OK, explanation still doesn`t show different models

 

I think different models...

I agree different models

So it should be stated, no offense.

It`s a pity that it wasn`t clearly written before the start, but I`m sure that it will help during next stages and will be written next year in rules.

First two stages were a bit rough; I don't see any problems coming up with the next 4

 

But I agree - next year, more clear!

and finally these are models: KHX16000D3ULT1K3/6GX and KHX2000C8D3T1K3/6GX =)

and finally these are models: KHX16000D3ULT1K3/6GX and KHX2000C8D3T1K3/6GX =)

 

Ugh, you're right. I hate Kingston for using so similar partnumbers

 

9905403-039.A00LF (DFB)

9905403-067.A00LF (SV1SH)

Thanks for understanding. We will avoid such situations in future.

That is ok, but 31ns latency...

bugged run? Tweaked run ?

that's random, not tweakable

For such clocks ???

The scores with under 33ns latency are bugged... unless they are at 5.7GHz uncore

Gratz for Russian team! Was hard fight! Молодцы!

Edited January 13, 201114 yr by GunGod

The scores with under 33ns latency are bugged... unless they are at 5.7GHz uncore

 

Oh! http://hwbot.org/community/submission/2105354_micutzu_maxxmem_ddr3_sdram_2358_marks

uncore only 5000mgz lat 31.3ns

The scores with under 33ns latency are bugged... unless they are at 5.7GHz uncore

This one ??

http://www.hwbot.org/community/submission/2105354_micutzu_maxxmem_ddr3_sdram_2358_marks

 

Is there any special tweak ?

Gratz for Russian team! Was hard fight!

 

Yes it was difficult. At you very good results!

Well, you guys name it. Bug or tweak. But if it is bug, all <33ns scores must go

For low latency have one tweak, no comments more, need using true benchmarks)))

That I agree, MaxxMemm is not quite reliable, and that is a problem, because if we start to check which is bugged and which not, it is definately a pain in the but.

For low latency have one tweak, no comments more, need using true benchmarks)))

 

I'm now playing this card: "Any special software tweaks must be explained to the HWBOT staff during the competition"

I'm now playing this card: "Any special software tweaks must be explained to the HWBOT staff during the competition"

All trick in optimal selection subtimings in CPU-Tweaker, loading windows on BLCK 250+ and lowering of BLCK frequency in windows. It is all.

It is not software tweak, it is right settings in BIOS =) need higher boot fsb frequency!

Tricky!

Guys, I think that it can be fair to remove stage with bugged benchmark or accept as it is.

I`m sure in future competitiions MaxxMemm will be updated or removed at all.

Our time, ln2, money, experience, died hardware and now remove this stage? I am disagree