It's like they dumbed down the threading to one per cpu core, just so they could make a claim that single core performance is better... look at the details, it's not and this is simply deceptive marketing. You can double the performance of the AMD "single thread performance" and then we a more comparable benchmark of "single core performance" and AMD yet again words the floor with Intel; they lose in every metric, it's not even close.

They're a joke and have to cheat by building the chip with these specific metrics in mind now just to seem reliant? Must be DEI initiatives at work, idfk.

This is nuts.

Quote from: Borg on February 05, 2025, 11:08:28I'm a software engineer,

Lol. Good one.

So the new Intel CPU is actually slower than an older AMD CPU. That's progress, I guess. As for the guy still clinging to 'single-threaded performance is the most important metric for productivity'—he either:
A) Thinks Notepad and Photoshop are the only productivity apps
B) Has a time machine and is posting from 1995.

I'm a software engineer, and I actually pay attention to my CPU usage. In the past 8 years, I haven't seen a workload that uses fewer than 8 threads—even when I'm not running VMs. Even basic development work today, like data processing, compilation, rendering, or containerized workloads, eats up as many cores as you give it. Even with 16 cores, I sometimes feel limited.

In 2025, pretending single-threaded performance is king is just coping. Multithreading has been the standard for years. But hey, Intel has to push that narrative because it's the only thing they have left. It's the same pattern—Intel fanboys convincing themselves their overpriced space heaters are the best, while ignoring decades of bribery, anti-competitive behavior, security flaws, and inefficiency. Meanwhile, AMD keeps winning in real-world performance.

The real question is: how much is Intel paying you?

Quote from: A Guy on February 05, 2025, 01:10:47

Quote from: Worgarthe on February 04, 2025, 10:23:44

Quote from: Illrigger on February 04, 2025, 09:45:14The desktop parts also beat AMD in single core on the desktop and we know how that panned out.

How did that pan out? To be the faster CPU for productivity workloads? Because that's the case, and people use PCs for other things too, not just to play games.

Quote from: Illrigger on February 04, 2025, 09:45:14Which just shows us how useless single core benchmarks are in 2025, when most apps will use as many cores as you have.

Single thread is the most beneficial for at least 80% of all non-gaming workloads, and it's easy to verify that single core is being used by simply opening a task manager if you don't want to measure more precisely with HWiNFO's graphs. There's a reason why Threadrippers and Xeons are barely usable at all for consumers, as despite their tremendous multi core performance, their single thread is quite weak.

You're pretending most workstation tasks aren't parallelized to hell? Brainlet.

Take a Threadripper and use it for a heavily single-threaded app like Photoshop against something with very fast single thread, say Apple M4. Enjoy.

Quote from: Worgarthe on February 04, 2025, 10:23:44

Quote from: Illrigger on February 04, 2025, 09:45:14The desktop parts also beat AMD in single core on the desktop and we know how that panned out.

How did that pan out? To be the faster CPU for productivity workloads? Because that's the case, and people use PCs for other things too, not just to play games.

Quote from: Illrigger on February 04, 2025, 09:45:14Which just shows us how useless single core benchmarks are in 2025, when most apps will use as many cores as you have.

Single thread is the most beneficial for at least 80% of all non-gaming workloads, and it's easy to verify that single core is being used by simply opening a task manager if you don't want to measure more precisely with HWiNFO's graphs. There's a reason why Threadrippers and Xeons are barely usable at all for consumers, as despite their tremendous multi core performance, their single thread is quite weak.

You're pretending most workstation tasks aren't parallelized to hell? Brainlet.

Quote from: Illrigger on February 04, 2025, 09:45:14The desktop parts also beat AMD in single core on the desktop and we know how that panned out.

How did that pan out? To be the faster CPU for productivity workloads? Because that's the case, and people use PCs for other things too, not just to play games.

Quote from: Illrigger on February 04, 2025, 09:45:14Which just shows us how useless single core benchmarks are in 2025, when most apps will use as many cores as you have.

Single thread is the most beneficial for at least 80% of all non-gaming workloads, and it's easy to verify that single core is being used by simply opening a task manager if you don't want to measure more precisely with HWiNFO's graphs. There's a reason why Threadrippers and Xeons are barely usable at all for consumers, as despite their tremendous multi core performance, their single thread is quite weak.

Quote from: puiu on February 04, 2025, 07:39:33

Quote from: cantstandstupidpeople on February 03, 2025, 20:17:38

Quote from: Serhii on February 03, 2025, 19:25:59In reality, not sponsored by Intel author, must has writing that Ryzen 370 by 21% more powerful than new Intel chip, as it clearly shows on comparison

Do you even understand what single thread means?

not much since the desktop chips also had high synthetic scores.

The desktop parts also beat AMD in single core on the desktop and we know how that panned out. Which just shows us how useless single core benchmarks are in 2025, when most apps will use as many cores as you have.

Besides, comparing to Strix Point when Fire Ridge is starting to ship is kinda pointless to begin with.

Quote from: cantstandstupidpeople on February 03, 2025, 20:17:38

Quote from: Serhii on February 03, 2025, 19:25:59In reality, not sponsored by Intel author, must has writing that Ryzen 370 by 21% more powerful than new Intel chip, as it clearly shows on comparison

Do you even understand what single thread means?

not much since the desktop chips also had high synthetic scores.

Quote from: Worgarthe on February 03, 2025, 23:24:50How would single thread use 50W? It uses 20-25W at most. So yes, it would still win in this case with 50W cap, especially given the already high difference.

Edit: Just ran a quick single thread test in Cinebench R23, my i7 14700HX consumes 35-37W, CBR23 score 2092. And that's a VERY power-inefficient CPU :D

Well, if it manages to keep the high performance of single-thread when the CPU is limited to 50W... then great for Intel.
I'm assuming that if the CPU is limited to 50W, the single-thread performance is also reduced. If it can run 50W on single-core, and also 50W on multi-core... then I'd be impressed.

N3B vs N4P. Hmmm...

Quote from: heffeque on February 03, 2025, 23:10:07Now do the same benchmark with CPUs set to consume a maximum of real 50W.
Let's see if it still trounces the 370 in single-thread or not.

How would single thread use 50W? It uses 20-25W at most. So yes, it would still win in this case with 50W cap, especially given the already high difference.

Edit: Just ran a quick single thread test in Cinebench R23, my i7 14700HX consumes 35-37W, CBR23 score 2092. And that's a VERY power-inefficient CPU :D

Now do the same benchmark with CPUs set to consume a maximum of real 50W.
Let's see if it still trounces the 370 in single-thread or not.

Quote from: Serhii on February 03, 2025, 19:25:59In reality, not sponsored by Intel author, must has writing that Ryzen 370 by 21% more powerful than new Intel chip, as it clearly shows on comparison

Do you even understand what single thread means?

In reality, not sponsored by Intel author, must has writing that Ryzen 370 by 21% more powerful than new Intel chip, as it clearly shows on comparison