Dragon quest VIII in HD hardware And settings opinion

[Squall Leonhart]

with nvidia cards atleast, NVAPI exposes the MCI bandwidth as a percentage and can be exposed with tools such as aida64.

[Alethram]

Total overall bandwidth as a rule is not to be relied on for a cards performance.

the Memory controller in the newer card for example might perform more efficiently, which brings up the Real memory access bandwidth making it perform better than the older cards.

I undestand your point But the new controllers compared

GTX 580 = 6x256MB

GTX 680 = 4x512MB

will give the extra 17% aprox needed to play the DQ8 at full speed?, is not a big assumption?, 17% more speed in pcsx2 is a loot just for memory controllers, and at least for me the 680 test aren't THAT impressive.

@miseru99

so then if with 194.4 GB/sec we got 50FPS (and hoping the bandwidth here is the only problem ) with 233.3 GB/sec and new memory controllers as squall said we can get DQ8 running at full speed with x5 ress, in nvidia cards at least.

Or maybe a little less bandwidth and a hell of memory controllers could do the trick, right?

that is pretty interesting maybe the next year we can got something like that, and will be about time for a upgrade.

[Alethram]

with nvidia cards atleast, NVAPI exposes the MCI bandwidth as a percentage and can be exposed with tools such as aida64.

I have aida64 right here, where can i check that?, i think i hijacked this thread, it is ok or maybe i should open a thread to talk about this in the hardware section?

[Squall Leonhart]

Well, it appears i was correct in assuming that the 680's real bandwidth capabilities are higher than the 580's

Pixel Fill is said to be memory bandwidth limited, and for good reason. Even with the lower clockspeed of the GTX 670, the fact that it has memory bandwidth equal to the GTX 680 means that it achieves equal performance in this test, confirming that GTX 670 can utilize its memory bandwidth just as well as GTX 680 can.

---

I have aida64 right here, where can i check that?, i think i hijacked this thread, it is ok or maybe i should open a thread to talk about this in the hardware section?

Display > GPU

[Squall Leonhart]

thats 33.641% more efficiency

[Alethram]

thats 33.641% more efficiency

That is nice, but it is a synthetic bench, the result of them not always are reflected on the applications, they are nice tools to see what our hard is capable of, but as i always say to a friend "i don't play 3d mark, that is for the bench boys" xD.

Thx for the aida tip, im going to check that tomorrow, now im going to take a sleep is 3am here and i have to wake up at 7 xD

and sorry again for my english i know that is awful but it is all i got atm xD

[Squall Leonhart]

Static benchmarks are perfectly fine for getting the peak bandwidth and fillrate capabilities between different gpu's.

[Omegaferrari]

Great info guys, it's great to see all these feedback, I'm truly happy to be able to play a few jrpgs and fighting games on my pc, that still look great, so I'm gonna donate to pcsx2 creators and website. pc gaming as whole lacks in these genres, so teoritically a 680 could run dq8 at 5x having the same results as my 580 at 4x, it doesn't go any lower than 47fps on my 580 at 5x.but it's still enough to make everything in game feel like bullet time at times.anyone with a gtx680 to confirm dq8 at 5x?

[rama]

Okay, I tested this with my GTX260 from 2008
I get 75FPS up until 4x scaling and 60FPS at 5x scaling.
So clearly memory bandwith isn't the only limiting factor in this game, my card has ~120GB/s.

I also fiddled a bit with GSdx settings and the skipdraw hack at 1 gives a free performance boost
back to 75FPS. It also fixes the shadows

[Squall Leonhart]

we didn't have the nvapi MCL information back in 2008 though ;p

at 4x, games like this can easily hit 85% on my 275