This isn’t the first time Intel’s flagship Arc Alchemist GPU setup has been benchmarked in the Geekbench OpenCL benchmark, but it is the first time it has been named properly. The Intel Arc A770 graphics card will be a high-end gaming graphics card with the complete ACM-G10 GPU, which is the more powerful of the two Arc Alchemist GPUs.
The flagship ACM-G10 GPU will be used in the Intel Arc 7 series, and we already know about the mobility variants, which include the Arc A770M and Arc A730M. Similarly, the Arc A770 is a high-end desktop PC that comes with the full ACM-G10 configuration, which includes 32 Xe-Cores, 4096 ALUs, and 32 ray tracing units.
Now, Geekbench doesn’t specify the precise VRAM configuration and reports 12.7 GB, which is an error, thus it’s most likely 16 GB GDDR6. Also, given that the Arc A780 is expected to be the flagship graphics card, this might be a clock-optimized GPU variation, with the A780 launching as the Limited Edition model teased during the Arc A-series launch event.
In terms of clocks, the GPU runs at a top boost of 2.4 GHz, which is always higher than the engine clock speed reported. The GPU should be capable of delivering close to 20 TFLOPs of FP32 horsepower at 2400 MHz.
The score of Intel ARC Alchemist in the OpenCL Graphics benchmark is pretty low
The Intel Arc Alchemist A770 received 85585 points in the OpenCL graphics benchmark, which is quite low for such a high-end graphics card, but that’s to be expected given that the graphics drivers aren’t yet optimised for high-end parts.
Even the entry-level ACM-G11 GPUs haven’t been adequately tuned through drivers, so performance should increase in the next months. The card is currently comparable to the NVIDIA GeForce RTX 2070, and we can expect it to be comparable to the RTX 3070 and 6700 XT when it launches in the second quarter of 2022.
Also Read:
