Quote from: _MT_ on April 21, 2022, 12:08:20

Quote from: rs on April 20, 2022, 19:16:30
Obviously this Passmark run is borked. 5800X3D shouldn't be slower than 5800X clock for clock. If 5800X at 4.7 GHz scores 3486 on average in the single thread rating then 5800X3D at 4.5 GHz should score at least ~3340.

You do realize that the 5800X can overclock itself if you have sufficient cooling while the 5800X3D can't. It's the price you pay for the extra cache. The frequencies you quote are not actually achieved values during the test therefore you can't draw the conclusion that you have drawn.

The passmark scores are similar to a 3700x so I would extremely say its borked if we aren't seeing the IPC improvements without overclock.

It's always the case that you have to look at relevant benchmarks. Workloads differ. If you play a particular game, ideally, you want a benchmark based on that game. Or find a good proxy (a game that tracks well). Also, if you're playing on high resolution with high details, you're probably limited by your GPU. There is not much the CPU can do about it.

Quote from: rs on April 20, 2022, 19:16:30
Obviously this Passmark run is borked. 5800X3D shouldn't be slower than 5800X clock for clock. If 5800X at 4.7 GHz scores 3486 on average in the single thread rating then 5800X3D at 4.5 GHz should score at least ~3340.

You do realize that the 5800X can overclock itself if you have sufficient cooling while the 5800X3D can't. It's the price you pay for the extra cache. The frequencies you quote are not actually achieved values during the test therefore you can't draw the conclusion that you have drawn.

It's obvious that only applications that can use the extra cache can benefit from it. And scientific benchmarks tend to be optimized to use smaller caches as efficiently as possible, minimizing the room for improvement. Efficient data access patterns are a huge factor in performance. Because processors are so much faster than memory. Games, on the other hand, often have reputation for poor optimization.

For the love of god STOP using PassMark for just about anything whether it favors AMD or Intel. They are notorious for being wrong, purposely manipulated, and even by their own admission not accurate.

Google Search PassMark Intel Bribe

Quote from: Big Black Clock on April 20, 2022, 18:21:23
"this AMD chip is for gaming" - nah, it's for the few spoiled brats who can afford to buy one

You realize that is cost more to buy a Intel CPU?

Shocker I know

Obviously this Passmark run is borked. 5800X3D shouldn't be slower than 5800X clock for clock. If 5800X at 4.7 GHz scores 3486 on average in the single thread rating then 5800X3D at 4.5 GHz should score at least ~3340.

"this AMD chip is for gaming" - nah, it's for the few spoiled brats who can afford to buy one