Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [WCCF] Nvidia's Volta could launch as early as 2H 2017 ON 16nm node
New Posts  All Forums:Forum Nav:

[WCCF] Nvidia's Volta could launch as early as 2H 2017 ON 16nm node - Page 21

post #201 of 272
A sample size of 1 (plus a few unicorn rare golden samples) doesn't invalidate what I said.
post #202 of 272
I am sure there are far more common 1550mhz 980tis than 1300mhz Fury X, hell, i am sure it will be more than 1200mhz Fury X lol.
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
post #203 of 272
Quote:
Originally Posted by magnek View Post

A sample size of 1 (plus a few unicorn rare golden samples) doesn't invalidate what I said.

It could if we consider the possibility of people who cares about overclocking going for the best overclocker, the 980Ti while people who just want something convenient and cool just buy the Fury X.

I believe the average Fury X owner is looking for great performance packed in a convinient closed loop that can serve them well for gaming without much thought into it, people who don't know don't want or can't install a custom loop in their cards

Also the Fury X is exponentially less popular compared to Nvidia's offering which lowers the accuracy of any statistic like these, the more units measured = the more accurate it becomes.

However, even if the Fury X statistic is accurate AS IS then those 15MHz will probably provide less performance than what the next Crimson Driver will smile.gif
Edited by Dargonplay - 7/30/16 at 6:55pm
post #204 of 272
Quote:
Originally Posted by Kpjoslee View Post

I am sure there are far more common 1550mhz 980tis than 1300mhz Fury X, hell, i am sure it will be more than 1200mhz Fury X lol.

If you truly believe that HWBOt is right about the average Fury X being 1135MHz then you'd have no problem believing that the average 980Ti is 1382MHz

http://hwbot.org/hardware/videocard/geforce_gtx_980_ti/

The difference being 247MHz between the Fury X and 980Tis average, if the 980Ti tops out at 1550MHz for 5% of samples then using these AVERAGES the Fury X should top out at 1303MHz for the same percent of samples.

But yeah you're right, there are probably more 1550MHz 980Tis than 1200MHz Fury Xs... But only because the 980Ti have sold so many more cards compared to the fury X that even if only a 5% of 980Tis can reach 1550MHz there would still be tons more 1550MHz 980Tis than 1200MHz Fury Xs, even if 95% of Fury Xs could reach 1200MHz like mine lol.

I mean there are 25K of 980Tis on HWBOt to average them down, even if only 5% of that can do 1550MHz that's still 1250 different 980Tis that can do 1550MHz, that's literally more than the entire HWBOt samples of 540 Fury Xs lmao.
Quote:
Originally Posted by NikolayNeykov View Post

Every Ti can reach 1455 Mhz that's for sure, if it cannot there is problem with the card.
I'd say average on this card is 1480 -1500 Mhz
It's just quality made card which can be pushed a lot.

That would make every 980Ti is as fast as a GTX 1080 at 1700MHz boost, why do benchmarks comparing the two of them don't agree with you? redface.gif

I've personally seen a lightning 980Ti with triple PCIe Power connectors top out @ 1450MHz stable and it didn't looked like something every 980Ti could do.
post #205 of 272
Quote:
Originally Posted by Dargonplay View Post

If you truly believe that HWBOt is right about the average Fury X being 1135MHz then you'd have no problem believing that the average 980Ti is 1382MHz

http://hwbot.org/hardware/videocard/geforce_gtx_980_ti/

The difference being 247MHz between the Fury X and 980Tis average, if the 980Ti tops out at 1550MHz for 5% of samples then using these AVERAGES the Fury X should top out at 1303MHz for the same percent of samples.

