TR Forums

JohnC wrote:

B.t.w, it looks like you're not using the latest drivers... Upgrade them :wink:

314.07 is the latest non-TITAN driver, huh? The 314.09 driver is only for the TITAN I think. I guess he could be using the BETA drivers, but meh.

Since overclocking seems to be a "thing" to do (just kidding), here is mine with clocks 950/1450 vs the original post (and speed) of 880/1375:
Again, 13.2 beta 6 drivers, 4.1Ghz CPU. (Love how most tools don't pick up on the turbo speed settings...CPUZ does though)
EDIT: About 980Mhz is the highest I can get my card without it freaking out completely. Hitting 1.0 with this card is my goal....someday.

...well, just re-ran same test, using the stock settings (no TDP or temp limit changes). The good news is the difference between "overclocked" settings seem to be pretty insignificant:



The "bad" news is that the card reaches the 80C temp limit very fast at max GPU load (using stock fan speed settings), and after that - starts throttling down to non-boost "base clock" (which is 876MHz on this model), making the GPU clock speed graph look not very smooth... This temp limit throttling can be easily fixed by bumping up the temp limit slider to its max value (95C), making the GPU clock speed graph completely "smooth", but I'm very curious as to what effect it may have on framerate "smoothness" or something like this (personally I haven't noticed any visual issues when playing BF3 or Crysis 3) when running at completely stock temp limit settings... And I've also read curious things about this new "Boost 2.0" at overclock.net's forums (it seem to be working in a slightly mysterious ways). Would be VERY interesting for Scott to test this feature in more details, using different Titan cards (the "regular" and "overclocked" versions like EVGA's SC models) and different cooling methods and different games/tests with high GPU utilization (like Crysis 3 or this Valley benchmark)!

JohnC wrote:

...well, just re-ran same test, using the stock settings (no TDP or temp limit changes). The good news is the difference between "overclocked" settings seem to be pretty insignificant:

I'm rather shocked that your minium FPS is only 2.5 FPS greater than mine.

End User wrote:

JohnC wrote:

...well, just re-ran same test, using the stock settings (no TDP or temp limit changes). The good news is the difference between "overclocked" settings seem to be pretty insignificant:

I'm rather shocked that your minium FPS is only 2.5 FPS greater than mine.

Eh, it's pretty random... I've seen ppl on overclock.net site getting a 7.5FPS as minimum even with powerful GPU's... The "maximum" seem to be more reliable indicator.

End User wrote:

JohnC wrote:

...well, just re-ran same test, using the stock settings (no TDP or temp limit changes). The good news is the difference between "overclocked" settings seem to be pretty insignificant:

I'm rather shocked that your minium FPS is only 2.5 FPS greater than mine.

You shouldn't really be; Valley is really memory-bandwidth-limited and your cards have the same memory interface width (384-bit), so ...

auxy wrote:

End User wrote:

JohnC wrote:

...well, just re-ran same test, using the stock settings (no TDP or temp limit changes). The good news is the difference between "overclocked" settings seem to be pretty insignificant:

I'm rather shocked that your minium FPS is only 2.5 FPS greater than mine.

You shouldn't really be; Valley is really memory-bandwidth-limited and your cards have the same memory interface width (384-bit), so ...

Nah, it's just a poor benchmark overall... If you'll watch it running - you'll notice that the lowest "minimum FPS" drop occurs during the scene "transitions", where nothing is rendered on screen... Also, if you have some kind of background process that is using up considerable system resources - you might also get a "dip" in minimum FPS. Nothing to do with "memory bandwidth"

JohnC wrote:

Nah, it's just a poor benchmark overall... If you'll watch it running - you'll notice that the lowest "minimum FPS" drop occurs during the scene "transitions", where nothing is rendered on screen... Also, if you have some kind of background process that is using up considerable system resources - you might also get a "dip" in minimum FPS. Nothing to do with "memory bandwidth"

Well the bench is still version 1. I go off the max too.

End User wrote:

Like that's real. What kind of magical 570 do you have?

This is an SLI setup... Which this garbage benchmark doesn't properly detect

JohnC wrote:

This is an SLI setup...

Bingo

It's also rather easy to cheat this benchmark with speedhacks ... eheh. ┐(￣ヮ￣)┌

End User wrote:

JohnC wrote:

This is an SLI setup...

Bingo

Ah.. I see. Well SLI is at least.. um .. moderately magical.

auxy wrote:

It's also rather easy to cheat this benchmark with speedhacks ... eheh. ┐(￣ヮ￣)┌

Care to enlighten us?

cynan wrote:

auxy wrote:

It's also rather easy to cheat this benchmark with speedhacks ... eheh. ┐(￣ヮ￣)┌

Care to enlighten us?

Hehe, I guess you aren't much of a cheater. Errm, not that I am! But those who fight devils... etc, etc.

I don't really want to get into specifics here because someone might use it. Well, I probably shouldn't have brought it up altogether, eheh. It's like screaming HAX!! in an FPS game... (;´・`)

I don't know about "Valley", but previous Unigine's benchmark used to calculate the "score" by taking the total number of rendered frames and dividing it by a predefined, fixed number (that number was the time it supposedly takes the benchmark to run under normal conditions). Sooo... If you can make the benchmark to run for longer period of time (which will increase the total number of rendered frames) - you can increase the final "score"
http://hwbot.org/forum/showthread.php?t=13369
However, that doesn't really change the FPS scores.

