Intel’s debut with the Alchemist A-Series GPUs was commendable, but the Battlemage B-Series, takes things a step further. With 12 GB of GDDR6 memory and a maximum power draw of just 190 W, the B580 strikes a fine balance between power efficiency and performance.
In our 1440p testing, it delivered robust results, particularly in titles like Red Dead Redemption 2, where it performed closer to the RTX 4060 Ti, while in others, it matched or outperformed the RX 6700 XT. However, occasional stability issues, such as crashes and black screens, were noted during testing. While these problems were inconsistent and represent an improvement over the A-Series, they highlight the importance of considering driver maturity and system compatibility when opting for this card. As with the last series, Resizable BAR is necessary, which might complicate compatibility with older games.
Labs tested the Intel Arc B580 Limited Edition, which is currently sold out as of the publication date with Intel promising more stock later this month. The Limited Edition, or reference card, is available for $249 USD MSRP. Partner manufacturer cards are available across various retailers ranging from $249 to $279 USD.
WHAT YOU NEED TO KNOW
THE GOOD
- Price-to-Performance
At a launch MSRP of $249 USD, the B580 provides an impressive performance-per-dollar ratio, making it an excellent choice for budget-conscious gamers.
- 1440p Gaming Performance
The card excels in 1440p gaming, offering competitive performance against NVIDIA and AMD's mid-tier GPUs, including the GeForce RTX 4060 Ti and Radeon RX 6700 XT.
THE BAD
- Future-Outlook
Stability and compatibility issues remain a concern, and Intel's long-term commitment to its GPU lineup is uncertain.
THE REST
- Overall
The B580 delivers robust value for gamers on a budget. While it doesn't dominate the mid-tier market, it's a solid performer in 1440p gaming and offers competitive results against NVIDIA and AMD's GPUs.
The B580's hardware design is grounded in Intel's Xe2 architecture, which features 2560 shading units, 20 ray tracing cores, and 19,600 million transistors packed into a 272 mm2 die. Its dual-slot, dual-fan design ensures it fits comfortably in most PC cases while maintaining a manageable thermal design power (TDP) of 190 W. These specifications are comparable to those of the RTX 4060 Ti, making the B580 a competitive alternative in the mid-tier segment.
1 of 4
Front of the box.
1 of 3
1 of 5
Front of the card.
The B580 supports DirectX 12 Ultimate, AV1 Encode/Decode, and Intel's XeSS upscaling technology, which collectively enhance its utility for both gaming and productivity. Intel has also made strides in optimizing drivers for DirectX 9 and DirectX 11 titles, showing noticeable improvements over its predecessors. Additionally, Intel Graphics Software now includes useful tools such as frames-per-second limiter, a driver-level low-latency mode, and advanced display settings like color and scaling adjustments.
In terms of performance, the B580 shines brightest in 1440p scenarios, consistently outperforming the RX 6700 XT and occasionally surpassing the RTX 4060 Ti in select titles. At 1080p the card's performance aligns more closely with the RX 6700 XT, with occasional spikes nearing the RTX 4060 Ti. Titles such as The Last of Us Part I saw the B580's performance fluctuate, but overall, it proved to be a reliable contender in most modern games. When aggregated across all 1440p tests, the B580 demonstrated near-parity with the RTX 4060 Ti, while also clearly outperforming the RX 6700 XT. Productivity and synthetic benchmarks reveal a similar story, with the B580 showing competent results.
We will periodically retest graphics cards with updated drivers. We continuously monitor and update our results for accuracy. Stay tuned for the latest test results and insights on this product and others as we aim to provide up-to-date information on a consistent basis.
We actively address test bench variance in our validation process. Our methodology involved validation and calibration procedures to ensure consistency and reliability across all tests. The verified testbed variance for benchmarks in this product page is +/- 0.25%.
By monitoring and controlling variables such as hardware configurations, ambient temperatures, and system settings, we minimize potential sources of variance that could affect our results. This detailed approach enables us to deliver the most objective and accurate assessments.