Is it really that deep?

Mem freq subs will be allowed with empty slots on the motherboard ? I guess, yes. It's somewhere the same. Max freq is about the highest, reachable freq with the current architecture. If You won't allow disabled cores or even turned off ht, than what about spi for example or some legacy 3D ? If it will be forbidden in the one and only pure freq based benchmark, it will produce a few questions right after.

18 minutes ago, Leeghoofd said:

Correct Mat, I compare CPUZ max Freq, by measuring the Horsepower of an engine on a testbench. On a track it can pop in under a lap.

But plz finish the latest BenchMate before indulging in another project

Of course. It would need to be done using BenchMate anyway because as soon as we are talking validation, security and timers, you will always end up with the same conclusion: it all sucks and needs a whole concept to make sure that it is foolproof, hard to cheat and easy to use.

New BM version will also let you close the client at any point. So only a background service is running to measure sensors and the driver enables correct time measurement and that's it. That said the saving of validation files should be done differently as well. Also without the client and with a single hotkey.

So new BM versions first, new suicide runs later.

Something we should think about is the definition of "maximum frequency".

Right now with CPU-Z the definition is that the CPU's sensors can be read and the CPU can produce an encrypted result file (which implies disk I/O, multiple levels of drivers invoked, CPU-Z version-dependent code, ...). It's also not a very stable definition as it can change (and has changed) between different versions of CPU-Z. The right term instead of "max frequency" would actually be "CPU-Z validation stable".

So before it makes sense to code anything, we should try to define what "max frequency" really should mean for us and also for the outside world. And how this value is communicated to be credible.

How it's technically done is for now irrelevant. The requirements will define that.

Edited April 5, 20214 yr by _mat_

Can liquid helium be banned?

2 minutes ago, Takukou said:

Can liquid helium be banned?

why

8 minutes ago, keeph8n said:

为什么

Why? If it is disgraceful to ban the core, then it is worth advocating to use lower temperature to defeat the opponent? I think we should focus on skills, not more money!

1 minute ago, Takukou said:

Why? If it is disgraceful to ban the core, then it is worth advocating to use lower temperature to defeat the opponent? I think we should focus on skills, not more money!

99% of the time, the vendors are the ones supplying it. MSI, Gigabyte, and ASUS have all used LHe to chase records.

 

If someone is willing to spend the money, who are we to stop them?

Not sure if there's any skill involved in CPU Freq, the magic word is Binning silicon/cores... As with nearly all silicon at the top of the charts, but for other benchmarks more tweaks etc come into play....

LHe is only accessible anyway for manufacturers like stated above. Plus its no warranty if a record was done on LHe that it will remain on top of the charts.  When using LHe the focus is far less on efficiency and more on raw clocks.

16 minutes ago, keeph8n said:

99％的时间是由供应商提供。MSI，技嘉和华硕都使用LHe来追逐记录。

 

如果有人愿意花这笔钱，我们将阻止他们吗？

If someone is willing to close the core to increase the frequency, why not? He didn't cheat

If we are about to change things then the only thing I would really want to see to be changed is a separate top10 list for sponsored uploads. This way anyone could get to be 1st, not just people who were pre filled in the pull down menu from vendor clockers lists.
No hard feelings, just the way I feel about it and my main reason to avoid new hardware for overclocking. 

Btw if we won't accept different clocks within the cores it will mean You can't valid your cpu with factory specs. I know it's hwbot and it's not about base specs but this is sounds pretty crazy to me. In the future, clock management will be pushed even more into this way. All-core setups, with everything fixed is the past. You have right about cpuz validation and the pretty small group who committed to this. It probably, won't be any "big" thing, if rules going to be changed, but I have problems only on theoretical levels. I don't think it would lead to the future.

Soon Intel will release a big-little cpu. How about that ?

Edited April 5, 20214 yr by Alpi

2 hours ago, Leeghoofd said:

Correct Mat, I compare CPUZ max Freq, by measuring the Horsepower of an engine on a testbench. On a track it can pop in under a lap.

But plz finish the latest BenchMate before indulging in another project

How about an added column in the CPUZ ranking? Is that a feasable solution for all parties ?

 

We still have the issue of affinity incorrectly locking the frequency on all cores to that of the core the affinity is set too, which is a cpu-z issue obviously. So not sure you can fix that without forcing integration of benchmate or something similar.

Also, all this drama for a bench anyone who has any interest in efficiency or tweaking will ignore?? Seriously?

Edited April 5, 20214 yr by Noxinite

Is it feasible to have 2 different rankings for cpu-z max validation ? One all out single core balls to the walls quick validation Liquid helium / Ln2 allowed and another one for all cores active  ht/smt enable + some stability test added on whatever cooling allows ?