Long story short: The "Score" value of Unigine's benchmarks is pointless, so is "Min FPS"

I'll post some TITAN benchmarks tonight when I get home! I need some 'alone time' with my PC and the card first tho.

^{Because it's gonna take me a really long time to install it -- moving that big heavy thing around is hard! (┳Д┳)}

auxy wrote:

I'll post some TITAN benchmarks tonight when I get home! I need some 'alone time' with my PC and the card first tho.

^{Because it's gonna take me a really long time to install it -- moving that big heavy thing around is hard! (┳Д┳)}

All the sexual puns intended.

ClickClick5 wrote:

auxy wrote:

I'll post some TITAN benchmarks tonight when I get home! I need some 'alone time' with my PC and the card first tho.

^{Because it's gonna take me a really long time to install it -- moving that big heavy thing around is hard! (┳Д┳)}

All the sexual puns intended.

ClickClick5 wrote:

All the sexual puns intended. :D

W-WHAT? (((((;ﾟДﾟ)))))
I h-have no idea what you're talking about! ｡：ﾟ(｡ﾉД＼｡)ﾟ･｡

auxy wrote:

I'll post some TITAN benchmarks tonight when I get home! I need some 'alone time' with my PC and the card first tho.

^{Because it's gonna take me a really long time to install it -- moving that big heavy thing around is hard! (┳Д┳)}

Longer than I thought!

See my rant here.

auxy wrote:

auxy wrote:

I'll post some TITAN benchmarks tonight when I get home! I need some 'alone time' with my PC and the card first tho.

^{Because it's gonna take me a really long time to install it -- moving that big heavy thing around is hard! (┳Д┳)}

Longer than I thought!

See my rant here.

Sooooo, the bench pic is where?!

Also, nice overall FPS. I still cant fathom buying a $1k GPU, CPU, or a single part. My complete system costs about $1200 each build.

ClickClick5 wrote:

Sooooo, the bench pic is where?!

Did you read it?! I didn't get a chance to run it! I was up 'til 4:30AM just getting my machine to POST! (╥﹏╥) And now I think I'm sick...

ClickClick5 wrote:

Also, nice overall FPS. I still cant fathom buying a $1k GPU, CPU, or a single part. My complete system costs about $1200 each build.

That wasn't Valley FPS! x3 That was just a quick playthrough of Blacklight: Retribution.

Personal story incoming: I'm not a wealthy person; I have a nice enough PC, but all my "disposable" income more or less has gone into my PC for the last few years, and I work full-time. Until not-too-long-ago, I lived with my parents, and I was able to spend (read: waste) a lot of my money on computer parts. The last year and a half or so my girlfriend and I had our own place, but even limiting my spending to virtually nothing, we still can't afford it on my income plus what she makes as a bartender, so now we moved back in with her parents, since they have a gigantic house and we can have an entire floor (the top floor) all to ourselves.

Anyway, the point is, I would never have spent $1k on a GPU myself; there's no way I could ever afford something like this. I could maybe save and buy one, for sure, but justifying the expense would be impossible; I have too many other things in my life that I have to pay for.

So, Thanks JohnC! ♥～(‘▽^人)

So, Thanks JohnC! ♥～(‘▽^人)

That's pretty cool.


When you get home take a nap, then run the benchmark for us

Back on subject

My desktop got a bit under 1100 for those who care.

Surface pro (it's what I use as my laptop)
66 (I let it warm up for 5 minutes then ran the test to keep turbo off the score)

thecoldanddarkone wrote:

Surface pro (it's what I use as my laptop)
66 (I let it warm up for 5 minutes then ran the test to keep turbo off the score)

LOL! I'm pretty surprised it was even able to run it on that pathetic Intel graphics~! Pretty amazing! Does it run okay without AA, or maybe in 720p without AA? (⊙△⊙✿)

auxy wrote:

ClickClick5 wrote:

Sooooo, the bench pic is where?!

Did you read it?! I didn't get a chance to run it! I was up 'til 4:30AM just getting my machine to POST! (╥﹏╥) And now I think I'm sick...

ClickClick5 wrote:

Also, nice overall FPS. I still cant fathom buying a $1k GPU, CPU, or a single part. My complete system costs about $1200 each build.

That wasn't Valley FPS! x3 That was just a quick playthrough of Blacklight: Retribution.

Personal story incoming: I'm not a wealthy person; I have a nice enough PC, but all my "disposable" income more or less has gone into my PC for the last few years, and I work full-time. Until not-too-long-ago, I lived with my parents, and I was able to spend (read: waste) a lot of my money on computer parts. The last year and a half or so my girlfriend and I had our own place, but even limiting my spending to virtually nothing, we still can't afford it on my income plus what she makes as a bartender, so now we moved back in with her parents, since they have a gigantic house and we can have an entire floor (the top floor) all to ourselves.

Anyway, the point is, I would never have spent $1k on a GPU myself; there's no way I could ever afford something like this. I could maybe save and buy one, for sure, but justifying the expense would be impossible; I have too many other things in my life that I have to pay for.

So, Thanks JohnC! ♥～(‘▽^人)

Lol, sick or dead, where are my results?!

ClickClick5 wrote:

Lol, sick or dead, where are my results?!

I'm still at work! ｡･ﾟﾟ･(>д<)･ﾟﾟ･｡

I work 2:30pm - 11pm Central US time (UTC-5 because of daylight savings time).

here's my video for VALLEY

http://www.youtube.com/watch?v=svs-TdDGPvA

[quote="auxy"] pic here[quote]

ABOUT BLOODY TIME!!! Not bad I say. Not bad.