But yeah you're right, there are probably more 1550MHz 980Tis than 1200MHz Fury Xs... But only because the 980Ti have sold so many more cards compared to the fury X that even if only a 5% of 980Tis can reach 1550MHz there would still be tons more 1550MHz 980Tis than 1200MHz Fury Xs, even if 40% of Fury Xs could reach 1200MHz like mine.
Every Ti can reach 1455 Mhz that's for sure, if it cannot there is problem with the card and i feel pity for the owner who doesn't return it.
I'd say average on this card is 1480 -1500 Mhz
It's just quality made card which can be pushed a lot.
post #206 of 272
Quote:
Originally Posted by magnek View Post


980 Ti @ 1405/7200 matches a 1080 @ 1801/10000, so probably looking at another CPU bottleneck. Did they test 4K? Otherwise something's really funny with the results.
found another ones. Warning: Spoiler! (Click to show)

I wondered why they reduced the details to high or even medium with resolution but kept the TSAA8X on, then I realized that the "wonder card" has 4GB of vram, but it has async shaders
post #207 of 272
Quote:
Originally Posted by Klocek001 View Post

Quote:
Originally Posted by magnek View Post


980 Ti @ 1405/7200 matches a 1080 @ 1801/10000, so probably looking at another CPU bottleneck. Did they test 4K? Otherwise something's really funny with the results.
found another ones. Warning: Spoiler! (Click to show)

I wondered why they reduced the details to high or even medium with resolution but kept the TSAA8X on, then I realized that the "wonder card" has 4GB of vram, but it has async shaders
Those results look funny. Funny as in, not congruent with what we've been seeing from major reviewers so far.
    
CPUMotherboardGraphicsRAM
Intel 6700k 4.5 GHz 1.3v Asus Z170i MSI 980Ti 1490/7760 MHz G.skill DDR4 8 GB x2 3733 MHz 15-15-15-35-1 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 1 TB Crucial M4 256 GB NH-C14S Windows 10 Student 
MonitorKeyboardPowerCase
TBD Cooler Master Quick Fire TK Corsair SF600 Fractal Core 500 
MouseMouse PadAudioAudio
Zowie EC2-A Zowie G TF-X Fiio E17 v1 Sennheiser HD 598 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel 6700k 4.5 GHz 1.3v Asus Z170i MSI 980Ti 1490/7760 MHz G.skill DDR4 8 GB x2 3733 MHz 15-15-15-35-1 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 1 TB Crucial M4 256 GB NH-C14S Windows 10 Student 
MonitorKeyboardPowerCase
TBD Cooler Master Quick Fire TK Corsair SF600 Fractal Core 500 
MouseMouse PadAudioAudio
Zowie EC2-A Zowie G TF-X Fiio E17 v1 Sennheiser HD 598 
  hide details  
Reply
post #208 of 272
Quote:
Originally Posted by ToTheSun! View Post

Those results look funny. Funny as in, not congruent with what we've been seeing from major reviewers so far.
I like the DF video, seems like they're trying to achieve objectivity

Shows 1080 ahead of all the others (21% over Fury X, 30% over 980Ti), Fury X just managing to beat a reference 1070 and 980Ti by 7.5% in Vulkan and reference 980 tied with rx480.
Vulkan shows little gains for nvidia cards, but then again they were already pushing the game really fine under OGL.

1440p/ultra/8x TSSAA



another scene




CPU intensive scene





avg. of these 3 scenes in Vulkan:

1080: 155 fps
1070: 119 fps (-23%)
980Ti:118 fps
Fury X:125 fps (-19,5%)
rx480:




