PCMark05 , keep or remove from hwbot.

[chispy]

, chispy I take long time without the help of my daughter to write in English and is well understood, sorry my delay, please read my previous post

 

Camstudio is not a tweak, is software to record video

 

EDIT: Uff be late again hahahahaha

 

Hahha , entiendo ahora , no worries i understand now. Same here my first language its spanish and i have to edit my post many times to write in english and get my point across

[chispy]

allow everything, or remove it.

 

However, this benchmark is a joke. Should be rather obvious when you have to cap a subtest, so it's no longer about getting the highest Xp score, it's about getting as close as possible to a fixed number - after crippling the performance of your raid array... When a system is defined as too fast for a benchmark it's a sign that it's outdated and should be removed.

 

+100% agree with you here knut

[Sweet]

Ohhh welll I'm happy , someone understand spanish, me alegra y mucho chispy, sometimes is hard write in english, cos i'm write in spanish here, and I dont realize

[Antinomy]

Well, first I agree to Bruce - the HWBot rules are too lousy. Apart from that they aren't always applied.

And another example is PCM.

Why is updating the browser allowed while updating codec (not replacing with another, just updating with a more optimized version that does the same thing) not?

Why is updating to Vista/7 and thus using GPU rendering for floating windows, "Any software or human interaction causing the benchmark to use different calculation algorithms as intended by the original software creator. E.g. replacing the codec or browser in PCMark2005." e.g. using Aero legal while updating a codec isn't? Another thing is that besides the cheaty settings like resolution, Powertoy allows setting up priority and system tweaks (multicore or am I wrong here?) - how does that alter the algorythm?

If we allowed to change LOD, are we allowed to set resolution of, say, the floating windows?

 

Rules is a thing that must be done right. And it definitely isn't now. If it's possible to make a normal statement so that everyone understands what's in the rules and what does break them - let's do it. If it's impossible - let's dump it. I agree to Bruce in this case.

[chispy]

If you guys take for example the top 20 scores , the sub-test on every score in the top 20 are all over the place , the disparity and abnormalies going on with those sub-scores are abysmal , massive difference , in everything from low cpu clocks with very high video encoding and web browsing test to HD test Virus Scan over 2000+ with only one or two I-Rams. I have seen it all in the top 20 scores , good ones , bad ones , some legit and some obviously cheated but there is no way to go trhu every single score and every single test and sub-test to make sure its 100% legit , some people lack the hardware and thru the means of the new so called software ***TWEAKS*** make up for it on some sub-test wich gives them the edge at the total score. But when you combine both , the right hardware and the so call new ***TWEAKS*** massive scores can be achieved no doubt , that it is the clue that the benchmark its no longer reliable and cannot be moderated correctly , we need to find a new reliable benchmark wich its more apropiate and suitable to replace this old dying , outdated , sabotaged , obsolete , raped PCMark05. I sincerely believe this benchmark has to be remove.

Edited July 25, 2011 by chispy

[chispy]

clear example of this whole mess wich is PCMark05 its the many many treads with open disscusions thru the years about the cheating or legit tweaks on this benchmark.

[S_A_V]

Why is updating the browser allowed while updating codec (not replacing with another, just updating with a more optimized version that does the same thing) not?

IE9 is default browser for now. 12 may 2011 Microsoft updates their distributives of Windows 7 with integrated SP1 and IE9.

[chispy]

DejaVu lol ...

 

read here , year 2009 , same thing all over again and again...

 

http://hwbot.org/forum/showthread.php?t=2661

 

richba5tard said on that PCMark05 2009 tread Post # 11 """I agree with SF3D. As soon as a benchmark can not produce reliable results anymore due to new technology, it's obsolete... """

Edited July 25, 2011 by chispy

[Gautam]

I don't agree with most of you at all, but it seems like the majority? has spoken. Almost every benchmark out there today is just about having the best CPU and best GPU. You can bin your way through having poor tweaks and bad OS'es. PCMark takes some thinking and a lot of investigation to run properly. I don't even understand why the PowerToy or switching browsers was banned. What I see with every PCMark discussion is, certain people find tweaks, and the people that don't know what they are get mad and brand them as cheats. The only thing that I'm strongly opposed to with regards to PCMark (besides speedhacks and mipmap) is the usage of software ram drives.

 

Why allow the other stuff and not the software ram drives? The reason is simple but for some reason very hard for people to comprehend nevertheless. This I know from PCMark discussions I've seen and participated in in the past. I will try to spell it out nevertheless.

