Achievements - What do you want?

[Massman]

It's not a right measurement since you run one benchmark multiple times for one score. If you see 6.7s in 1M, it means you spend 10 hours getting that score

[Antinomy]

I know. But this is a thing that can't be measured. The total benchmark time - can be

[Massman]

Just tried the query - the problem is that some impossible scores (that are not blocked) mess up the ranking.

[Thor941]

too many achievements kill the achievements IMO

[Massman]

... "Achievement King", "Achievement Guru", "Achievement Junkie"

[Massman]

Oh, and don't add the achievement for the newcoming socket at least half a year - until the first wave of popularity is over.

 

And that's when exactly?

 

Initially, I was thinking about only doing this for discontinued platforms, but it's not a bad idea to also do this for very new platforms (it's maybe easier to be king, but if it's easier more people will try to get it, so it becomes difficult again?)

 

If we revise caps every 6 months, we can add this kind of achievement to every new platform no problem.

[Antinomy]

Well, every platform becomes discontinued sooner or later, I don't think there's a difference when to add it to the achievements.

Maybe someone likes to perform in only new hardware. And the old school guys will try to get the Socket A award most likely. So I see no problem. But add only desktop platforms - the server and mobile ones are not that fun IMO.

[knopflerbruce]

I think all platforms should be added. Server stuff isnt that exciting, but that's not really an excuse to not include those IMO.

[Splave]

I think knopfler deserves the Golden Jesus award look at those medals

[Thor941]

I think knopfler deserves the Golden Jesus award look at those medals

 

[Antinomy]

Another suggestion about the achievements placement.

Please, put the manufacture achievements (i.e. AMD, Intel, NVIDIA and ATI) in one line. Otherwise the page body is a bit too long for a comfortable view.

[Antinomy]

You can start looking for AMD K5 benches for the long Spi32M

But I agree about the link on the results.

[Massman]

writing queries for the socket king/guru/junkie achievements at the moment. I'm using the socket characteristic of the cpu model for this achievement, not the mainboard model. This means that you can get achievements by using an AM3 cpu on an AM2 mainboard.

 

The alternative is to go by mainboard socket characteristic, but this will require people to select a motherboard model. In addition, it doesn't solve the socket combination problem: there are socket convertors out there as well. Imho, the first alternative (go by cpu socket characteristic) is the best alternative

[knopflerbruce]

Just curious, do I have 754, 939, am2 and 940?

[Massman]

939 - yes

754 - yes (close)

AM2 - yes

940 - yes

AM2+ - yes (close)

AM3 - no

[Turrican]

no wonder, since knut is only benching a64s

[knopflerbruce]

939 - yes

754 - yes (close)

AM2 - yes

940 - yes

AM2+ - yes (close)

AM3 - no

 

Uhh, I have am2+ as well? LOL.

 

Ahh, I forgot I still have some work to do on 754 Will do it next month:D

[Antinomy]

Imho, the first alternative (go by cpu socket characteristic) is the best alternative

I completely agree with this one. Sometimes socket converters are the part of the game in pushing the CPU to the edge so it doesn't have anything about the CPU itself.

[knopflerbruce]

CAn't both be implemented? I mean... you go by motherboard first, and if there's no mobo info, you go by the CPU?

[Antinomy]

FWIW?

The motherboard is inprecise - so why don't just use the more accurate one instead of implementing two methods?

[Massman]

CAn't both be implemented? I mean... you go by motherboard first, and if there's no mobo info, you go by the CPU?

 

It's an incredibly complicated workaround since mb socket can differ from cpu socket.

[knopflerbruce]

Maybe you can add some box for the sockets this applies to - like "socket xxx to yyy adapter used".

[Antinomy]

Still too complicated. And once again - FWIW?

 

If I don't fill up the MoBo info cause I'm lazy, I won't be able to get the achievement? Also, this splits competition into too many classes.

As Massman said much more than once - "don't implement a rule that is too hard to trace out".

The CPU socket is easy to tell - it's CPU-Z.

 

Let's take the socket 5 and socket 7. Can you tell whether a certain result was achieved with a socket adapter or not? CPU-Z is too buggy when working with old hardware. It can't even detect the multiplier though third-party software is able in the same case.

[Massman]

Maybe you can add some box for the sockets this applies to - like "socket xxx to yyy adapter used".

 

Technically, yes ... but for what purpose?

 

Solution A: use CPU characterisic + risk for socket convertors (minor issue)

Solution B: use MB characteristic + risk for socket convertors + necessity to add MB info (bigger issue)

 

Solution C: use MB characteristic + CPU characteristic + socket convertor checkbox + risk for socket convertors + write code in case MB isn't equal to CPU (huge workload)

 

So, CPU socket it is

[richba5tard]

hey Thor, notice anything special about my sig?

 

 

Signature engine v3 is in development, it shows your best or last achievement in the sig.