wPrime 1024 WR

[<<HANNIBAL>>]

Nothing strange?

like 8 cored X7460? or 8 socket Xeon server?

[NeoForce]

http://www.unisys.com/products/enterprise__servers/high_d_end__servers/models/index.htm

It is 8 socket platform

________

Anal live

Edited May 13, 2011 by NeoForce

[OgaiB]

http://www.unisys.com/products/enterprise__servers/high_d_end__servers/models/index.htm

It is 8 socket platform

In this link you didn't see 4 server is rack?

here is specification:

http://www.unisys.com/eprise/main/admin/corporate/doc/ES7000_Model_7600R_Spec_Sheet_Apr09.pdf

 

You can see there are 4 cell

CPU scalability • 4, 8, 12 or 16 processors (four processors per cell, one to four cells)

 

Maximum memory • Per cell: 256 GB (8 GB DIMMs)

• Per system: 1 TB (8 GB DIMMs)

 

This isn't nod?

[ncc-1709]

if your a tech who can get away with benching on something like that, then fair game.

you could allways be the owner of a data centre and try doing a cell bench of a full server rack.....

[NeoForce]

Cell benching is not forbidden.

It is up to users.

________

MAGIC FLIGHT LAUNCH BOX REVIEWS

Edited May 13, 2011 by NeoForce

[OgaiB]

Cell benching is not forbidden.

It is up to users.

ok.

 

Blade Server is also not forbiden?

[NeoForce]

Tell me reason - WHY it should be forbidden?

there is no limitation to amount of cores

________

Georgian cooking

Edited May 13, 2011 by NeoForce

[gantevebis_vaci]

i have read here that only true configurations were allowed

i have no problem with andre

he is a great bencher

but if such validation is Ok than i have acces to super computing centres in germany

next record will be like 0.XXX sec wprime32

and sub 10sec 1024

[gantevebis_vaci]

sorry but wasnt it andreyang who submitedd this score yesterday???

[K404]

Please can we split off multi-CPU results into their own leaderboard? I've asked before and I still think its a valid idea

[K404]

Multi-CPU CPUs have different names, I think it can be done. Obviously, similar to multi-GPU Vs single 2,3,4 are lumped in together it wouldnt be possible to try and class a 4-way system in a 2-way leaderboard for example.

 

CPU-Z will deal with 90%+ of naming screwups.

 

Same rules as always- if it looks dodgy, someone will report it and the score can be analysed.

Edited June 3, 2009 by K404

[K404]

Make the submission process as hard as ya like to make sure results are solid

 

Theres a massive community here who are largely happy to help out...if ya need more mods just post on the forum

[gantevebis_vaci]

please close this thread

thanks jmke

[gantevebis_vaci]

i rather break record than discuss it with you

our leader was banned for that

nothanks

[utilizatori]

that doesn't answer my question as to why this thread should be closed...

here is the reason why the thread was opened

Nothing strange?

like 8 cored X7460? or 8 socket Xeon server?

 

as we get the answers

Cell benching is not forbidden.

It is up to users.

Tell me reason - WHY it should be forbidden?

there is no limitation to amount of cores

 

there is no reason for this topic any more

[gantevebis_vaci]

you have right

OK

[gantevebis_vaci]

thanks jmke

i preciate

[r1ch]

it has to be solid; not just analyzed; current workload for moderation is high enough;)

 

Hi jmke, what are the problems with splitting into single socket and multi socket categories like the 3D benchmarks?

 

I think this would be a good idea. If we know what the issues are, then hopefully we can find a way of solving them.

[knopflerbruce]

Hi jmke, what are the problems with splitting into single socket and multi socket categories like the 3D benchmarks?

 

I think this would be a good idea. If we know what the issues are, then hopefully we can find a way of solving them.

 

Singlecores VS dual + quad + hexa etc? Sounds like a bad idea, as there are very few singlecore models left for sale these days:p Perhaps single + dual + tricores vs the rest would be a good solution TODAY, but not for more than a year or two max. WHich means that by adding that category the rules will have to change with time, which is not right IMO.

[der8auer]

I think r1ch doesn't mean The cores of one CPU. Hes talking about multiple CPUs. Like Single, Dual and QuadCores in one categorie and another for stuff like servers or the Skulltrail plattform with several CPUs.

[knopflerbruce]

I think r1ch doesn't mean The cores of one CPU. Hes talking about multiple CPUs. Like Single, Dual and QuadCores in one categorie and another for stuff like servers or the Skulltrail plattform with several CPUs.

 

That makes more sense:D And I agree. It's a bit like PCMark05 without HDD limits now, it all comes down to who can buy the most expensive CPUs and ditto mobo;)

[Hyperhorn]

IMO Wprime belongs to the server guys as all the other benchmarks belong to the overclockers with access to extreme cooling. New CPUs will arrive (e.g. Gulftown) and the extreme overclockers will decrease their times, too. It`s not that hwbot gets flooded by server farms. If you look at the Top 20 you will see many Skulltrail system, actually more than professional server racks.

[Massman]

Isn't this more interesting? Now, you have to compete at lower rankings, which means that the battle for those last points is more tight.

[K404]

I believe that multiple physical CPUs should be split off into their own categories, the same as SLI and CF are split now.

 

Just because enthusiast-level CPUs will improve over time and "our" times will come down- I do not think thats a valid arguement. Smae thing has happened in GPU categories but they are still split.

 

Can servers be classed as readily available? Its not just a question of money, but of space and power too (from what I know of servers)

[der8auer]

Haha nice thought, Massman

 

I don't see a problem in the current wPrime rankings. Like Hyperhorn said

 

It`s not that hwbot gets flooded by server farms. If you look at the Top 20 you will see many Skulltrail system, actually more than professional server racks.

 

100% agreed