 

A software ram drive is not indicative of the actual performance of the system.

 

Changing the browser, enabling the video codec to use more threads, all that might be manipulating the benchmark, but whether you guys like it or not, it is REAL performance. If Chrome is a faster browser than IE, then it's a faster browser than IE, period. If you enable the usage of more threads, you get more REAL performance.

 

The distinction with the software ram drives is that it's not accurately measuring the performance of the storage subsystem. You get results that are basically fluff and that you can practically arbitrarily manipulate, making it akin to speedhacking.

 

(Ignoring the off chance that you're actually using a ram drive as your actual boot drive and reinstalling the OS every time the system boots...)

 

The other tweaks still result in accurate measurements of the system at hand, just that the system itself has been manipulated in a way that is undesirable to much of the PCMark community. These rules are literally just in place because people got mad about certain things and decided to brand them as "cheats".

 

My actual preference is to remove pretty much all regulations besides prohibiting software ram drives, but since this seems like a topic that the community will never come to a consensus on, removing the benchmark might be the best way to proceed.

[Massman]

All goes, but:

 

- 220 startup limit

- no ramdisks

[Predator]

i have to agree on Gautam here, i'd let it as everything goes except the ramdisk software

 

voted #2

[pro]

okay i think it would be stupid to remove pcmark, it is the only storage based benchmark we give points to

 

lets think about this for a second, biggest bottleneck in PC is storage, yet we only have 1 benchmark that tests storage and we are considering removing it? i dont get it,

 

yes there are lots of interesting things in pcmark but thats good isnt it? do we just want boring benchmark that all you need is the best cpu or gpu.. lets think about all the benchmarks we use.. i can only think of a few examples 3d01, 32m where you can actually catch up without the BEST equipment but you still need near the best..

 

the only real example of something that takes real skill is 4-way and thats too expensive for everyone to play..

 

you can get top 3 pcmark with 4 i-rams or 2 acards, an old 5870 and a sandy bridge...

 

the main issue here as, the benchmark has evolved, the rules are still set at 3 years ago, lets look at the rules, what should be changed or stay, here is my opinion.

 

 

1. keep 220 startup limit, it adds some character, often i get a great score but startup is 240, it makes this bench harder and better for it

 

2. KEEP no MFT and ramdisks... these are not real performance figures, its like mipmapping, it shows a score, but it doesnt actually work

 

3. allow BROWSER CHANGING and optimization, okay lets look at it, if anyone else has tested, they will see its extremely hard to make pcmark use anything except ie, and at the moment IE9 is the fastest browser, and what i have been using since it was released, before that i used IE 5.5 just for clarification... i think if you are clever enough to make another browser faster (and i sure as hell never been able to) then good luck to you, you deserve it.. after all its called web browsing we are still web browsing right?, and were not replacing part of the benchmark, were replacing part of the OS.. how is this different than using a different driver, or operating system, or using nvidia inspector to force d3d compability mode bringing our gtx580's in line with gtx275

 

4. allow all the powertoy encoding tweaks.... this is a sour point for alot of people, is it allow, isnt it, we really need to address this, it certainly should be allowed, then there is no debate..

 

 

i think of course we should keep banned anything that alters the benchmark perception, ie speedhack, etc.. but changing internet explorer options, optimizing your hdd array, changing inbuilt video encoding settings or moving your mouse during a test are not altering the perception of the benchmark.. its just a quirky bench and thats what makes it so cool..

 

i think its sad that we cant have a bench thats tweakable, do we need to bin 30 PCS of gpu to have a benchmark now? thats so lame.. this community has a serious problem with anyone getting an advantage with 'tweaking' as we call it... if the tweak is in the public domain its a tweak, if only one person or a team know it, its a cheat, STRAIGHT UP AND DOWN.. if you have hardware skills and can mod your hardware to give an extra 300 mhz you dont need to explain shit, but if you get an extra 2 FPS in GT3 in 3dmark 05 then your a cheat and should be burned at the steak... real talk

Edited July 25, 2011 by pro

[Lucky_n00b]

While I yet still learning on how to get good scores for this benchmark, I think it's still quite interesting

Voted #2, make new set of up-to-date rules ()b

[Hondacity]

even my choice in the poll can't be tweaked...lmao

[chispy]

All goes, but:

 

- 220 startup limit

- no ramdisks

 

All goes but 220 start up limit and no ramdisks , thats it .

