A few days ago Tech PowerUp released a review of the 14900k with various power restraints. This was both telling and interesting for me, and something I really wish I had seen a week ago. However, those other chips are running at stock. It’s not an apples-to-apples comparison. Is there anything similiar to that for any of Ryzen 7000 CPUs? I’ve read plenty of anecdotes, but such evidence is plentiful and all over the place.
Absolute balls-to-the-wall benchmarks are meaningful and have their uses, but when that’s all you have, it’s just a fruitless dick measuring contest. Why Intel is playing that game, I can’t imagine since it’s a really bad look, but that isn’t the subject of this post.
Intel are playing that game because they’re still behind in power efficiency. It’s part process and part architecture, but in practice the result is simply that for any given power limit zen4 will outperform 12th gen and its various refreshes. Heterogeneous architecture ought to give Intel a slight power advantage outside benchmarks and chiplets put AMD at a disadvantage, but in practice e-cores are more space efficient than they are energy efficient (letting them put more cores in a smaller area die and offsetting the cost of large monolithic dies) and TSMC 5nm beats Intel 7nm. Intel are catching up, but it will take time, and until they do become competitive in efficiency they prefer to market on pure performance (because their performance/W is quite unimpressive)