Well since the last thread about this made some movment here, this time I've created a poll for voting in what do you think about Maxxmem2 giving points.

 

Here are the options:

 

1 - Yes, Maxxmem2 should give points and be ranked for the submition points, as it is;

 

2 - Yes, Maxxmem2 should give points, but it should be based on Read\Write\Copy, excluding latency;

 

3 - Yes, Maxxmem2 shuold give points, but it can only be ranked by Read result directly, excluding the others (Copy\Write\Latency);

 

4 - No, I don't want Maxxmem2 to give points.

 

Come on, let's vote for this, and remember that HWBot doesn't have realy a benchmark so directed to memory performance, as this one. On this benchmark you can show your memory tweaking capabilities, without always beeing overwhelmed by ultra CPU's

 

Thank you for participating!

 

Best regards

 

Morphling

Edited September 23, 201113 yr by Morphling

Come on people! Come here and vote

214 Million views and only a handful of votes,

 

I'll try this out first.

i like alot maxxmem but it's bugged as hell and also quite CPU dependant.

better to not award it with points.

i like alot maxxmem but it's bugged as hell and also quite CPU dependant.

better to not award it with points.

 

Hi there If you think it's so bugged you have other 2 options those don't have bugs, and many other benchs have bugs too

 

If you think that is so CPU dependent give me your best score with maximum CPU frequncy and I'll beat it with 100mhz less on CPU

 

Come on, let HWbot have a memory performance benchmark

 

Thank you for participating!

 

Regards

i'm not against memory benchmarks, i'm against bugged benchmarks like this one is

 

and BTW, when you said "i can beat your score with CPU 100 MHz lower than you" you automatically assumed that CPU frequency have some kind of impact on the final score.

i'm not against memory benchmarks, i'm against bugged benchmarks like this one is

 

and BTW, when you said "i can beat your score with CPU 100 MHz lower than you" you automatically assumed that CPU frequency have some kind of impact on the final score.

 

You have a point there

 

But still, it's the best we can find for memory performance

 

Come on, with the other two options there are no bugs

Even though I love running 2.5V+ through DDR3, I don't think that MaxxMem in its current form should be worth points.

IMO, for benchmark to be able to recieve points it must scale logically and be completely bug-free ... which MaxxMem isn't.

Even though I love running 2.5V+ through DDR3, I don't think that MaxxMem in its current form should be worth points.

IMO, for benchmark to be able to recieve points it must scale logically and be completely bug-free ... which MaxxMem isn't.

 

But that's why there are 2 more options that brings Maxxmem without bugs Option 2 and 3.

 

Come on, at least vote people

 

Regards

2 and 3 both seem to lose hard so far.

2 and 3 both seem to lose hard so far.

 

But at least it is indeed memory performance Otherwise what can we found that is a measuring memory performance benchmark?

But at least it is indeed memory performance Otherwise what can we found that is a measuring memory performance benchmark?

 

Better not to add than to add the wrong benchmark Something decent will show up one day, that's always the case.

Well, maybe you're right

 

It's just because Maxxmem is so adict

 

Well, let's see what the voting will tell.

 

Thanks pal

Votation is ending in 2 days!

 

Please, vote on whatever you think

 

Regards

Better not to add than to add the wrong benchmark Something decent will show up one day, that's always the case.

Memory overclocking is getting very popular in the last 2 years. Additionally the move to integrated graphics and their improved performance forces faster memory in general. Guess that helps too and will finally bring us a good benchmark tool.

OK, so this result is just very interesting. From what I can see Yes have won. Now, the problem is that there are 3 diferent "yes's".

 

I'd like to hear what does the moderators think about this.

 

Still the votations where really competitive.

 

So, what do all the comunity think about this?

 

Regards to all!

From my point of view, there are two options - no boints at all, or keep the current system ("as it is"). However, the difference is very small, both are really close to 50%, so no changes should be made. "IRL" you usually need 2/3 in a poll like this to be sure the change is for the better. (imagine if you have a poll that's for/against something, 51% for one week, and 49% next - then one week after the change you have to undo it, and so on).

Find which are bugged and which are not then you WIN!

http://hwbot.org/forum/showthread.php?t=22522

 

Poll seems to be closed. I was going to vote no. Maxxmem is just way to buggy. Would love to see a memory benchmark, but not maxxmem as it stands now.

aida64 ?

aida64 ?

 

Not a bad idea, the problem is that we can't have a direct result like on maxxmem

 

The perfect bench would be Aida64 with the same system of points of Maxxmem

 

But I would never put direct pontuation, like read results because it would not be fair. The biggest deference between PSC's, Elpida BBSE's, Hynyx, etc compared to the only good IC's, whitch of course are Elpida Hyper MGH-E and MNH-E is the latency.

 

But also, using latency as a direct pontuation would also not be fare for the other IC's

 

It needs to be combined.

 

I think we need programmer hehe

 

Regards

So, what will happen now?

 

We need a decision

 

Regards

Frist of all, sorry for digging this thread up...

 

Then, can anyone give a quick and fast explanation of what is the latency bug about? Is a software (MaxxMem2) bug? Or what is it?

 

Cheers