PCMark05 , keep or remove from hwbot.

[[GF]Duke]

Mr.Scott and i are on the same channel on every aspect of this argument. I spend a ton of money on hardware dude. Including storage. And thats what pisses me off most. With all the tweaks going on, i feel like i've blown a ton of cash for nothing. Again. Because of rule changes here. This is not the only situation. It's prevelant thoughout HWBOT. I feel like a broken record here. You're right i should just give up on what i believe because i am out numbered. Then again that pretty much means giving up competing here. And thats probably not a bad idea either. I could buy a new Harley every year ffs.

[ZMBKLN]

Duke;158208']Mr.Scott and i are on the same channel on every aspect of this argument. I spend a ton of money on hardware dude. Including storage. And thats what pisses me off most. With all the tweaks going on' date=' i feel like i've blown a ton of cash for nothing. Again. Because of rule changes here. This is not the only situation. It's prevelant thoughout HWBOT. I feel like a broken record here. You're right i should just give up on what i believe because i am out numbered. Then again that pretty much means giving up competing here. And thats probably not a bad idea either. I could buy a new Harley every year ffs.[/quote']

 

Now hey... Something we can both agree on! This crap is retarded hard on the pocket book if you want to be competitive.

 

Again... Just to clarify... I mean no ill will to anybody. I apologize if it came off that way.

 

Now... Having invested a large amount of cash in storage myself (JUST for this bench) I see where you're coming from. However for me... I take solice in the fact that I'll eventually (once all my other ducks are in a row) be able to put up scores that the "tweaks only" guys can't compete with.

 

If a person has truly spent the cash to get the right gear... and knows how to properly utilize it... There is no reason they shouldn't be at the top of the rankings.

 

To allow software RAMDisk on the other hand... Would change this... And in my opinion would truly break the bench.

 

I myself... Don't particularly agree with WBC on onboard RAID systems being legal.

Edited February 29, 2012 by Convicted1

[Mr.Scott]

Did you mean to have the same MHz for both examples?

Yes, but they are just examples. My point was that a single core at the same speed of a dual core, with the same periferal extra hardware, using tweaks/hacks, should never be able to beat said dual core that's using no tweaks. Unless I'm mistaken and PCM05 is a totally single threaded app.

A tweaked single core should never beat a dual at the same speed ever. Yet I can find examples of it in the rankings. That's why this is no longer a hardware bench.

Edited February 29, 2012 by Mr.Scott

[ZMBKLN]

Yes, but they are just examples. My point was that a single core at the same speed of a dual core, with the same periferal extra hardware, using tweaks/hacks, should never be able to beat said dual core that's using no tweaks. Unless I'm mistaken and PCM05 is a totally single threaded app.

A tweaked single core should never beat a dual at the same speed ever. Yet I can find examples of it in the rankings. That's why this is no longer a hardware bench.....at least not a CPU bench, as storage is much more a factor. Since there is no category for storage benchmarks, it should no longer be considered a hardware bench at all.

 

I can see your point here... But at the same time... Look at it like this...

 

The Dual Core... with the same tweaks would then be able to stomp all over the Single Core.

 

That is why my stance is that this is still very much a hardware bench. The hardware STILL matters. The tweaks only affect the efficiency of the given hardware.

 

Something I myself would love to see though... Is a "total system performance" category added... I think thats much more fitting of what PCM05 actually is.

 

BUT... I think PCM05 is the only one that truly fits this category. I haven't run PCMVantage or any of the other PCM's enough to truly make that assumption though. I leave that to you more experienced guys to debate.

[K404]

I see you changed the FX-55 to the Opty 165

 

Well......We see "weird" leaderboards during the transition, but once the new rules affect all the leaderboards, things will look "right" again. ie.... single core loses to dusl-core, loses to dual-core w/HT etc.

 

There are plenty of GPU categories where higher spec doesn't have better scores. E.G.... GTX260-216 Vs GTX280.

[Mr.Scott]

I can see your point here... But at the same time... Look at it like this...

 

The Dual Core... with the same tweaks would then be able to stomp all over the Single Core.

 

That is why my stance is that this is still very much a hardware bench. The hardware STILL matters. The tweaks only affect the efficiency of the given hardware.

 

See....it's a vicious circle pending your point of view. Being a hardware bench, I see the hardware first and foremost.

You see the tweaks being more important than the actual hardware.

[Mr.Scott]

I see you changed the FX-55 to the Opty 165

 

Well......We see "weird" leaderboards during the transition, but once the new rules affect all the leaderboards, things will look "right" again. ie.... single core loses to dusl-core, loses to dual-core w/HT etc.

