Its the same reason AMD only uses a single K8 against a quadcore K10 instead of a dual K8s. It could end up embarrasing with specfp_rates due to the dual K8s NUMA/extra bandwidth.
You can also learn more and check out other results on www.spec.org
Printable View
Its the same reason AMD only uses a single K8 against a quadcore K10 instead of a dual K8s. It could end up embarrasing with specfp_rates due to the dual K8s NUMA/extra bandwidth.
You can also learn more and check out other results on www.spec.org
could it be posible if it run crossfire + physic?
looks like theo was lying to us about dual cards then or what
I don't think he lied,he simply didn't know what was inside and just guessed.
And you have to rememeber one important thing here:the system was air cooled.No extreme cooling ,just some good ol' air.
hmm.. rd790 support ati autoexpress.. do you think it will boost the score to reached 30k?
I wasted 30 minutes reading through this.
Possibly the most attack filled e penis marathon ever at XS.
Useless information for all.
Fuad left INQ for reasons. IMO reasons like this. THEO exaggerates himself into a frenzy and then prints it.
If anyone believes Theo or this Hypothesised result at this point shame on you!
The only result of his article has been Flames across the intarweb! :(
Why doesn't anyone just listen to Perkham. 3d05 score.
Oh and I am an AMD fanboi! Touche!
it is not 3DMark05 score either since with that configuration it will be closer to 25k than 30k ;)
look at the CPU situation logically, if they bench 3d01 only really 1 CPU is being used as the bench is not multi threaded, now i know dual cores seemed to do better so it may use 2 of the 4 cores but the performance is still going to be a little lower. I actually feel 790 and a 6400+@3.6GHZ would be faster than a Phenom for 01 but only tests will show.
Now moving to TRi crossfire as that is what Theo saw, the fact is he did not report it but we have seen enough pics on the net since to show that there had got to have been 3 cards in that system...the issue for us is we don't know if the 3 cards had a single or had dual GPU's on each card.
CrossfireX has to be massively exciting for gamers the world over as they have known for quite a while money is better spent on video cards than getting the fastest CPU and with the possibility of running 4 cards linked up for massive res and huge frame rates this has to be truly mouth watering to them...i know it is for me and I hardly play any games ;)
In the end though all you guys doing math to try and disprove or those who are trying to prove whether this 30k run is true remember this...you have not yet got all the info you need, so all you are doing is speculating.
And lets be honest the degree of accuracy while speculation is in play could be miles off.
I will say this to finish, I see a new era of multi GPU's over hugely fast CPU's coming, ATI have Crossfire X and NV has TRI SLI all ready to go. I feel we will see some massive scores being posted real soon and they will all be using 3 and 4 card setups.
Now if NV are reading this I have a nice 4x4 system and 4xGTX's here....please show us how to align the bridges for TRI and quad SLI so we can start showing just what improvements are really made with multiple cards.
I ALSO have to add:
ATI haven't YET MANAGED to overclock it's CrossFire config SINCE the 1900s and you think that they can Overclock the Tri-Fire?.......and IF yes, well ONLY THEM could Overclock it a bit, thing that DOESN'T COUNT if people out there CAN'T do the SAME.......No?..... ;)
NVIDIA on the other hand, encoded PROBLEMS running those dual 7950s in SLI mod......Do you think that they can easily run THREE of them?.......It's JUST the begining so everything will NOT be as we want/imagine to.... ;)
EDIT: We have a TRI-SLI system here......do you think that it works as it should be?.... ;)
SO what do we have?......CPU POWER with TWO VGAs.......for now......so ......GIMME PLEASE an X38 chipset with SLI SUPPORT......Period...... :D :p: