• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Admits "Stars" in Ryzen Master Don't Correspond to CPPC2 Preferred Cores

Solaris17

Super Dainty Moderator
Staff member
Joined
Aug 16, 2005
Messages
26,925 (3.83/day)
Location
Alabama
System Name RogueOne
Processor Xeon W9-3495x
Motherboard ASUS w790E Sage SE
Cooling SilverStone XE360-4677
Memory 128gb Gskill Zeta R5 DDR5 RDIMMs
Video Card(s) MSI SUPRIM Liquid X 4090
Storage 1x 2TB WD SN850X | 2x 8TB GAMMIX S70
Display(s) 49" Philips Evnia OLED (49M2C8900)
Case Thermaltake Core P3 Pro Snow
Audio Device(s) Moondrop S8's on schitt Gunnr
Power Supply Seasonic Prime TX-1600
Mouse Lamzu Atlantis mini (White)
Keyboard Monsgeek M3 Lavender, Moondrop Luna lights
VR HMD Quest 3
Software Windows 11 Pro Workstation
Benchmark Scores I dont have time for that.
huh? I dont get alot of this thread. Microsoft is not responsible for writing the scheduler around a particular manufacturer. If a host device wants something a certain way you need to include that in the driver.
 

bug

Joined
May 22, 2015
Messages
13,755 (3.96/day)
Processor Intel i5-12600k
Motherboard Asus H670 TUF
Cooling Arctic Freezer 34
Memory 2x16GB DDR4 3600 G.Skill Ripjaws V
Video Card(s) EVGA GTX 1060 SC
Storage 500GB Samsung 970 EVO, 500GB Samsung 850 EVO, 1TB Crucial MX300 and 2TB Crucial MX500
Display(s) Dell U3219Q + HP ZR24w
Case Raijintek Thetis
Audio Device(s) Audioquest Dragonfly Red :D
Power Supply Seasonic 620W M12
Mouse Logitech G502 Proteus Core
Keyboard G.Skill KM780R
Software Arch Linux + Win10
huh? I dont get alot of this thread. Microsoft is not responsible for writing the scheduler around a particular manufacturer. If a host device wants something a certain way you need to include that in the driver.
That's just common sense. It does not apply when one party is the darling of the internet ;)
 
Joined
Aug 14, 2009
Messages
216 (0.04/day)
Location
Denmark
System Name Bongfjaes
Processor AMD 3700x
Motherboard Assus Crosshair VII Hero
Cooling Dark Rock Pro 4
Memory 2x8GB G.Skill FlareX 3200MT/s CL14
Video Card(s) GTX 970
Storage Adata SX8200 Pro 1TB + Lots of spinning rust
Display(s) Viewsonic VX2268wm
Case Fractal Design R6
Audio Device(s) Creative SoundBlaster AE-5
Power Supply Seasonic TTR-1000
Mouse Pro Intellimouse
Keyboard SteelKeys 6G
I never said that I want to feel, the way you feel.
Also I never said that I would like to describe myself, the way you described yourself.
Well, i did get bamboozled by buying it /:

Life goes on, if you have no sense of humor its already over! :)
 
Joined
Dec 22, 2011
Messages
289 (0.06/day)
Processor Ryzen 7 5800X3D
Motherboard Asus Prime X570 Pro
Cooling Deepcool LS-720
Memory 32 GB (4x 8GB) DDR4-3600 CL16
Video Card(s) PowerColor Radeon RX 7900 XTX Red Devil
Storage Samsung PM9A1 (980 Pro OEM) + 960 Evo NVMe SSD + 830 SATA SSD + Toshiba & WD HDD's
Display(s) Samsung C32HG70
Case Lian Li O11D Evo
Audio Device(s) Sound Blaster Zx
Power Supply Seasonic 750W Focus+ Platinum
Mouse Logitech G703 Lightspeed
Keyboard SteelSeries Apex Pro
Software Windows 11 Pro
"admits"? Really? @btarunr you're trying to make it sound like AMD has tried to fool people or has done something to hide the reality or some other such, why?
They've told exactly what the "Stars" etc mean and show, and that hasn't changed (until the update that changes them to show what Windows sees as preferred cores) - they show the highest clocking core of the CPU, and 2 best clocking cores of each CCX based on their properties.
 

