News Cyberpunk 2077 adds core prioritization for hybrid CPUs, which would be great if it didn't cause other problems

[TheHerald]

You have to look at benches that show both power and scores in the same bench, otherwise you get confused, you can see here with the 13900k that the 7950x wins by 10-13% at every power level down to 65W and the 13900k only wins at 45W by about 10% because that's when the e-cores are finally increasing the power efficiency instead of just increasing the performance.
Their p-cores are more efficient than zen but as a whole CPU intel is less efficient overall in multithreading.
https://www.computerbase.de/2023-10...i5-14600k-test/2/#abschnitt_multicoreleistung

Or here as well, the 14900k limited to 200W draws 141W average compared to the 128W average of the 7950x while having the same performance, while the default draw of the 14900k is 170W average compared to the 128W average.
(Application 47 test average)
https://www.techpowerup.com/review/...ke-tested-at-power-limits-down-to-35-w/8.html

I have, the 7950x is a little bit more efficient than the 14900k at all core workloads. Maybe not all of them, but in general it has the upper hand. They are both incredibly efficient though for heavy workloads, set them to 150-200w power limit and they are basically untouchable.

 

[brandonjclark]

As someone who helped work to bring this patch to the game due to my ongoing issues with crashes on Cyberpunk 2077 with an i9-13900k, allow me to explain:

Many, many modern games, including Returnal, Remnant from the Ashes 2, and a slew of other modern games, do not know how to schedule and prioritize the workload effectively across P and E cores. You will get either random crashes or error notifications about a "lack of video memory," even while running a 4090.

These crashes are very unfortunate, and the fix remains constant across all these games, including Cyberpunk 2077: A manual downclock of about 200-300mhz across all P cores done through Intel Extreme Tuning Utility (Intel XTU).

Now, with these fixes in place, this enables users to play the game with the full processing power of their 12th, 13th, and 14th gen processors, without thermal throttling, poor performance, crashes, and without having to run an entire separate program just to downclock their CPUs.

If it introduced stuttering, that is unfortunate, but Cyberpunk deserves credit for being the only game so far to even acknowledge this widespread problem exists, let alone implement an actual solution to fix it. I am sure further tuning and tweaking will be needed to prevent micro stuttering other users are claiming to have since patch 2.11, but simply turning off the Hybrid CPU Utilization option, or leaving it to "auto" should restore behavior and performance to how it was in the prior patch.

If you just do a brief Google search for "i9 13900k out of video memory," you can find many, many such users across MANY modern games with this exact same problem, being resolved in this exact same way.

Again, the Cyberpunk team deserves credit for not only acknowledging this issue exists, but actually implementing a fix, even if it's imperfect.

Hrmm, I'm playing the hell out of Returnal right now (on an Intel 10th gen 10900kf) and I don't think I've ran into any leak issues. Must be lucky or it doesn't affect 10th gen. I also have an RTX3090.

 

[TheHerald]

Hrmm, I'm playing the hell out of Returnal right now (on an Intel 10th gen 10900kf) and I don't think I've ran into any leak issues. Must be lucky or it doesn't affect 10th gen. I also have an RTX3090.

I've played around 180 hours of cyberpunk - on both 12900k, 13900k and 14900k. Never noticed any of the issues he is mentioning.

 

[JamesJones44]

Feels like this should just be a setting in DirectX and other gaming focused APIs to simply say "don't use non-performance cores". Seems crazy they are expecting every game to come up with their own scheduler solution. For some games it may make sense to priorities one type over another but for most games having a switch is likely the right solution.

 

[Pierce2623]

Nope, it won't make any difference nor will it work on AMD's hybrid CPUs either, because the Zen 4 and 4c cores don't behave in the same way like Intel's P and E cores. P and E cores are based on different architectures.

You guys must already be aware by now that ZEN 4c cores instead feature an identical IPC to "Zen 4," as well as an identical ISA. So, that's the same execution and issue width, the same number of registers, the same internal latencies, including multi-threading.

"Zen 4c" CPU core is just a compacted version of the "Zen 4" core without the subtraction of any hardware components, but rather a high density arrangement of them, which are generally clocked lower than "Zen 4" cores, as they can operate at lower core voltages.

BTW, it appears that there's much more asymmetry between Intel's core types, like differences in issue width, latencies and even instruction set. And more importantly, Intel's Efficient cores do not support multi-threading.

So there seems to be some complication to manage thread allocation and load balancing on Intel's hybrid design.

Just for future reference, Intel’s P cores are also about to drop multithreading completely. They’ve decided e cores are the new way to build multithreaded performance because they can achieve a given Cinebench score with a smaller die area, they want you to ignore the fact that the e cores only really benefit them instead of the customers.

 

[TerryLaze]

Feels like this should just be a setting in DirectX and other gaming focused APIs to simply say "don't use non-performance cores". Seems crazy they are expecting every game to come up with their own scheduler solution. For some games it may make sense to priorities one type over another but for most games having a switch is likely the right solution.

