Hi everyone,

I’m currently mapping out an efficiency-first upgrade for my volunteer computing setup. I’m planning to deploy a high-TDP (80W) Panther Lake Mini PC in early February to help with the current Minecraft seed searches.

I noticed the project has an active opencl_intel_gpu binary (v1.03) in beta. I'm very interested to see how the Xe3 architecture on the 18A node compares to the current project averages. Given the restored hardware-native math blocks and the move to a 2nm-class node, I'm expecting a significant jump in GigaFLOPS-per-watt over previous Intel generations.

I’ll be running 64GB of DDR5-9600 to ensure the iGPU has a massive shared memory buffer. I'll post clinfo logs and benchmarks once the unit arrives in Denmark. Is the dev team currently looking for more testers for the Intel GPU beta?

Hi everyone,

I’m currently mapping out an efficiency-first upgrade for my volunteer computing setup. I’m planning to deploy a high-TDP (80W) Panther Lake Mini PC in early February to help with the current Minecraft seed searches.

I noticed the project has an active opencl_intel_gpu binary (v1.03) in beta. I'm very interested to see how the Xe3 architecture on the 18A slope game node compares to the current project averages. Given the restored hardware-native math blocks and the move to a 2nm-class node, I'm expecting a significant jump in GigaFLOPS-per-watt over previous Intel generations.

I’ll be running 64GB of DDR5-9600 to ensure the iGPU has a massive shared memory buffer. I'll post clinfo logs and benchmarks once the unit arrives in Denmark. Is the dev team currently looking for more testers for the Intel GPU beta?

That setup should be interesting for iGPU testing, especially with high memory bandwidth helping OpenCL workloads. Have you checked if the current beta fully utilizes Xe3 features or if there are known driver limitations?