The TechTuber says that it is very likely that Battlemage will be housing 64 Xe Cores, which again, is double the current 32 Xe Cores that is being utilised by the A770, the current “top-tier” discrete graphics in Intel’s Arc arsenal. Additionally, it is speculated that the GPU will be categorised under the codename BGM-G10, in the same way the current Alchemist GPU goes by the code AGM-G10. RedGamingTech also hints at other specifications of the Battlemage GPU, including a targetted clockspeed of 3GHz or higher, a 256-bit memory bus, and possibly the adoption of GDDR6X memory, although to date, NVIDIA remains the only GPU brand that uses the memory standard in its high-end RTX 30 and 40 Series GPUs. It is also possible that the 2nd generation Arc GPUs will receive 48MB of L2 Cache, and have a die size similar to that of the green team’s AD103 GPU, which is currently being used in the RTX 4080. At the same time, RedGamingTech suggests that Intel may switch gears with Battlemage’s process node, using TSMC’s 4nm die lithography instead of its own process. Now, the likelihood of this happening isn’t impossible, given that the Taiwan-based semiconductor manufacturer already has its silicon fingers stuck in the pies of virtually every major tech player in the industry. However, this rumour also hints at the possibility that Intel is unable to hit its own target of manufacturing its own chips in time for the launch of Battlemage. Further, and based on Intel’s roadmap, there is still the arrival of Alchemist+ that the chipmaker is planning on releasing later, during the second half of this year, most likely as a buffer before they roll out Battlemage early next year. In any case, we advise that you take this bit of news with the recommended consumption of sodium, at least till something more substantial or concrete – hopefully, from Intel directly – presents itself. (Source: Videocardz, RedGamingTech)