 

There are plenty of GPU categories where higher spec doesn't have better scores. E.G.... GTX260-216 Vs GTX280.

 

I wanted a more drastic example. You caught the post as I was editing.

I understand the board will level out eventually, but that doesn't make it right.

Besides, almost nobody will bench using the tweaks on old hardware. Somebody could make a substantial killing in hardware points on the old platforms should they decide to apply themselves to it, and those rankings may never level out.

If it's now just about global points, then omit hardware points for PCM05.

Edited February 29, 2012 by Mr.Scott

[ZMBKLN]

See....it's a vicious circle pending your point of view. Being a hardware bench, I see the hardware first and foremost.

You see the tweaks being more important than the actual hardware.

 

LOL... Chicken or the Egg I suppose.

 

I just feel that the tweaks don't mean the bench is "broken"... Because it does still reward faster hardware with faster scores.

 

I do see your point of view now though... That lesser hardware can be made to put up scores equivalent or better than faster hardware that isn't tweaked.

 

I suppose thats a disservice to what we're trying to accomplish here....

 

But... I look at it as being the same as any other benchmark. They are all tweakable.

 

LOL... Guess there is no perfect solution.

 

We both have valid points. (I hope you'd agree...) Problem is... Neither outweighs the other.

Edited February 29, 2012 by Convicted1

[Mr.Scott]

I'm more pleased the we can have a civil discussion, than anything about PCM05. Thank you for that.

[ZMBKLN]

Yeah... I was being a bit of douche at first.

 

For that I do apologize.

 

Just seemed like some guys were attacking the bench for (as I understood it) no real reason.

 

Its one of my favorite benches (oddly enough... Because of the tweakability)... So I'm a bit biased I think. LOL

 

To me... It's more fun than watching SuperPi loops repeatedly for hours. LOL

[K404]

Besides, almost nobody will bench using the tweaks on old hardware. Somebody could make a substantial killing in hardware points on the old platforms should they decide to apply themselves to it, and those rankings may never level out.

 

Well.... the more popular a category is, the more popular it will become. If a category only has a few points, only those who...."care" about that piece of hardware or put emphasis on silverware will bench it. Applies to every category.

[Mr.Scott]

Well.... the more popular a category is, the more popular it will become. If a category only has a few points, only those who...."care" about that piece of hardware or put emphasis on silverware will bench it. Applies to every category.

Yes, I guess that's true.

[I.nfraR.ed]

Haven't seen anyone raised that question.

I wonder why there are global points for 5x, 6x, 8x, 12x and so on number of cores?

Hardware points - ok, but as I've read somewhere PCMark05 utilizes up to 4 cores. It's not like wprime. Correct me if I'm wrong here.

So, a guy with "1 of 1" score in unpopular category still gets full global points. He can even run at stock, the outcome would be the same.

On the other hand if you manage to score 4th in a tough category like wprime1024 with a single core processor, you get less points.

How fair is that?

Well, if there are not global points for these kind of systems, they would be not appealing for submissions, that's also true.

 

Not going against anyone personally. If there are flaws, they will be always exploited. Just sharing my thoughts.

I'm not even a points collector, I just do it mostly for fun.

 

In my opinion there should be global points for PCMark05 as a whole, but not based on core count.

Edited March 1, 2012 by I.nfraR.ed

[thebanik]

Haven't seen anyone raised that question.

I wonder why there are global points for 5x, 6x, 8x, 12x and so on number of cores?

Hardware points - ok, but as I've read somewhere PCMark05 utilizes up to 4 cores. It's not like wprime. Correct me if I'm wrong here.

So, a guy with "1 of 1" score in unpopular category still gets full global points. He can even run at stock, the outcome would be the same.

On the other hand if you manage to score 4th in a tough category like wprime1024 with a single core processor, you get less points.

How fair is that?

Well, if there are not global points for these kind of systems, they would be not appealing for submissions, that's also true.

 

Not going against anyone personally. If there are flaws, they will be always exploited. Just sharing my thoughts.

I'm not even a points collector, I just do it mostly for fun.

 

In my opinion there should be global points for PCMark05 as a whole, but not based on core count.

 

I agree,

[I.M.O.G.]

I think massman commented before, and the high core count point rewards in pcmark05 were kind of an unexpected side effect of the rev4 point scheme. It was sort of highlighted when 1st place in the enthusiast league was taken rather easily from just running high core count benches at stock - I don't remember where exactly he posted about it, but it was a conversation related to that. Kinda cool to see what those sorts of rigs are capable of, but the points rewarded don't reflect the effort involved.