1usmus

AMD Memory Guru
Joined
Jul 28, 2018
Messages
61 (0.03/day)
Location
Ukraine
"admits"? Really? @btarunr you're trying to make it sound like AMD has tried to fool people or has done something to hide the reality or some other such, why?
They've told exactly what the "Stars" etc mean and show, and that hasn't changed (until the update that changes them to show what Windows sees as preferred cores) - they show the highest clocking core of the CPU, and 2 best clocking cores of each CCX based on their properties.

this is not always the case, unfortunately
 
Joined
Dec 30, 2010
Messages
2,198 (0.43/day)
Looking at the way it's programmed to switch in between cores, these CPU's are fragile. It's just not all about thermal management, but pretty much degradation. Look at 2x00 series. Many reviewers just boot a 1.4V into a CPU and hit a OC on that. The 1.4V is seen to degrade the CPU in just months (!).
 
Joined
Dec 29, 2010
Messages
3,809 (0.75/day)
Processor AMD 5900x
Motherboard Asus x570 Strix-E
Cooling Hardware Labs
Memory G.Skill 4000c17 2x16gb
Video Card(s) RTX 3090
Storage Sabrent
Display(s) Samsung G9
Case Phanteks 719
Audio Device(s) Fiio K5 Pro
Power Supply EVGA 1000 P2
Mouse Logitech G600
Keyboard Corsair K95
If this were Intel, the internet would Explode. Instead, its "thank you for much for this news" :laugh: :kookoo:

Has the internet exploded due to all the Intel security flaws? Where are the class action lawsuits? :rolleyes:
 
Joined
Oct 21, 2005
Messages
7,057 (1.01/day)
Location
USA
System Name Computer of Theseus
Processor Intel i9-12900KS: 50x Pcore multi @ 1.18Vcore (target 1.275V -100mv offset)
Motherboard EVGA Z690 Classified
Cooling Noctua NH-D15S, 2xThermalRight TY-143, 4xNoctua NF-A12x25,3xNF-A12x15, 2xAquacomputer Splitty9Active
Memory G-Skill Trident Z5 (32GB) DDR5-6000 C36 F5-6000J3636F16GX2-TZ5RK
Video Card(s) ASUS PROART RTX 4070 Ti-Super OC 16GB, 2670MHz, 0.93V
Storage 1x Samsung 970 Pro 512GB NVMe (OS), 2x Samsung 970 Evo Plus 2TB (data), ASUS BW-16D1HT (BluRay)
Display(s) Dell S3220DGF 32" 2560x1440 165Hz Primary, Dell P2017H 19.5" 1600x900 Secondary, Ergotron LX arms.
Case Lian Li O11 Air Mini
Audio Device(s) Audiotechnica ATR2100X-USB, El Gato Wave XLR Mic Preamp, ATH M50X Headphones, Behringer 302USB Mixer
Power Supply Super Flower Leadex Platinum SE 1000W 80+ Platinum White, MODDIY 12VHPWR Cable
Mouse Zowie EC3-C
Keyboard Vortex Multix 87 Winter TKL (Gateron G Pro Yellow)
Software Win 10 LTSC 21H2
"admits"? Really? @btarunr you're trying to make it sound like AMD has tried to fool people or has done something to hide the reality or some other such, why?
They've told exactly what the "Stars" etc mean and show, and that hasn't changed (until the update that changes them to show what Windows sees as preferred cores) - they show the highest clocking core of the CPU, and 2 best clocking cores of each CCX based on their properties.
Its typical btarunr writing, it is sensational, there's melodrama to it, with a hero and a villain, though the reality is often quite dry and undramatic.
 
Last edited:
Joined
Oct 1, 2006
Messages
4,931 (0.74/day)
Location
Hong Kong
Processor Core i7-12700k
Motherboard Z690 Aero G D4
Cooling Custom loop water, 3x 420 Rad
Video Card(s) RX 7900 XTX Phantom Gaming
Storage Plextor M10P 2TB
Display(s) InnoCN 27M2V
Case Thermaltake Level 20 XT
Audio Device(s) Soundblaster AE-5 Plus
Power Supply FSP Aurum PT 1200W
Software Windows 11 Pro 64-bit
Lmao, with every passing day i feel more and more scammed for buying a 3700x and Crosshair VII x470 board

You dumb asshole, Axaion.
That if they cant even get their own software right, then what hope is there?
AMD software maybe right but what matters just as much is microsofts software and how it works AMD software into windows. If AMD and m$ dont work together on the integration of their software with each others software, than when issue rise, it comes down to finger pointing. Its no secret that m$ doesnt appear to like AMD as AMD loves to pass around a headache or two.
huh? I dont get alot of this thread. Microsoft is not responsible for writing the scheduler around a particular manufacturer. If a host device wants something a certain way you need to include that in the driver.
https://www.techpowerup.com/261094/microsoft-releases-windows-10-november-2019-update-1909
Only as recently that M$ finally optimized for Turbo Boost Max 3, which was launched with the X299 platform from 2017.
I think people are assuming too much here, M$ is just incompetent at fixing their Windows Scheduler.

