Intel readies Arc Pro B70 and B65 Battlemage GPUs for pro visualization and artificial intelligence

Intel is preparing Arc Pro B70 and B65 Battlemage GPUs based on the BMG-G31 die, targeting professional visualization and artificial intelligence workloads with significantly higher core counts and memory than current Arc Pro models.

Intel has confirmed the Arc Pro B70 “Battlemage” graphics card as a larger, professional visualization and artificial intelligence focused GPU, surfaced through usage in its LLM Scaler software with limited performance testing in a non-ideal scenario. The company plans to release this GPU within the current quarter, which means we could see the official launch and availability in about a month. Both the Arc Pro B70 and Arc Pro B65 are based on the long-rumored BMG-G31 die aimed at higher-end configurations.

For the Arc Pro B70, Intel is planning a BMG-G31 configuration featuring 32 Xe2 cores and 32 GB of GDDR6 memory on a 256-bit bus. This setup translates to approximately 4,096 FP32 cores in its full configuration, which effectively doubles the core count and memory capacity of the current Arc Pro B60 in a single-GPU card. The design targets professional users who need more compute resources and memory bandwidth than the existing Arc Pro B60 can provide, particularly for visualization and large model workloads.

The Arc Pro B65 scales the same BMG-G31 silicon to a smaller configuration that supports 2,560 FP32 cores, with a total of 20 Xe2 cores. While this matches the FP32 and Xe2 core configuration of the Arc Pro B60, the B65 comes with 32 GB of GDDR6 memory, which is 8 GB more than the Arc Pro B60. Dual-GPU configurations were common with the Arc Pro B60, so there is a possibility that board partners such as Maxsun could introduce dual-GPU PCBs using the Arc Pro B70, mirroring their earlier Arc Pro B60 Dual-GPU card and further extending compute density for professional visualization and artificial intelligence workloads.

52

Impact Score

Tsmc debuts A13 process technology

Tsmc has introduced its A13 process at its 2026 North America Technology Symposium as a tighter version of A14 aimed at next-generation Artificial Intelligence, high performance computing, and mobile designs. The company positions the node as a more compact and efficient option with backward-compatible design rules for faster migration.

Google unveils eighth-generation tensor processor units

Google introduced its eighth generation of custom tensor processor units with separate designs for training and inference. The new TPU 8t and TPU 8i are aimed at large-scale model training, serving, and agentic workloads.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.