AMD Ryzen 9 7900X / Ryzen 9 7950X Benchmarks Show Impressive Zen 4 Linux Performance

[coder]

read, i didn't say none, i said not too creative. enough to catch up with alder lake but nothing like zen 3 did back in the day.

Maybe you can visit AMD and tell them how to design a CPU, then.

I can't believe how many people seem to think IPC is just a knob that you can trivially dial up or down. Increasing IPC is hard, and what they did with Zen 3 was pretty amazing. You can't just expect that to be the norm, from now on.

[jrch2k8]

No, AMD says Zen 4 offers better perf/W.





Source: https://www.anandtech.com/show/17552...coming-sept-27


Cost-wise, sure. AM4+DDR4 is the better deal, if you don't need a lot of performance.

what i meant is sure, you can have a 1000w cpu that is more efficient than 65w cpu but that doesn't mean i'm comfortable with that power draw.

Unless the laws of physics changed, regardless of efficiency more draw mean more amperage and that means more stress on all the components and their prices because they have to be more robust due to the extra heat.

those slide are marketing slides for "Efficiency, plz plz plz don't look at our power draw, shiny numbers, more shiny number, magic"

yeah because a 5950x became a celeron and offer barely any performance ...

Note, i have the same thoughts on alder lake and expect even bigger numbers with raptor lake with the customary marketing slide of "Efficiency"

[WannaBeOCer]

As expected Zen 4's IPC is only 3.85% faster than Alder Lake. Raptor Lake will definitely be the go to for performance. They already confirmed a 40% increase in multi-core workloads compared to Alder lake. While rumors for single-core workloads will be between 8-15%.

[coder]

what i meant is sure, you can have a 1000w cpu that is more efficient than 65w cpu but that doesn't mean i'm comfortable with that power draw.

As you increase power, efficiency typically decreases quite rapidly. It would be extremely surprising to see a 1 kW CPU that's more efficient than a 65 W model. I think it's quite impossible (on a similar manufacturing process), as long as we're staying in the realm of classical computing.

those slide are marketing slides for "Efficiency, plz plz plz don't look at our power draw, shiny numbers, more shiny number, magic"

No, they're comparing performance at the same power, as the slides say. It's not smoke & mirrors.

And if you read other reviews of AM5, they provide BIOS options to restrict power, which you can do without much loss in performance.

Note, i have the same thoughts on alder lake and expect even bigger numbers with raptor lake with the customary marketing slide of "Efficiency"

No, they cannot. It's basically the same design, on the same manufacturing node. Sure, they tweaked a few things, but it can't meaningfully move the bar on efficiency from Alder Lake.

[jrch2k8]

Maybe you can visit AMD and tell them how to design a CPU, then.

I can't believe how many people seem to think IPC is just a knob that you can trivially dial up or down. Increasing IPC is hard, and what they did with Zen 3 was pretty amazing. You can't just expect that to be the norm, from now on.

That is not what im saying, what im saying is after the IPC adjustment to match alder lake like intel did frequency brute force was the name of the game hence the extra power draw. im not comfortable with the power draw of either vendor atm hence i will wait another node shrink/redesign(it should come with zen5) for something more to my taste but in the meantime an used 5950x will do just dandy if i need to upgrade.

same reason i will forget RTX 40 even existed

[Mahboi]

I'm very confused. How do we go from a TSMC 7nm drawing between 65W-150W to a TSMC 5nm drawing 105W-250W, two years of work from AMD on IPC, a whopping 1Ghz improvement across all processors, and we end up with a broad 15-20% increase in most benchmarks? Yes some are stellar (compilation and some AVX-512 related things), but if my poor brain can math for once in my life:

7nm to 5nm: expected roughly 30% performance increase or lower consumption
150W top to 250W top: 66% wattage increase
2 years of work and a promised ~8% IPC increase
1Ghz on top of a broadly 4Ghz lineup: 20-25% increase

So I count 8% IPC + 20% speed + 30% lower consumption + 66% wattage increase. Even if you know that more watts != more speed, I'm still baffled here. How are we seeing such a massive power draw and such a "good" growth of speed? Since some benchmarks are incredibly faster, I'm thinking it could be a microcode question, that AM5 and Zen4 are entirely new and that AMD will have months of optimisation in the coming months/years to fully exploit them, but otherwise, I really really do not understand how these numbers add up.

In any case, my little 5600x is doing amazing in perf/watts, so I'm very very happy I bought it.

[Etherman]

I'm very confused. How do we go from a TSMC 7nm drawing between 65W-150W to a TSMC 5nm drawing 105W-250W, two years of work from AMD on IPC, a whopping 1Ghz improvement across all processors, and we end up with a broad 15-20% increase in most benchmarks? Yes some are stellar (compilation and some AVX-512 related things), but if my poor brain can math for once in my life:

7nm to 5nm: expected roughly 30% performance increase or lower consumption
150W top to 250W top: 66% wattage increase
2 years of work and a promised ~8% IPC increase
1Ghz on top of a broadly 4Ghz lineup: 20-25% increase

So I count 8% IPC + 20% speed + 30% lower consumption + 66% wattage increase. Even if you know that more watts != more speed, I'm still baffled here. How are we seeing such a massive power draw and such a "good" growth of speed? Since some benchmarks are incredibly faster, I'm thinking it could be a microcode question, that AM5 and Zen4 are entirely new and that AMD will have months of optimisation in the coming months/years to fully exploit them, but otherwise, I really really do not understand how these numbers add up.

In any case, my little 5600x is doing amazing in perf/watts, so I'm very very happy I bought it.

Please find a comparison of the new processor in 65 W "ECO" mode vs Zen 3. There one can really see perf/watt gains.

[WannaBeOCer]

Intel confirmed Raptor Lake will be 41% faster in multi-core workloads and 15% faster in single-core workloads than Alder Lake.

https://www.youtube.com/watch?v=XyFrxMRdAf0

[Anux]

Intel confirmed Raptor Lake will be 41% faster in multi-core workloads and 15% faster in single-core workloads than Alder Lake.

https://www.youtube.com/watch?v=XyFrxMRdAf0

Yes but those are Intels "real world benchmarks", you know that it will be lower in real reality.

[WannaBeOCer]

Yes but those are Intels "real world benchmarks", you know that it will be lower in real reality.

Ever since their Sunny Cove roadmap their performance benchmarks have been accurate. Looking at GeekBench leaks the performance numbers seem very close to accurate with 41% multi-core/15% single-core uplift. Considering these could have still been engineering samples.

https://browser.geekbench.com/v5/cpu/16375702