News Intel's New Application Optimizer Yields Up to 31% Higher Frame Rates On i9-14900K

Status
Not open for further replies.
The only dependency they list on the Intel website is DTT which is available for everything 11th gen and newer with Xe based graphics (so for desktop this would be 12th-14th gen).

If it doesn't continue getting expanded over time to cover at least Raptor Lake K SKUs we'll know it's because they don't want to not that it won't work:
Why did Intel only choose to enable Intel® Application Optimization on select 14th Gen processors?
Settings within Intel® Application Optimization are custom determined for each supported processor, as they consider the number of P-cores, E-cores, and Intel® Hyperthreading Technology. Due to the massive amount of custom testing that went into the optimized setting parameters specifically gaming applications, Intel chose to align support for our gaming-focused processors.
 
Someone give me a prod when there's an example that serves some kind of useful purpose in the real world, such as pushing 50 to 60+ for example.

On the other hand, going from 659 FPS to 867 FPS is useful for marketing purposes I guess ...
 
  • Like
Reactions: P.Amini
Someone give me a prod when there's an example that serves some kind of useful purpose in the real world, such as pushing 50 to 60+ for example.

On the other hand, going from 659 FPS to 867 FPS is useful for marketing purposes I guess ...
These are examples of real world improvement, the same technique can be used for any and all examples of performance degradation that people had from the e-cores, where everybody would tell you to just disable them from the bios, now you can keep them and still have full performance.
It's just easier to get a final number if your examples are games with countable FPS.
 
Would be very interesting to understand how it is locked to 14th gen and thread director. but I have been running 12th and 13th without thread director (only Windows 11 compatible). I like the direct optimization work being done, allowing programs to interact more intelligently with the platform closer to metal. I think there is lots of room for efficiency improvement in the modern software stack and APO seems to prove that point is true.
 
Interesting stuff.
I wonder if there is any reason for AMD to not do the same thing.
Other than they don't have the money to pull it off no there is no reason whatsoever. This is basically what sense MI was supposed to be, which amd heavily used to promote zen and then did nothing to make it happen and nobody so much as raised an eyebrow at amd completely lying to people.
So... Process Lasso, disabling HT/Ecores and higher priority threads.

Got it.

Thanks for revolutionizing the market with 14th gen, Intel. This will show AMD how its done, for sure. And yes, this is sarcasm.

Regards,
At least they are doing something actively instead of just waiting for the devs to do everything for them.
Also if peoples argument for intel's high power draw is that people are too dumb or just can't be bothered to change PL settings then yes, this is a big deal because such a person will definitely be too dumb or won't be bothered to figure out process lasso, especially making settings for every game individually.
 
  • Like
Reactions: KyaraM
At least they are doing something actively instead of just waiting for the devs to do everything for them.
Also if peoples argument for intel's high power draw is that people are too dumb or just can't be bothered to change PL settings then yes, this is a big deal because such a person will definitely be too dumb or won't be bothered to figure out process lasso, especially making settings for every game individually.
Reading that, made me chuckle, Terry. Thanks.

I'll add then: I applaud Intel for introducing user-dependent software crutches to help fix their incompetence to implement a better, invisible, scheduler to the user.

Ah, the times when people gave AMD a lot of crap for toying around the power profiles to fix performance, but Intel can do whatever to fix their incompetence and it's received with open arms. "Oh, but it's different!". Yeah, ok, sure.

Regards.
 
  • Like
Reactions: bit_user
Couldn't microsoft just use their silly xbox gamebar app to do the same thing. I know when 7950x3d came out there was lots of discussion of having to use this software to best use these cpu. Then again maybe it does I tend to kill off any processes that I can't see a need for when I run games.
 
These are examples of real world improvement, the same technique can be used for any and all examples of performance degradation that people had from the e-cores, where everybody would tell you to just disable them from the bios, now you can keep them and still have full performance.
It's just easier to get a final number if your examples are games with countable FPS.
All I can say to that is we will see. These improvements here are real and measurable, but in those 2 particular examples they serve no useful purpose if the best gains are to be had only from games that already run at a million FPS. I will be far more interested in games that are heightened from jittery to smooth using this technique.

It reminds me a little bit of how the best results from using frame generation with GPUs is when you already have fairly high performance.

Feel free to brush off my mutterings; being super-cynical is my thing.
 
  • Like
Reactions: P.Amini
Well, how exactly is a CPU running much lower than they should clocks on a lower power profile anything like directing the threads to where they should go?!
Someone mentioned the gamebar above, so there's another one if the context of "software crutches" is hard to grasp.

I still remember AMD had to put kernel patches and a CPU driver for Windows when the first Athlon64 X2's hit the market back when Windows only knew about HT (barely). Maybe I'm missing some details on this, but it was more or less a patch for the CPU and it didn't need user intervention (other than installing it) to get more performance out of the CPUs.

Regards.
 
Someone mentioned the gamebar above, so there's another one if the context of "software crutches" is hard to grasp.

I still remember AMD had to put kernel patches and a CPU driver for Windows when the first Athlon64 X2's hit the market back when Windows only knew about HT (barely). Maybe I'm missing some details on this, but it was more or less a patch for the CPU and it didn't need user intervention (other than installing it) to get more performance out of the CPUs.

Regards.
But what you said makes zero sense.

Disabling some cores is not the same as "lowering frequency".

Even less with the Athlon 64 X2.. because they were of the first dual cores ever to slot in a consumer OS.
 
  • Like
Reactions: KyaraM
when they realize they can use process lasso... XD
This is definitely doing something at a lower level than what process lasso is achieving, Does appear to be a bit kludgy though as Intel will need to work with each dev to utilize new code to enable this optimizer. You can not recreate these gains with process lasso, all process lasso does well is to help manually direct threads where you want them(Do not run on e-cores, utilze only P, or utilize all cores).

These gains are impressive, would think there would be more appetite for seeing these types of optimizations as they appear to have legs and are well worth the effort.
 
This is definitely doing something at a lower level than what process lasso is achieving, Does appear to be a bit kludgy though as Intel will need to work with each dev to utilize new code to enable this optimizer. You can not recreate these gains with process lasso, all process lasso does well is to help manually direct threads where you want them(Do not run on e-cores, utilze only P, or utilize all cores).

These gains are impressive, would think there would be more appetite for seeing these types of optimizations as they appear to have legs and are well worth the effort.
driver level instead of software/service level?
 
driver level instead of software/service level?
Yes apparently so, Intel has this on the only application info page I have seen on this:
How will an end user know that Intel® Application Optimization is running/working?
Intel Application Optimization will be enabled by default on a supported system that is running Intel® DTT, it runs silently as part of the Intel® DTT driver.
Missed opportunity to include on 14600k, appears to definitely be arbitrary differentiating by SKU.

Originally I thought this might be an actual hardware change, perhaps on the packaging or some feature not fused off, I don't believe this anymore and think it is purely if cpu-id=x enable feature... SO someone might hack in the feature to enable 12/13/14 gen support in windows 11, would love to see if there is a way to enable the gains in Windows 10 but as thread director does not exist there, more of a stretch.

Its just like Optane, worked perfectly with 12th gen (H10 hybrid nand optane device). But when I upgraded to 13th gen on platform, gooddbye optane acceleration, would not boot properly, had to recover to find out it purposely broke the ability. Intel also does things differently, NAND division same way where the drive goes to read only mode after warranty writes level reached. Not a fan but they are consistent...
 
Status
Not open for further replies.