 In my humble opinion that should keep everyone happy.

All this doesn't solve the question "But can it run Crysis at that frequency?", so why bother?

I'm all for leaving it as is and don't make things more complicated.

CPU-Z doesn't even show what each core is running at, it has to be modified to suit that need and I agree with others that mentioned big.LITTLE and all other complications.

PS: I stand corrected, it apparently does now. It shows when I've last "benched" cpuz 

 

Quote

3. Do you think moving only ocing only one of the cores should be allowed for CPUZ submissions? IE. Ryzen 5950x all cores at 1600mhz one at 6400mhz

Technically that's not possible  Rest of the cores in the same CCD will have much higher frequency. I think the maximum difference is 800MHz or something like that. Effective frequency is another thing though... but that's also applicable to all-core OC.

Edited April 5, 20214 yr by I.nfraR.ed

3 hours ago, Takukou said:

Why? If it is disgraceful to ban the core, then it is worth advocating to use lower temperature to defeat the opponent? I think we should focus on skills, not more money!

Well..... If this is an arguement that's based on whether a bencher is sponsored or one of the "In-House" guys, it doesn't have alot behind it.
There are some here that can get results similar to or even beat these guys while doing it on their own dime like I have to do for anything I do, regardless of how it turns out.

Money, although it DOES make alot of difference isn't the only variable in play here, some do have the skill and determination (Insanity?) to make it happen. While it is true those guys have a real, inherent advantage it's not the end-all as some make it out to be.

Everything I've ever done here has been 100% on my own dime and anyone that knows me can and will vouch for that as fact.
As for the insanity - I can't deny it on my part.  ?

Edited April 5, 20214 yr by Bones

Stick to the topic gents, personal or other beer powered derails will be removed... 

I hope my post is safe.

But so far it seems the community wants things to stay how they are.

Max Freq = Max Freq

 

Just an idea, introduce this All Core Max Freq as a Beta, Lord knows there is some not so great benches already in Beta, and see if people actually try or participate in it.

I'd be glad to see nothing ever change with pre-existing benchmarks and rules. When rules stay the same you can get an apple-to-apples comparison generation after generation. If we're going to change stuff, let's change it all back the way it used to be. No seasons, no sliding scales, points is points, is points, the highest score wins. 

I see and like the thought behind requiring all cores and HT be enabled. That is a better way and ideally it would have been like that from the start. I don't use my computers with cores disabled or HT disabled. I bench my systems the way I use them. I think it is totally appropriate for hardware that can't overclock well with all cores and HT enabled to not have any opportunity to place well on a leaderboard. There shouldn't be a way for it to place well against hardware that does.

But, in spite of that I'd say leave it the way it has always been. Enough things have already been screwed up by change for the sake of change. And, there should be no rewards for the act of participation. Reward only results, not effort or participation.

Edited April 5, 20214 yr by Mr. Fox

I love to see the discussion happening; this is what's great about the community. All are coming together to share insights and opinions with a bit of banter.

Looking at both sides of the equation here, we can see both arguments and I want to touch on these.
The all-out frequency method that's historical to Overclocking and embodies a core pillar of its meaning and the new world realities of hardware, stability and what happens behind the scenes to make it happen.

We have the current method, which is historically accurate to overclocking: Maximum frequency, maximum volts, all the tweaks, none of the dead weight, get that highest score with no regrets. This might entail better cooling, like LN2 or LHe, or better tweaks or OS like XP, or even disabling cores/HT to get that maximum frequency.
This is, in my mind is a pillar of overclocking and one that is best shown in the frequency records on the CPUz validations. This is one style of overclocking and one that I think should still be around for the future.
 
On the other side, maybe with some pressure from vendors and OEMs to show what the hardware is capable of, their desire to have enabled all cores/HT on the hardware when setting records, especially when it comes to new hardware generations. These companies are throwing massive money into Research and Development, and world record scores would surely be a KPI against the iterative improvements they're making each generation. 
I DO NOT BELIEVE this has any influence on this decision or Hwbot directly, as some others maybe have insinuated.

So what does the future hold? 
I don't know, but I do want the all-out frequency method to be around and viable for the long term. If changes are to be made, let's take some suggestions from the world of sports, and maybe look to have classes or categories?
A class for all out, one that's got stabilitiy requirements, and naturally the one true class, Fully stock cooling and XMP memory settings. 

All out category? CPUz maximum frequency validation.
Fully featured? All cores/HT and validated with a stability test of X from Y application?
... others?
Stock cooling and XMP memory only? Validated in real-world settings? Maybe some game benchmarks or something visual for the Youtubers and hardware reviewers. (My personal favorite, as it's no longer pay to play)

More can be said on how to validate and enforce these. I don't have any opinion worth sharing about this now, but I would be open to exploring how it goes.

Thank you for coming to my ted talk.

 

2 hours ago, Leeghoofd said:

Stick to the topic gents, personal or other beer powered derails will be removed... 

what topic? roman sounds like hes going to do whatever the banana he wants to.

On 4/5/2021 at 7:26 AM, der8auer said:

The good thing is that we don't need a poll to decide this. I already talked to Albrecht last week to tell him we will not allow CPU-Z submissions anymore with disabled cores. We are currently thinking of when exactly we are going to enforce this rule. If already starting from this Gen or next one. You can give opinion on this if you want.

But for the future Only full cores + HT will be accepted. Everything else is just nonsense on a technical level.

does that mean old records will be scratched. as otherwise its simply impossible to reach the point levels set anyway.

1 hour ago, Matsglobetrotter said:

does that mean old records will be scratched. as otherwise its simply impossible to reach the point levels set anyway.

Also to make it more clear:

New submissions with old hardware: Old rules! You can still do single cores

New submissions with latest gen (STILL have to decide if we already start with 11th gen): New rule: All Cores + HT

This way pretty much nobody is affected. No old submission, no global ranking. No need to get mad 

 

4 hours ago, zeropluszero said:

what topic? roman sounds like hes going to do whatever the banana he wants to.

Ah so you are giving up so easily? ever thought about providing good arguments or even better maybe alternatives instead of feeding the troll machine or spicing up urban myths about HWBOT.

Just my 2 cents  or maybe 4 as the post is longer. not evaluating what has been and what will be but more looking at the hardware aspects as i also commented on a great video made by Buildzoid in the same line of thought.

I completely agree with the aspect of simply measuring cpuz and get it validated for a single core. Binned or not it shows a great complexity to be able to overclock up to 7 - 9 Ghz range on any CPU. Congratulations to those who managed and hold the records.  I also congratulate the manufacturers of both CPU and Motherboards and memory who made it possible.

Having eons ago worked in the Cable TV industry designing parts of line amplifiers I know how difficult it is to ensure that a 5Mhz analog TV channel is actually ending up on the screen with its sync-pulses across that 5 Mhz channel containing a wide amount of information when you introduce lousy coax cables and filters to have other communication going back towards the main central. Getting the signal through was one thing. Ensuring that the group time delay within the channel and its data bits was not twisted was a real challenge. Leading to a good picture or no picture or anything in between. Not to mention when I had to work on amplifiers for Sat TV at 10Ghz.

Relating that back the discussion of one core or many cores for CPUz. I actually consider both should be getting ranking.  getting a CPU and MB matched to run one core 7-9 Ghz is amazing. But what I find even more amazing is motherboard manufacturers that make sure high speed can happen across all cores on a CPU.  Considering a wavelength of the signal at 5Ghz is 5.99cm and at 7 Ghz its 4.2 cm is a huge difference. If I used an antenna for a transmitter it would be difference between having functional communication or having all the signal coming back to the transmitter due to a mismatched antenna. you can test it out, try use you 2Ghz wifi antenna exchanged for a 5Ghz antenna and see the result. Being able to match all the external components and leads to a CPU so that it still works with all the critical timings for a single core is difficult. To do the same across all cores with all external components is astonishing. No wonder there is need for multi layer PCB's closeness to the CPU of memory and other components to reduce noise and signal delays. I am sure you have seen the nice pattern some of the leads to the memory on your computer has as it surely is not straight lines as that would not work except at max 800Mhz?. Its made on purpose to hopefully ensure fast speed. are all the manufacturers focusing on this the same way.. not necessarily. but with a bit of competition here on HWbot there will be ?

So why not have two CPUZ core rankings. 

1. any amount of cores disabled

2. Fixed by CPU amount of declared physical cores.

I would propose to do  a similar category on Memory to make it more interesting. single channel dual channel and maybe quad channel. Getting  4-5 Ghz on a dual memory setup is definitely more difficult than on a single ram. Good expensive ram or not a lot comes down to if the design of the MB is equally good or not.

2 hours ago, Matsglobetrotter said:

So why not have two CPUZ core rankings. 

1. any amount of cores disabled

2. Fixed by CPU amount of declared physical cores.

I would propose to do  a similar category on Memory to make it more interesting. single channel dual channel and maybe quad channel. Getting  4-5 Ghz on a dual memory setup is definitely more difficult than on a single ram. Good expensive ram or not a lot comes down to if the design of the MB is equally good or not.

Proposal 2 has CPUZ limitations as mentioned in other posts, so that might be something for the future

Regarding the memory proposal we have already tried it out in Country and Team Cup and indeed it is more challenging, but again CPUZ is not 100% bulletproof.