the 5nm node.
Media and Display Innovations
Beyond raw performance, the Xe2 Battlemage architecture enhances media and display capabilities. The Arc B580 includes two multi-format encoders and decoders, making it adept at handling video workloads. The Display Engine supports four outputs, catering to various high-resolution and high-refresh-rate monitor setups.
Software Ecosystem and Latency Reductions
Intel pairs its hardware innovations with a revamped software suite. The “Intel Graphics Software” replaces the Arc Control utility, offering streamlined overclocking tools, driver-based latency optimizations, and integration with XeSS 2 technologies like Frame Generation and Low Latency (XeLL). These advancements underscore Intel’s commitment to improving both performance and user experience.
With the Xe2 Battlemage architecture, the Arc B580 strikes a compelling balance between price and performance. The design decisions, such as the 192-bit memory bus and advanced render pipeline, make it a strong choice for mid-range gamers seeking modern features like ray tracing and AI-enhanced upscaling. Intel’s focused efforts on efficiency, media processing, and low-latency gameplay could give it an edge in a crowded GPU market.
Final Word
Intel Arc B580 Sparkle Titan OC