It also took years after Intel Hyperthreading was introduced before M$ make Windows aware of HT.
IDK why it takes so long for this to become "breaking news". Windows 10 contains a lot of legacy code, some goes all the way to Win Vista.
 
Joined
May 30, 2018
Messages
1,890 (0.80/day)
Location
Cusp Of Mania, FL
Processor Ryzen 9 3900X
Motherboard Asus ROG Strix X370-F
Cooling Dark Rock 4, 3x Corsair ML140 front intake, 1x rear exhaust
Memory 2x8GB TridentZ RGB [3600Mhz CL16]
Video Card(s) EVGA 3060ti FTW3 Ultra Gaming
Storage 970 EVO 500GB nvme, 860 EVO 250GB SATA, Seagate Barracuda 1TB + 4TB HDDs
Display(s) 27" MSI G27C4 FHD 165hz
Case NZXT H710
Audio Device(s) Modi Multibit, Vali 2, Shortest Way 51+ - LSR 305's, Focal Clear, HD6xx, HE5xx, LCD-2 Classic
Power Supply Corsair RM650x v2
Mouse iunno whatever cheap crap logitech *clutches Xbox 360 controller security blanket*
Keyboard HyperX Alloy Pro
Software Windows 10 Pro
Benchmark Scores ask your mother
Looking at the way it's programmed to switch in between cores, these CPU's are fragile. It's just not all about thermal management, but pretty much degradation. Look at 2x00 series. Many reviewers just boot a 1.4V into a CPU and hit a OC on that. The 1.4V is seen to degrade the CPU in just months (!).
Is that really the case though? All I've seen are a handful of people saying things like "it's been proven." But I see many more people running them around or even a little beyond 1.4... with very few, if any definitive reports of degradation. I'd think if they were really so fragile we'd have more concrete reports of chips dying off. But if they're out there, I can't find em! Only a few sporadic accounts from people who think that's what happened to them. Which, maybe it did! But it reminds me of the RTX memory failure issue. Something that people say is a thing that's happening a lot, when the real number was probably still pretty low, and it was just one big, freak occurrence.

I could also go after the claim of switching cores being a sign of known fragility. It could just as easily be thermal and there's no logical reason to go one way or the other, because the longer a core is loaded down, the hotter it gets. It's one of those nebulous things. If you want it to be for degradation, there's a rationale. If you want it to be temperatures, there's also a rationale. Nobody but AMD really knows.

Reviewers pumping 1.4v in for their OC testing and calling it a day... yeah, not a fan of that myself. Probably encouraging a good chunk of less experienced builders to run way higher voltage than they need. 1.4 is too much, and there's a good chance, at least with the 12nm and 14nm, that you will see eventual degradation. That's undeniable. AMD itself has recommended not going over 1.35 for a continuous-use all-core, IIRC.

Whether that makes them weak or unreliable, I'm not sure. How many modern CPU's are really expected to run at 1.4v continuously? I'd think that most people with some experience from either side of the fence would say that's pretty high.

That aside, the way Ryzen 3000 currently jumps from core to core, as far as anybody can tell right now, isn't quiiite the intended behavior. I'd say the reason for what they're doing with this boost stuff it is to squeeze the max performance out for the least power. I'm not talking theory, here. It's just my conclusion looking at what's going down and what the outcome is. I mean, I have played with a couple handfuls of these Ryzen chips, now. All generations. I'm not an expert and don't know nearly as much as many of the regulars here, but from what I've seen a MAX (as in, approaching or fully hitting unsafe voltage) all-core OC is pretty much always just a *little* inferior in performance to an X model attaining max boost will pull off, for less power. For instance, I have a 2600 that will do an all-core of 4.2 steady.. at around 1.3v. I can even push it past that if I'm willing to go to ~1.4v... BUT, even doing that, it never benched as high as its X-model sibling and the thermals were just impossible. That seems to be how it goes, and is a big part of the prevailing wisdom that it's best to let it boost naturally. And that's still the consensus, because it just performs better.