So Fury X is 6% faster than reference 980Ti in Vulkan with async on. It'd smell the behind of both 1070 and 980Ti if they were custom ones.
Edited by Klocek001 - 7/31/16 at 10:42am
post #209 of 272
Quote:
Shows 1080 ahead of all the others (21% over Fury X, 30% over 980Ti), Fury X just managing to beat a reference 1070 and 980Ti by 7.5% in Vulkan and reference 980 tied with rx480.
that seems to be the case


for warhammer - http://www.overclock.net/t/1604750/dsog-report-total-war-warhammer-runs-27-slower-in-dx12-on-nvidia-s-hardware/330#post_25395562
The Green Beast
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 3770K @ 4500 Mhz ASRock Z77 Pro3 Gigabyte GTX 1080 G1 Gaming Crucial Ballistix 2x8GB DDR3-1600 
Hard DriveHard DriveHard DriveCooling
SSD Crucial M550 500GB SSD Samsung 850 Evo 1TB HDD Seagate 7200rpm 3TB Cooler Master Hyper 212 EVO; Xilence X5 
OSMonitorKeyboardPower
Windows 10 Pro 64-bit Acer Predator XB271HU 27" IPS Gsync 1440p 165Hz CM Storm QuickFire XT Cherry Red 800W modular 
CaseMouseMouse PadAudio
Fractal Design Define R4 Black Logitech G900 Chaos Spectrum SteelSeries QcK+ 4mm SK Gaming Realtek On-board 
  hide details  
Reply
The Green Beast
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 3770K @ 4500 Mhz ASRock Z77 Pro3 Gigabyte GTX 1080 G1 Gaming Crucial Ballistix 2x8GB DDR3-1600 
Hard DriveHard DriveHard DriveCooling
SSD Crucial M550 500GB SSD Samsung 850 Evo 1TB HDD Seagate 7200rpm 3TB Cooler Master Hyper 212 EVO; Xilence X5 
OSMonitorKeyboardPower
Windows 10 Pro 64-bit Acer Predator XB271HU 27" IPS Gsync 1440p 165Hz CM Storm QuickFire XT Cherry Red 800W modular 
CaseMouseMouse PadAudio
Fractal Design Define R4 Black Logitech G900 Chaos Spectrum SteelSeries QcK+ 4mm SK Gaming Realtek On-board 
  hide details  
Reply
post #210 of 272
Quote:
Originally Posted by Klocek001 View Post

I like the DF video, seems like they're trying to achieve objectivity

[...]

Shows 1080 ahead of all the others (21% over Fury X, 30% over 980Ti), Fury X just managing to beat a reference 1070 and 980Ti by 7.5% in Vulkan and reference 980 tied with rx480.
Vulkan shows little gains for nvidia cards, but then again they were already pushing the game really fine under OGL.
Yeah. It's a shame the Fury X has 4GB of VRAM. At discounted prices, with its advantage in DX12/Vulkan, it could be a nice alternative to a 1070. Alas, it's still slightly more expensive and with half the VRAM.

I'm honestly looking for any and all reasons to stick with AMD instead of switching over to the green camp, but Raja is not making it easy.
    
CPUMotherboardGraphicsRAM
Intel 6700k 4.5 GHz 1.3v Asus Z170i MSI 980Ti 1490/7760 MHz G.skill DDR4 8 GB x2 3733 MHz 15-15-15-35-1 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 1 TB Crucial M4 256 GB NH-C14S Windows 10 Student 
MonitorKeyboardPowerCase
TBD Cooler Master Quick Fire TK Corsair SF600 Fractal Core 500 
MouseMouse PadAudioAudio
Zowie EC2-A Zowie G TF-X Fiio E17 v1 Sennheiser HD 598 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel 6700k 4.5 GHz 1.3v Asus Z170i MSI 980Ti 1490/7760 MHz G.skill DDR4 8 GB x2 3733 MHz 15-15-15-35-1 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 1 TB Crucial M4 256 GB NH-C14S Windows 10 Student 
MonitorKeyboardPowerCase
TBD Cooler Master Quick Fire TK Corsair SF600 Fractal Core 500 
MouseMouse PadAudioAudio
Zowie EC2-A Zowie G TF-X Fiio E17 v1 Sennheiser HD 598 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Rumors and Unconfirmed Articles
Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [WCCF] Nvidia's Volta could launch as early as 2H 2017 ON 16nm node