IanCutress

GT430: 96 SPs @ 700 MHz GT520: 48 SPs @ 810 MHz GT610: 48 SPs @ 810 MHz

? Been sitting on my shelf for at least a couple of years

I didn't say it would beat it, just that they exist Seriously though, it's either a 5450/430/520/610 for PCI in that round.

It's either that or a GT 430

Yeah, there are a few dual GPU PCI cards worth searching for

BOOOOM great score

Save your result, then upload it to the website, rather than submit direct from the benchmark.

GPU-Z window?

I admit that the whole debacle has made it confusing. What AMD call Hybrid graphics is different to how it's classified here at HWBot. I think AMD steered away from the hybrid naming scheme in Llano because you can have two different GPU combinations ended up with the same name, and end users might get confused. But the end of the day: Best desktop hybrid combination for the contest is HD6550D + HD6670 = HD6690D2. If MM wants to declare that any 'Dual Graphics' AMD solution (which will be in the GPU-Z name, and should probably get its own HWBot category) is fine then I'm ok with that, but as it is written, the list above I provided is what's required. On a side note. Next year: Lucid graphics? Must use 1 AMD and 1 NVIDIA GPU? How about 'PCIe at x1'

Valid combinations for Hybrid stage: HD6410D + HD6350 = HD6430D2 HD6530D + HD6450 = HD6550D2 HD6530D + HD6570 = HD6610D2 HD6530D + HD6670 = HD6690D2 HD6550D + HD6450 = HD6550D2 HD6550D + HD6570 = HD6630D2 HD6550D + HD6670 = HD6690D2 HD6410D + HD6350A = HD6430D2 HD6410D + HD6450A = HD6510A2 HD6410D + HD6550A = HD6570A2 HD6530D + HD6450A = HD6550A2 HD6530D + HD6550A = HD6590A2 HD6530D + HD6650A = HD6670A2 HD6530D + HD6750A = HD6710A2 HD6550D + HD6450A = HD6550A2 HD6550D + HD6550A = HD6610A2 HD6550D + HD6650A = HD6690A2 HD6550D + HD6750A = HD6730A2 HD6480G + HD6430M = HD6510G2 HD6480G + HD6450M = HD6510G2 HD6480G + HD6470M = HD6510G2 HD6480G + HD6490M = HD6515G2 HD6520G + HD6450M = HD6540G2 HD6520G + HD6470M = HD6540G2 HD6520G + HD6490M = HD6545G2 HD6520G + HD6630M = HD6580G2 HD6520G + HD6650M = HD6720G2 HD6520G + HD6730M = HD6740G2 HD6520G + HD6750M = HD6755G2 HD6520G + HD6770M = HD6775G2 HD6620G + HD6450M = HD6640G2 HD6620G + HD6470M = HD6640G2 HD6620G + HD6490M = HD6645G2 HD6620G + HD6630M = HD6690G2 HD6620G + HD6650M = HD6740G2 HD6620G + HD6730M = HD6760G2 HD6620G + HD6750M = HD6755G2 HD6620G + HD6770M = HD6775G2 Basically anyone who could read a table should have got that from the CPU-World page.

He probably has not read the rules. It seems counter intuitive to let people change # of threads for HyperPi, but I have a feeling that the software doesn't always spawn them equally to one thread each.

Correct, according to what MM has said in this thread, those would not be valid.

Not a valid one for the contest. It'll just show HD8670D + 6670 in GPU-Z. You need one with a whole new designation: anything in this link that has a new name is valid - http://www.cpu-world.com/info/AMD/Recommended_graphics_cards_for_AMD_dual-graphics.html

Ticket ID: 1810 Priority: Low http://valid.canardpc.com/a/cfn0b3\r\n\r\nRequest adding Snapdragon 600 (Krait 300)

PCIe x1 to PCI adaptor, so you don't have to go find a board with PCI. Though does it work for video output?

Memories?

Points make a lot of difference

Damn, only 102 motherboards. I know there's several Z87 I've tested that I forgot to BCLK test...

Yes, Lippokratis reported three of my scores today. HyperPi rules: allowed optimsations: change number of processes http://forum.hwbot.org/showthread.php?t=80198&highlight=hyperpi

Quickfast: yes, you can submit on a new socket, as long as all seven scores that count are with different sockets Lucky_n00b: nice list

Good stuff

wPrime has issues where spawning threads takes longer than running threads. You also have to set the number of threads in the advanced options menu. Even with 160 threads, takes 2-3 secs.

I was saying that to Scott's misrepresentation to diminishing returns on performance without the qualifier on hardware to which he was referring.

Nope, I write code I deal in performance first I have written a couple of benchmarks (not on HWBot) that act like this, to see where L2/L3 cache plays a role as you increase problem set sizes. Sometimes it's better to test this way in order to maximise throughput and minimise out-to-memory transfers, and you get an overall peak value at the end. I change my problem set sizes over time as well. If anyone has ever looked at n-body calculations, the complexity of a problem rises as a square of the number of bodies in a simulation. At some point your memory requirements will be more than L2, more than L3, and then require looking back to main memory. Keeping out of main memory is a lot quicker in simulation steps. Or a good coder can hide that latency

Law of diminishing returns in the context of code applies to performance. You mean to apply it to hardware stress, which is valid but confusing without an additional qualifier in your phrasing.