So I see what you're saying - you can easily cook a Ryzen running the voltage too high, but I'm much more willing to believe it's for max performance and thermals primarily, as they actually do run better when you run them the way AMD wants them.


I don't get why people get up in arms about this stuff. It's annoying, sure, but at this point it's to be expected. AMD's boost system is both very advanced and very young. It's going to be a while before they get it right. In the meantime, the Ryzen 3000 line still has some of the best-performing CPU's on the market. It's not like they're completely broken by these little quirks. It's nothing like the whole bulldozer situation where people completely didn't get what they thought they were getting and the performance wasn't what it should've been. You're still getting a good, working CPU for a good price - it just boosts funny lol.
 
Joined
Dec 30, 2010
Messages
2,198 (0.43/day)

It happens. It's the reason why i am not even touching a manual overclock on my 2700x at all. The degradation usually occurs when going around or beyond 1.4V. Some say it cant even hold any boost anymore without crashing. AMD said before that as long as the current usage of a core is low, the voltage can go up on light weight threads. Once you have a heavy workload the voltage and core clocks go down, indicating that current is a factor and can kill these CPU's on long term base. If you consider on how windows switches in between one and two cores inside a unit, AMD states because of thermals but i personally think these chips are fragile to degradation in a fast way.

24/7 safe voltages should never be exceeded on these chips. This aint the Bulldozer FX that could eat 1.65V all day from 1.32V or so.
 
Joined
Aug 9, 2006
Messages
1,065 (0.16/day)
System Name [Primary Workstation]
Processor Intel Core i7-920 Bloomfield @ 3.8GHz/4.55GHz [24-7/Bench]
Motherboard EVGA X58 E758-A1 [Tweaked right!]
Cooling Cooler Master V8 [stock fan + two 133CFM ULTRA KAZE fans]
Memory 12GB [Kingston HyperX]
Video Card(s) constantly upgrading/downgrading [prefer nVidia]
Storage constantly upgrading/downgrading [prefer Hitachi/Samsung]
Display(s) Triple LCD [40 inch primary + 32 & 28 inch auxiliary displays]
Case Cooler Master Cosmos 1000 [Mesh Mod, CFM Overload]
Audio Device(s) ASUS Xonar D1 + onboard Realtek ALC889A [Logitech Z-5300 Spk., Niko 650-HP 5.1 Hp., X-Bass Hp.]
Power Supply Corsair TX950W [aka Reactor]
Software This and that... [All software 100% legit and paid for, 0% pirated]
Benchmark Scores Ridiculously good scores!!!
1.4375V on a Ryzen 1600X since August 2017. Holding clocks just fine. In fact with some BIOS updates I was able to hit new SuperPi and 3DMark 01/03 records just this past summer. That said, it is water cooled with a externally mounted 360mm radiator and 6x120mm fans, so nice thermals might play a part.

That said, I feel your pain Axain. Zen platform has issues galore. Every itteration of it so far. Performance wise, it has been over-hyped beyond belief (thanks Youtube "tech journalists"! How are those paid-for Taiwan vacations factory tours working out?! *cough*GamersNexus*cough*) In single threaded loads, you are really on par with Nehalem in some benchmarks, and anything using the x87 instruction set (like legacy gaming, benchmarks, and similar) and you might be more than a few % behind Nehalem even! Also, no way around it, even in 2019 Mhz’s do matter and AMD is wayyyy beyond in that race!

Anyway, all that aside, I like it. It is a very very very nice value CPU. (I still own a CPU from every AMD platform, from K5 to this Ryzen and same goes for Intel as well as a few Socket 370 Via Samuel's) For last-year gaming/benchmarking though, it managed to bottleneck my decade old GTX 580, a single one, in certain benchmarks. (I own three) Much less a single RX 580 (I own four), if we are going to go by 580-anything. Just not a good gaming platform outside modern gaming titles that can actually take advantage of all the cores.
 
Last edited:

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,600 (2.41/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Thanks for a much more sensible explanation compared to what Anandtech posted, as I couldn't make head or tails of what they were trying to say in their explanation.
That said, no need for the sensationalist headline.
 
Last edited:
Joined
Oct 10, 2009
Messages
937 (0.17/day)
System Name Desktop
Processor AMD Ryzen 7 5800X3D
Motherboard MAG X570S Torpedo Max
Cooling Corsair H100x
Memory 64GB Corsair CMT64GX4M2C3600C18 @ 3600MHz / 18-19-19-39-1T
Video Card(s) EVGA RTX 3080 Ti FTW3 Ultra
Storage Kingston KC3000 1TB + Kingston KC3000 2TB + Samsung 860 EVO 1TB
Display(s) 32" Dell G3223Q (2160p @ 144Hz)
Case Fractal Meshify 2 Compact
Audio Device(s) ifi Audio ZEN DAC V2 + Focal Radiance / HyperX Solocast
Power Supply Super Flower Leadex V Platinum Pro 1000W
Mouse Razer Viper Ultimate
Keyboard Razer Huntsman V2 Optical (Linear Red)
Software Windows 11 Pro x64
Make sense. The Windows scheduler is either too stupid and/or too focused on the Intel design philosophy to optimally load and manage threads within AMD's CCX architecture, so using a bunch of metrics such as, individual core performance, CCX performance, cache location, etc, they have the firmware/drivers fool Windows in an attempt to have the scheduler use the most suitable core.

The stars and dots are a ranking for the physical quality and performance of a core, whereas the CPPC2 'preferred cores' is a ranking based on the aforementioned metrics and are used to 'trick' the scheduler.
 
Joined
Jan 11, 2005
Messages
1,491 (0.21/day)
Location
66 feet from the ground
System Name 2nd AMD puppy
Processor FX-8350 vishera
Motherboard Gigabyte GA-970A-UD3
Cooling Cooler Master Hyper TX2
Memory 16 Gb DDR3:8GB Kingston HyperX Beast + 8Gb G.Skill Sniper(by courtesy of tabascosauz &TPU)
Video Card(s) Sapphire RX 580 Nitro+;1450/2000 Mhz
Storage SSD :840 pro 128 Gb;Iridium pro 240Gb ; HDD 2xWD-1Tb
Display(s) Benq XL2730Z 144 Hz freesync
Case NZXT 820 PHANTOM
Audio Device(s) Audigy SE with Logitech Z-5500
Power Supply Riotoro Enigma G2 850W
Mouse Razer copperhead / Gamdias zeus (by courtesy of sneekypeet & TPU)
Keyboard MS Sidewinder x4
Software win10 64bit ltsc
Benchmark Scores irrelevant for me
windows scheduler may not be good but is taking the info from cpu; i agree with solaris that driver should give instructions to the scheduler and not vice-versa...
 
Joined
Aug 6, 2017
Messages
7,412 (2.78/day)
Location
Poland
System Name Purple rain
Processor 10.5 thousand 4.2G 1.1v
Motherboard Zee 490 Aorus Elite
Cooling Noctua D15S
Memory 16GB 4133 CL16-16-16-31 Viper Steel
Video Card(s) RTX 2070 Super Gaming X Trio
Storage SU900 128,8200Pro 1TB,850 Pro 512+256+256,860 Evo 500,XPG950 480, Skyhawk 2TB
Display(s) Acer XB241YU+Dell S2716DG
Case P600S Silent w. Alpenfohn wing boost 3 ARGBT+ fans
Audio Device(s) K612 Pro w. FiiO E10k DAC,W830BT wireless
Power Supply Superflower Leadex Gold 850W
Mouse G903 lightspeed+powerplay,G403 wireless + Steelseries DeX + Roccat rest
Keyboard HyperX Alloy SilverSpeed (w.HyperX wrist rest),Razer Deathstalker
Software Windows 10
Benchmark Scores A LOT
Explains not admits.
 
Joined
Jun 19, 2010
Messages
409 (0.08/day)
Location
Germany
Processor Ryzen 5600X
Motherboard MSI A520
Cooling Thermalright ARO-M14 orange
Memory 2x 8GB 3200
Video Card(s) RTX 3050 (ROG Strix Bios)
Storage SATA SSD
Display(s) UltraHD TV
Case Sharkoon AM5 Window red
Audio Device(s) Headset
Power Supply beQuiet 400W
Mouse Mountain Makalu 67
Keyboard MS Sidewinder X4
Software Windows, Vivaldi, Thunderbird, LibreOffice, Games, etc.
@DeathtoGnomes yes, i believe the intle CPUs with the mesh-architecture could gain an advantage if the windows scheduler would be aware of the core near or far from I/O or memory, even if it is monolitic.