Seriously Pieter you guys at hwbot need to rewrite the whole rules with an updated version to reflect the real changes in hardware and software over time , it needs to be done , James has some good ideas to keep this benchmark going , but can you tell me how acurately the mods will be able to moderate an old benchmark like this one with all this new hardware and software available ? If the answer is , i dont know then the moderation of this benchmark will be useless. If a benchmark can not produce reliable results anymore it should be removed as richba5tard said it before.

 

Is it PCMark05 a hardware benchmark or a Software benchmark ?

 

Is there a posibility to replace PCMark05 with a new benchmark ?

 

Keep the opinions coming guys , lets air this out once and for all.

[T_M]

keep it cos its the only thing pro is good at

[SteveRo]

I strongly agree with Gautam and Pro - both provide good reasoning and sufficient justification to keep pcm05. I also believe we should update the rules IF it will help us moderate the bench.

 

I agree that ram disks should NOTbe allowed -but don't we need better detection than just saying anything over 2000 in virus is ram disk?

[rsannino]

Honest guys are penalyzed by people that don't follow rules and limitations about what is and is not allowed. So, without limitations, everybody will be in the same situation. But, that way, you'll also have a "far west" competition, that is why I voted for "delete pcmark05 from hwbot"

 

Roberto

 

P.S. Sorry for my English

Edited July 25, 2011 by rsannino

[El Gappo]

It doesn't need new rules, it needs the current rules to be enforced. I've reported numerous submissions from the same few people for cheating this bench, the users are still active and still doing it openly. No fear of "being removed instantly".

 

You need to punish the people cheating rather than the people playing fairly or this is going to be a reoccurring problem, not just pcmark.

[pro]

Cheating is a strong word and just reflects the ignorant perspective of a number of members if this community,

 

I have the answer to solve all problems, let's ban all tweaking,

 

No tweaking allowed, just install os, install driver and run bench, let's do it

[Freakezoit]

Maybe we need something like we have on Unigine . And small program wich starts PCM . But that is only one solution. And ban all tweaks wouldn`t help there will be allways an way to tweak any benchmark. So that would be an wrong way to go. But some clear Rules about Subtests would help but there is no perfect solution , only an improvement .

[saint19]

I vote for keep it, but re-lease the rules. I saw that some of the rules are very old and even PCMark Vantage doesn't show the proper example picture in rules.

Edited July 25, 2011 by saint19

[uncle fester]

Yeah seriously guys WTF, what are you guys trying to do ???

next you will say lets ban all reg tweaks for super pi, FFS people software tweaking for efficiency is half of what we do.

Yes wankers cheat but they do it in everybench, now lets get this right EVERY BENCH can be cheated, that is no reason to ban them from the bot, is it just people with the inablilty and time to work out how to bench good that want this to turn to who has the best hardware wins game ???

 

You guys have no idea how long some people work on software tweaking and keeping within the rules, hell with your mentality things like the 580, 01 tweak would not be allowed to be used, FFS people put the time in and learn.

[saint19]

Agree with uncle fester and El Gappo, we are the 1st line to keep the benchmark clean and between rules.

[uncle fester]

This is my favorite benchmark , since 2006 i have been playing it , but as of lately people have started to look for loopholes and abuse those loopholes to get Very high scores , too much tinkering and manipulation of each test with the benchmark and too many programs running in the background (manipulating how pcmark05 scores) does not make any sense at all since its giving you an artificially (false) very high scores wich does not represent a real and legit score ,Now Its not PCMark05 running as it was meant to be run , Now its all about what program can i find to cheat the score lol. I had it with this BS , Massman please remove this benchmark from hwbot or make new revised and stronger Rules for PCMark05 , otherwise i dont see any real use of this benchmark in hwbot since you can run this bench cheating your way up to 40K + just using (cheat) background programs that manage (tinkering) how PCMark05 its running and scores , that alone break the rules and integrity of this benchmark. Sad very Sad the way its going for us long time players of this benchmark. Please guys vote accordangly and thank you in advanced , lets decide once and for all the faith of this benchmark.

Regards: Angelo.

 

wow im sorry chispy, i should have read this before i posted,

But WTF just becasue you cannot figure out this bench out you want it banned??? , oh ffs i cant figure out a good run order for 01 i say lets ban it because dino and vince kick my arse, jesus mate i know you have been benching pcmark for ages and asking james for his secrets, but this is just rediculas.