help in upgrading the pc...
#41
Most of the games use various caching techniques for the modern gpu's so, the performance varies most of the time. The passmark benchmarks seem much more accurate for emulation excluding some games like shadow hearts which favour NVIDIA gpu's.
We're supposed to be working as a team, if we aren't helping and suggesting things to each other, we aren't working as a team.
- Refraction
Reply

Sponsored links

#42
Passmark benchmarks(Especially CPU ones) are extremely good for predicting emulation performance. They are pretty good overall too, as they use real world type loads.

I don't think they favor Nvidia. I mean, it's a compilation of benchmarks users run on their own machines. In fact, Comparing my overclocked 7870 to my 670, it predicted the 670 would only be about 10% faster, when in reality it's about 20%. So in that case it favored AMD. It just depends on the load.
[Image: XTe1j6J.png]
Gaming Rig: Intel i7 6700k @ 4.8Ghz | GTX 1070 TI | 32GB RAM | 960GB(480GB+480GB RAID0) SSD | 2x 1TB HDD
Reply
#43
(04-22-2015, 07:12 PM)Blyss Sarania Wrote: Passmark benchmarks(Especially CPU ones) are extremely good for predicting emulation performance. They are pretty good overall too, as they use real world type loads.

I don't think they favor Nvidia. I mean, it's a compilation of benchmarks users run on their own machines. In fact, Comparing my overclocked 7870 to my 670, it predicted the 670 would only be about 10% faster, when in reality it's about 20%. So in that case it favored AMD. It just depends on the load.

AMD is generally better in OS X. Linux... eh, it really depends.

Windows itself isn't biased, but many games are Nvidia biased*cough*WATCHDOGS*cough*. The Kaveri APUs seem to work better in CAD-based sh*t than even high-end Nvidia gaming cards from what I hear

Emulators don't really seem to be biased at all. Just what works and what doesn't.
Moe: [to Homer] And I'm pulling your favorite song out of the jukebox.
Homer: "It's Raining Men"?!
Moe: Yeah, not no more, it ain't. [throws the record out the window which lands into Smithers' car]
Smithers: Ow! [looks at the record] Ohhh.

what would ;e; do
Reply
#44
(04-22-2015, 07:12 PM)Blyss Sarania Wrote: Passmark benchmarks(Especially CPU ones) are extremely good for predicting emulation performance. They are pretty good overall too, as they use real world type loads.

I don't think they favor Nvidia. I mean, it's a compilation of benchmarks users run on their own machines. In fact, Comparing my overclocked 7870 to my 670, it predicted the 670 would only be about 10% faster, when in reality it's about 20%. So in that case it favored AMD. It just depends on the load.

I wasn't saying the company is biased, just that the tests seem to favor nvidia.
Maybe only in some cases though, but the 780 Ti is definitely not 30% faster than a 290x.

It looks like their test only uses DX10 anyway. Which precludes them from using a lot of features that modern games use. Features that AMD GPUs are particularly good at.

I guess that doesn't really matter in the case of pcsx2 though.
Reply
#45
Nah the test runs DX9, 10 and 11 test separately, as well as OpenCL tests.

(04-22-2015, 07:20 PM)Marge Simpson Wrote: AMD is generally better in OS X. Linux... eh, it really depends.

Windows itself isn't biased, but many games are Nvidia biased*cough*WATCHDOGS*cough*. The Kaveri APUs seem to work better in CAD-based sh*t than even high-end Nvidia gaming cards from what I hear

Emulators don't really seem to be biased at all. Just what works and what doesn't.

Many games are biased to either AMD or Nvidia. The companies pay for it to happen. "AMD Gaming Evolve" "Nvidia the way it's meant to be played" etc
[Image: XTe1j6J.png]
Gaming Rig: Intel i7 6700k @ 4.8Ghz | GTX 1070 TI | 32GB RAM | 960GB(480GB+480GB RAID0) SSD | 2x 1TB HDD
Reply
#46
(04-22-2015, 07:37 PM)Blyss Sarania Wrote: Nah the test runs DX9, 10 and 11 test separately, as well as OpenCL tests.


Well, I see that they have DX9/DX10 and Direct Compute tests.

Anyway, I still wouldn't use or recommend it as a guide for PC game performance.

For pcsx2, I dunno. Any decent graphics card should do the trick.
Reply
#47
I've used it for every system I've built. Whereas half the "review" sites on the net are bought out by either Nvidia or AMD, Passmark is fed user data only, so even if one or two people managed to fake it, it would get lost in the average.

And I run the test on my own PC when I overclock.

There are two DX9 tests - they involve jets flying around over a swampy place. One windowed, one fullscreen.
The DX10 test is a meteor shower falling on floating islands with trees.
The DX11 test is on the moon with a moonbase and tons of floating jellyfish aliens.
Then direct compute tests are bitonic sorting, mandelbrot fractal zoom, and sphere deformation.
[Image: XTe1j6J.png]
Gaming Rig: Intel i7 6700k @ 4.8Ghz | GTX 1070 TI | 32GB RAM | 960GB(480GB+480GB RAID0) SSD | 2x 1TB HDD
Reply
#48
(04-22-2015, 07:37 PM)Blyss Sarania Wrote: Nah the test runs DX9, 10 and 11 test separately, as well as OpenCL tests.


Many games are biased to either AMD or Nvidia. The companies pay for it to happen. "AMD Gaming Evolve" "Nvidia the way it's meant to be played" etc

At least the games that are AMD biased work decent on Nvidia too
Moe: [to Homer] And I'm pulling your favorite song out of the jukebox.
Homer: "It's Raining Men"?!
Moe: Yeah, not no more, it ain't. [throws the record out the window which lands into Smithers' car]
Smithers: Ow! [looks at the record] Ohhh.

what would ;e; do
Reply
#49
(04-22-2015, 07:53 PM)Marge Simpson Wrote: At least the games that are AMD biased work decent on Nvidia too

Not always true.

Neither is it true that all Nvidia biased games perform ***** on AMD. It all depends on how far it's been taken.
[Image: XTe1j6J.png]
Gaming Rig: Intel i7 6700k @ 4.8Ghz | GTX 1070 TI | 32GB RAM | 960GB(480GB+480GB RAID0) SSD | 2x 1TB HDD
Reply
#50
(04-22-2015, 07:56 PM)Blyss Sarania Wrote: Not always true.

Neither is it true that all Nvidia biased games perform ***** on AMD. It all depends on how far it's been taken.

Nyeh.

Some Nvidia-biased games do perform well on AMD; FlatOut 2, The Crew (is that biased).

Most AMD-biased games in turn do really well on Nvidia too: Dirt Showdown, Saints Row IV, Grand Theft Auto V, etc.

Intel-biased games perform decently on both platforms *cough*GRID2*cough*
Moe: [to Homer] And I'm pulling your favorite song out of the jukebox.
Homer: "It's Raining Men"?!
Moe: Yeah, not no more, it ain't. [throws the record out the window which lands into Smithers' car]
Smithers: Ow! [looks at the record] Ohhh.

what would ;e; do
Reply




Users browsing this thread: 1 Guest(s)