The inherent goal of the windows scheduler has remained as a kind of time-slice and hierarchic priority planning until today, thats really old and comes out of the single-core era.
On top of that came the thermal and efficiency planning, wich looks like a kind of panicreaction microsoft made after the first dual-cores came out.
This legacy led to the scheduler making apps hop over the cores because of some insane temperature-abstraction-causes and to cores "more efficient" .. meaning lower clocked cores.
This legacy is now what gets manipulated by some kinds of "patches" and hidden settings and offsets in powerprofiles.
Microsoft eventually is not able to change the scheduler so courageous, how it would be needed by now. Thas very sad, i think.
The linux community is more courageous related to their scheduler.
 
Joined
Jan 8, 2017
Messages
9,434 (3.28/day)
System Name Good enough
Processor AMD Ryzen R9 7900 - Alphacool Eisblock XPX Aurora Edge
Motherboard ASRock B650 Pro RS
Cooling 2x 360mm NexXxoS ST30 X-Flow, 1x 360mm NexXxoS ST30, 1x 240mm NexXxoS ST30
Memory 32GB - FURY Beast RGB 5600 Mhz
Video Card(s) Sapphire RX 7900 XT - Alphacool Eisblock Aurora
Storage 1x Kingston KC3000 1TB 1x Kingston A2000 1TB, 1x Samsung 850 EVO 250GB , 1x Samsung 860 EVO 500GB
Display(s) LG UltraGear 32GN650-B + 4K Samsung TV
Case Phanteks NV7
Power Supply GPS-750C
And some applications, like CPU-Z in evaluating single-thread performance, aren't written to leverage any core except Core 0, regardless of power plan.

It's not the application that targets core 0, Windows does. This can't be fixed by AMD or anyone else except Microsoft, thread scheduling has always semi-broken even back when all CPUs had uniform topologies and clockspeeds.
 
Last edited:
Joined
Jun 2, 2017
Messages
9,127 (3.34/day)
System Name Best AMD Computer
Processor AMD 7900X3D
Motherboard Asus X670E E Strix
Cooling In Win SR36
Memory GSKILL DDR5 32GB 5200 30
Video Card(s) Sapphire Pulse 7900XT (Watercooled)
Storage Corsair MP 700, Seagate 530 2Tb, Adata SX8200 2TBx2, Kingston 2 TBx2, Micron 8 TB, WD AN 1500
Display(s) GIGABYTE FV43U
Case Corsair 7000D Airflow
Audio Device(s) Corsair Void Pro, Logitch Z523 5.1
Power Supply Deepcool 1000M
Mouse Logitech g7 gaming mouse
Keyboard Logitech G510
Software Windows 11 Pro 64 Steam. GOG, Uplay, Origin
Benchmark Scores Firestrike: 46183 Time Spy: 25121
Windows is exactly why Ryzen CPUs work better in Linux.
 
Joined
Oct 1, 2006
Messages
4,931 (0.74/day)
Location
Hong Kong
Processor Core i7-12700k
Motherboard Z690 Aero G D4
Cooling Custom loop water, 3x 420 Rad
Video Card(s) RX 7900 XTX Phantom Gaming
Storage Plextor M10P 2TB
Display(s) InnoCN 27M2V
Case Thermaltake Level 20 XT
Audio Device(s) Soundblaster AE-5 Plus
Power Supply FSP Aurum PT 1200W
Software Windows 11 Pro 64-bit
Make sense. The Windows scheduler is either too stupid and/or too focused on the Intel design philosophy to optimally load and manage threads within AMD's CCX architecture, so using a bunch of metrics such as, individual core performance, CCX performance, cache location, etc, they have the firmware/drivers fool Windows in an attempt to have the scheduler use the most suitable core.

The stars and dots are a ranking for the physical quality and performance of a core, whereas the CPPC2 'preferred cores' is a ranking based on the aforementioned metrics and are used to 'trick' the scheduler.
It is more than this.
The Windows Scheduler is worse even for Intel CPUs.
Often the difference is quite significant that not even the extra "bloat" in Windows can explain the difference.