That would be the worst thing ever, the e-cores can add a huge amount of performance to almost any game, the trick is to know what type of work, or how much of it, to send to the e cores and which ones to the p cores.

Just for future reference, Intel’s P cores are also about to drop multithreading completely. They’ve decided e cores are the new way to build multithreaded performance because they can achieve a given Cinebench score with a smaller die area, they want you to ignore the fact that the e cores only really benefit them instead of the customers.

It's intel and it's possible, but I don't think that they will drop HT since it adds a good amount of performance to cinebench.
It wouldn't make any sense to add a way for more performance and to nullify that by taking away a way to gain performance.
HT has a very small footprint in size and power while adding a good 30% to the CB score of the p-cores.

 

[digitalgriffin]

It makes perfect sense if you think like a company for a second, why would you put more than 8 very expensive cores in a CPU if you can instead fill it up with cheapo cores?
When all cores are working they all have to power down and clock lower so why not have your standard 8 cores and the rest can be cheap cores that always perform as if the whole CPU where under full load, because that's the only time they would do any work anyway. (as far as benchmarks are concerned anyway)

That is what AMD discovered now as well, after a few years of watching intel doing it and thought hey, why aren't we doing that as well, and hey presto c versions, because they can't afford a completely new design.

It takes 2+ years for a team to design a CPU. And at least 1+ previous years to set design targets. I don't think AMD was two years late to the party were they? When the Intel 12 series came out, I'm positive AMD was already working on it. And AMD did come up with a better solution.

 

[TerryLaze]

It takes 2+ years for a team to design a CPU. And at least 1+ previous years to set design targets. I don't think AMD was two years late to the party were they? When the Intel 12 series came out, I'm positive AMD was already working on it. And AMD did come up with a better solution.

But AMD didn't design a new CPU, or new cores, they just used TSMC's option to make the same design denser which forces it to use less power and run at lower clocks.

(Area reduction and low voltage)
https://www.anandtech.com/show/1272...-scaling-but-thin-power-and-performance-gains

 

[JamesJones44]

That would be the worst thing ever, the e-cores can add a huge amount of performance to almost any game, the trick is to know what type of work, or how much of it, to send to the e cores and which ones to the p cores.

Most games don't use that many CPU threads. They are largely GPU bound and game that are CPU bound do not typically use more than 8 cores (there are some exceptions like MS Flight Simulator, but those are not normal). Many games are still based on a single game loop methodology and simply offload non-time bound actions to background threads. However, your non-time related are not critical operations (hud updates, stats, multi player cheat detection, etc.). The single game loop is how games maintain frame consistency. This is very true for FPS style games like Cyberpunk. The single game loop does not mean they can't use threads, but it limits opportunities for them to be used, they will offload some work to CPU cores and then wait for all work to be complete before entering the next loop (or wait frame time to timeout and cancel the non-complete threads), but this has limits. This is why may games don't load out a 16 core CPU and a 7950x is less valuable for most games than an 8 core 7800x3d with a lot of cache and the 7800x3d ends up being the better CPU for gaming.

 

[rluker5]

Most games don't use that many CPU threads. They are largely GPU bound and game that are CPU bound do not typically use more than 8 cores (there are some exceptions like MS Flight Simulator, but those are not normal). Many games are still based on a single game loop methodology and simply offload non-time bound actions to background threads. However, your non-time related are not critical operations (hud updates, stats, multi player cheat detection, etc.). The single game loop is how games maintain frame consistency. This is very true for FPS style games like Cyberpunk. The single game loop does not mean they can't use threads, but it limits opportunities for them to be used, they will offload some work to CPU cores and then wait for all work to be complete before entering the next loop (or wait frame time to timeout and cancel the non-complete threads), but this has limits. This is why may games don't load out a 16 core CPU and a 7950x is less valuable for most games than an 8 core 7800x3d with a lot of cache and the 7800x3d ends up being the better CPU for gaming.

There's a guy who tested that:

View: https://youtu.be/I8DJITHWdaA

View: https://youtu.be/LcQUUmi3rWI

The long and short of it is that it depends on the game, but with a 13900k e-cores generally help and HT generally hurts.

Also that is with a 13900k that has ample threads, so having either of the weaker sets of threads enabled has less benefit than with a 4c8t or 6c12t CPU.

 

[TheHerald]

The long and short of it is that it depends on the game, but with a 13900k e-cores generally help and HT generally hurts.

Also that is with a 13900k that has ample threads, so having either of the weaker sets of threads enabled has less benefit than with a 4c8t or 6c12t CPU.

Yeap, ecores help in gaming, sometimes by a lot in 1% lows. Warzone with ecores off 1% lows = ~120, ecores on = closer to 200. That's with a 12900k.