I find the video from LTT kinda hilarious with the 96 core threadripper. Breaking records in cinebench but Cities Skylines 2 still runs like shit (in a 1mio pop city).
Because chances are the 7800X3D will be faster due to the cache.
Real-world applications often can only be parallelized so and so much, before you start hitting diminishing returns for many reasons. A lot of it is about the actual technical design as much as it is the technical execution (you can’t parallelize two operations if one depends on the result of the other).
That’s the point I think. I haven’t delved into the specifics too much but the 7800X3D favors big cache sizes (at all layers iirc) over cores/threads quantity. So, it should fare better with games that aren’t very optimized for multi threading (ie, most games)
I find the video from LTT kinda hilarious with the 96 core threadripper. Breaking records in cinebench but Cities Skylines 2 still runs like shit (in a 1mio pop city).
Because chances are the 7800X3D will be faster due to the cache.
Real-world applications often can only be parallelized so and so much, before you start hitting diminishing returns for many reasons. A lot of it is about the actual technical design as much as it is the technical execution (you can’t parallelize two operations if one depends on the result of the other).
Also cache optimization has been a huge trend in games programming in recent years
I’m not sure the 7800X3D would even run a mil pop city. They were using 64 of the 96 cores running Skylines 2.
Teeth don’t render themselves, buddy.
That’s the point I think. I haven’t delved into the specifics too much but the 7800X3D favors big cache sizes (at all layers iirc) over cores/threads quantity. So, it should fare better with games that aren’t very optimized for multi threading (ie, most games)