• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

(gamegpu) Battlefield 1 Benchmarks

lolfail9001 said:
Source for common knowledge of 960 being generally on par with 380?
Or Source for no so common knowledge of 380 tanking in slightly heavier titles [on CPU]?
Click to expand...

According to the link he posted the presumed 1050ti is 16.5% slower than the 380X in FSU, so it's a 50% exaggeration which , incidentally, coincides with the claims of new Polaris revision being 50% more efficient ;)
 
Ieldra said:
According to the link he posted the presumed 1050ti is 16.5% slower than the 380X in FSU, so it's a 50% exaggeration which , incidentally, coincides with the claims of new Polaris revision being 50% more efficient ;)
Click to expand...

Split the difference? :)

2225 vs 1853 => The leaked 1050ti need to be 20.1% than it is in the benchmark to match the 380x.

Your 16.5% would be from the 380x PoV. That it needs to be 16.5% slower to match a 1050ti.

Edit: which is actually what you said...sorry juggling work at the same time here. My bad.
 
Something was still bothering me, so I checked it a bit more...that 960 score looked *really* low (which it would need to be to support the 1050ti is 20% faster train of thought). Looking at 3dmark's actual site, for a i5-4460 + 960 (which is what PurePC reportedly used), shows that the average TS score is actually around...2330 for the top 300 or so, with plenty of plain old factory OC and stock cards in that mix. If a 1050ti is 20% faster than a 960, it certainly isn't in either of the benchmarks released so far. It must be 20% faster in everything else that hasn't been measured yet. :p
 
jbc029 said:
Something was still bothering me, so I checked it a bit more...that 960 score looked *really* low (which it would need to be to support the 1050ti is 20% faster train of thought). Looking at 3dmark's actual site, for a i5-4460 + 960 (which is what PurePC reportedly used), shows that the average TS score is actually around...2330 for the top 300 or so, with plenty of plain old factory OC and stock cards in that mix. If a 1050ti is 20% faster than a 960, it certainly isn't in either of the benchmarks released so far. It must be 20% faster in everything else that hasn't been measured yet. :p
Click to expand...
Sure, let's add 300 points to 1050 Ti score then, because 380 scores 2800 as well and not 2500. Same story!
 
lolfail9001 said:
Sure, let's add 300 points to 1050 Ti score then, because 380 scores 2800 as well and not 2500. Same story!
Click to expand...
I say let's wait for reviews. This GP107 part is rumored to have rated boost at around 1400mhz, let's assume it's 1500mhz in reality, it had 33% to go if it behaves like the other pascals. The way I see it this card will throttle to stay under 66W (which is the board limit for PCI-E) so the AIB version with supplemental power connector will probably perform better.
 
Ieldra said:
I say let's wait for reviews. This GP107 part is rumored to have rated boost at around 1400mhz, let's assume it's 1500mhz in reality, it had 33% to go if it behaves like the other pascals. The way I see it this card will throttle to stay under 66W (which is the board limit for PCI-E) so the AIB version with supplemental power connector will probably perform better.
Click to expand...
Well, The Stilt claims 1050 Ti is 100W TDP card, actually. So i would definitely agree to wait for reviews.
 
lolfail9001 said:
Sure, let's add 300 points to 1050 Ti score then, because 380 scores 2800 as well and not 2500. Same story!
Click to expand...

If you can find a commensurate sample size that show the 1050ti should have that adjustment, I'd be more than happy to agree.


Go right ahead, I'll wait.
 
jbc029 said:
If you can find a commensurate sample size that show the 1050ti should have that adjustment, I'd be more than happy to agree.


Go right ahead, I'll wait.
Click to expand...
My point is that we have no clue about circumstances of 1050 Ti's Time Spy run. So, why not join the overall trend and add 300 points for being pretty :)? Or just wait for actual reviews with actual games.
 
And my point is that unsubstantiated claims of outstanding performance gains based on dubious metrics should be just as scoffed at as any other claim. Yet somehow, they aren't.
 
jbc029 said:
And my point is that unsubstantiated claims of outstanding performance gains based on dubious metrics should be just as scoffed at as any other claim. Yet somehow, they aren't.
Click to expand...
Just like unsubstantiated claims of X being faster than Y when we have no actual clue about Y's performance. Remember those?
 
Plus updated drivers can help a lot. Like Vulkan Nvidia and Doom -> At first Nvidia kinda suck, now Vulcan and Doom works rather well with Nvidia Pascal cards. Maybe not as big as jump as AMD but rather decent.
 
AzixTGO said:
are those single figure minimums or do they include some type of frequency consideration? A single low minimum could happen for many reasons and be irrelevant for gameplay. eg. in some games on load a GPU might have a bit of hitching then be fine rest of the time.
Click to expand...
I think they're actually minimums, people should stop using those and just look at 90th and 99th percentiles
 
SWUUaGg.png

Battlefield 1 PC: Tolle Optik, saubere Performance- erste Benchmarks der Release-Version
 
You must log in or register to reply here.
Back
Top