 

There needs to be some sort of split rankings for pcmark however, otherwise no one would want to bother running single core, dual core, or other configs for top rankings... Those categories are pretty intensely competitive currently. The top 50 rankings for pcmark would be almost exclusively sandybridge chips if there was no split, making running anything else for the bench almost pointless.

[knopflerbruce]

There should be a maximum number of cores for certain benchmarks, yes (like for PCMark04 we only need single core and multicore, as it only benefits from two threads). PCMark05 up to 4. Vantage and 7 I don't really know. I bet the number of threads is specified in some white papers.

 

Same thing goes for 01 and AM3, no? neither of those take advantage over more than 2 GPUs AFAIK.

[K404]

This is a mostly noob comment, but for the Video encoding test, once it is made multi-threaded, how many threads can it use?

[S_A_V]

There should be a maximum number of cores for certain benchmarks, yes (like for PCMark04 we only need single core and multicore, as it only benefits from two threads). PCMark05 up to 4. Vantage and 7 I don't really know. I bet the number of threads is specified in some white papers.

Both PCMark04 and PCMark05 uses some OS components that scales from number of cores - web browser, audio-video codecs, etc.

 

Don't look at "whitepapers", just open process manager, run benchmark and see real processor usage per cores.

 

The same with 3DMarks03/05: they don't need multicore CPU, but video drivers support multicore CPU, so we have better scores with more cores.

 

PCMark04 - Web Page Rendering - Phenom II 560:

2-core - 60.6 - 6346mhz

3-core - 64.8 - 6300mhz

4-core - 70.9 - 6346mhz

 

PCMark04 WMV Video Compression subtest also scales from number of cores on Windows 8 with right settings.

[I.M.O.G.]

SAV: That is good proof of scaling. Do you have any proof of scaling beyond 4 cores? I haven't seen any scaling beyond 4 cores in any subtest.

 

This is a mostly noob comment, but for the Video encoding test, once it is made multi-threaded, how many threads can it use?

 

I haven't seen scaling past 4 threads testing on FX-8120 in this subtest. I haven't tested results much with other subtests and more than 4 cores.

Edited March 1, 2012 by I.M.O.G.

[S_A_V]

SAV: That is good proof of scaling. Do you have any proof of scaling beyond 4 cores? I haven't seen any scaling beyond 4 cores in any subtest.

 

I don't have direct comparison with more than 4 cores with exact the same settings.

Just found PCMark04 results with FX-4100 @ 6918 MHz and FX-8120 @ 6782 MHz:

4-core: 49.3 web page rendering, 317 fps WMV - http://hwbot.org/submission/2231053_

8-core: 55.2 web page rendering, 362 fps WMV - http://hwbot.org/submission/2240067_

Don't look at Overall score, FX-4100 result was done with wrong Grammar Check.

 

Windows was updated with AMD Bulldozer Performance Updates:

Windows6.1-KB2592546-x64.msu

Windows6.1-KB2645594-x64.msu

Windows6.1-KB2646060-v3-x64.msu

 

I haven't seen scaling past 4 threads testing on FX-8120 in this subtest. I haven't tested results much with other subtests and more than 4 cores.

Do you use WMV settings suggested in PCMark05 Tweaking Thread?

I mean this: "Force NumThreads"=dword:00000002

Try playing with other values

[knopflerbruce]

can you replace the grammar check?

[Mr.Scott]

can you replace the grammar check?

A better question would be, if you are, is it legal?

[S_A_V]

can you replace the grammar check?

I don't have FX-4100 anymore.

May be I'll do some testings with FX-8120 in 2-4-6-8 core mode, when I have time for that.

[S_A_V]

A better question would be, if you are, is it legal?

Grammar check "fix" was used by I.nfraR.ed for HWBOT OC Challenge June 2011, so I guess it is legal.

The only one thing that can be considered as illegal is registering the benchmark with a key.

But this benchmark not available for sale long time ago and we just not have any other options to get that "fix" working.

[Mr.Scott]

Grammar check "fix" was used by I.nfraR.ed for HWBOT OC Challenge June 2011, so I guess it is legal.

The only one thing that can be considered as illegal is registering the benchmark with a key.

But this benchmark not available for sale long time ago and we just not have any other options to get that "fix" working.

OK, I know the 'fix', but 'replace' was the word used so it threw me off.

 

Next question, why would registering the benchmark be illegal? I mean, if you have a key why not use it?

Edited March 2, 2012 by Mr.Scott