According to a leaked report, there are at least two known SKUs at this point: the Core i7-8705G (694E) and Core i7-8809G (694C). The Radeon GPU featured on the multi-chip module (MCM) reportedly features 24 compute units and a total of 1536 stream processors. The GPU in the Core i7-8705G is running at 1000MHz, while that clock rises to 1,190MHz on the Core i7-8809G. Though both GPUs have 4GB of HBM2, the low-end (694E) and high-end (694C) parts have memory clocks of 700MHz and 800MHz respectively.
As for the processors, both are listed as having 4 cores capable of executing total of 8 threads. Both processors appear to feature a base clock of 3.1GHz and boost clocks that max out at 4.1GHz.
Moving on to the leaked benchmarks, we see the Core i7-8705G and the Core i7-8809G duking it out in GFXBench. The latter far outperforms the former across the board in the graphics tests, leading us to believe that there has to be some more significant configuration differences between the parts to account for such a wide swing in performance.
There's also a Geekbench entry, in which the Core i7-8809G managed to put up an OpenCL score of 76607.
Likewise, we some additional benchmark goodness that compares the Core i7-8809G to the Core i7-8705G in 3DMark 11 (Performance Preset):
Rounding things out, we have a look at the Core i7-8705G running the Ashes of the Singularity benchmark; 1080p High and 1080p Low.
For an on-chip GPU, it looks as though Intel and AMD have a real winner here. If Intel’s OEM partners can deliver on the promises of thin and light designs, this could help reshape what we’ve come to expect with regards to computing and graphics performance in ultraportables. The only thing we want to know at this point is what NVIDIA must be thinking now that its most fierce graphics rival has teamed up with the dominant player in the processor market.