Note: the graphs are Time in ms, lower time is faster. All these test are run on the same Dell machine with Ice Lake 1065G7 CPU.
https://www.phoronix.com/scan.php?page=article&item=icelake-clear-windows&num=2
1574430023923.png

1574430054605.png
 
Joined
Dec 10, 2014
Messages
1,327 (0.37/day)
Location
Nowy Warsaw
System Name SYBARIS
Processor AMD Ryzen 5 3600
Motherboard MSI Arsenal Gaming B450 Tomahawk
Cooling Cryorig H7 Quad Lumi
Memory Team T-Force Delta RGB 2x8GB 3200CL16
Video Card(s) Colorful GeForce RTX 2060 6GV2
Storage Crucial MX500 500GB | WD Black WD1003FZEX 1TB | Seagate ST1000LM024 1TB | WD My Passport Slim 1TB
Display(s) AOC 24G2 24" 144hz IPS
Case Montech Air ARGB
Audio Device(s) Massdrop + Sennheiser PC37X | QKZ x HBB
Power Supply Corsair CX650-F
Mouse Razer Viper Mini | Cooler Master MM711 | Logitech G102 | Logitech G402
Keyboard Drop + The Lord of the Rings Dwarvish
Software Windows 10 Education 22H2 x64
So basically: Best Core is absolute best core as validated by AMD from the factory.
Preferred Core is relative best core assigned by Windows Scheduler.

AMD gonna update ryzen master in the future to match Windows Scheduler.

Is that everything.
 
Joined
Feb 3, 2017
Messages
3,747 (1.32/day)
Processor Ryzen 7800X3D
Motherboard ROG STRIX B650E-F GAMING WIFI
Memory 2x16GB G.Skill Flare X5 DDR5-6000 CL36 (F5-6000J3636F16GX2-FX5)
Video Card(s) INNO3D GeForce RTX™ 4070 Ti SUPER TWIN X2
Storage 2TB Samsung 980 PRO, 4TB WD Black SN850X
Display(s) 42" LG C2 OLED, 27" ASUS PG279Q
Case Thermaltake Core P5
Power Supply Fractal Design Ion+ Platinum 760W
Mouse Corsair Dark Core RGB Pro SE
Keyboard Corsair K100 RGB
VR HMD HTC Vive Cosmos
It is more than this.
The Windows Scheduler is worse even for Intel CPUs.
Often the difference is quite significant that not even the extra "bloat" in Windows can explain the difference.

Note: the graphs are Time in ms, lower time is faster. All these test are run on the same Dell machine with Ice Lake 1065G7 CPU.
https://www.phoronix.com/scan.php?page=article&item=icelake-clear-windows&num=2
The results in that article are more nuanced than that.
Windows wins 50% of these tests, Clear Linux 37% and Ubuntu 13%.
Windows 10 geometric mean is lowest by 4/8.5% but without redoing the calculations it would seem this is largely due to couple very bad results.
 

bug

Joined
May 22, 2015
Messages
13,755 (3.96/day)
Processor Intel i5-12600k
Motherboard Asus H670 TUF
Cooling Arctic Freezer 34
Memory 2x16GB DDR4 3600 G.Skill Ripjaws V
Video Card(s) EVGA GTX 1060 SC
Storage 500GB Samsung 970 EVO, 500GB Samsung 850 EVO, 1TB Crucial MX300 and 2TB Crucial MX500
Display(s) Dell U3219Q + HP ZR24w
Case Raijintek Thetis
Audio Device(s) Audioquest Dragonfly Red :D
Power Supply Seasonic 620W M12
Mouse Logitech G502 Proteus Core
Keyboard G.Skill KM780R
Software Arch Linux + Win10
Make sense. The Windows scheduler is either too stupid and/or too focused on the Intel design philosophy to optimally load and manage threads within AMD's CCX architecture, so using a bunch of metrics such as, individual core performance, CCX performance, cache location, etc, they have the firmware/drivers fool Windows in an attempt to have the scheduler use the most suitable core.

The stars and dots are a ranking for the physical quality and performance of a core, whereas the CPPC2 'preferred cores' is a ranking based on the aforementioned metrics and are used to 'trick' the scheduler.
So you're saying Microsoft should have had a CCX aware scheduler, before AMD launched the CCX?
All cores were created equal until Ryzen, the scheduler was good enough for that. Now that even AMD can't point out the "best" core accurately, it's still Microsoft's fault for not being able to pick it from the line up?

To be clear: all the scheduler should do is ask the driver which is the best core and put it to good use. The driver is supposed to come from AMD first and foremost.

Also, I'm not saying Windows's scheduler is perfect. No scheduler ever is (Linux itself support like a dozen scheduler, none of which is better than the rest is every situation).
 
Top