Overclock.net › Forums › Graphics Cards › AMD/ATI › ATI Drivers and Overclocking Software › AMD GPU Drivers: The Real Truth.
New Posts  All Forums:Forum Nav:

AMD GPU Drivers: The Real Truth. - Page 40

post #391 of 451
Quote:
Originally Posted by Kana-Maru View Post

Interesting.

As a previous Kepler user for 2+ years [now Fury X] I witnessed and actual benchmarked the constant loses. My FireStrike scores dropped tremendously. Performance was clear getting worse when the 970 dropped. My performance with AMD this time around has increased and that's what I continue to look forward to.

It's actually not interesting at all. It's complete speculation based on data that has since been proven wrong.

The entire part where he speculates about why Fallout 4 "lost" performance because of what some random forum user stated has all turned out to be utter nonsense.

Besides the fact that no actual users reported any performance losses, we now have results from two sources that render the gamegpu results invalid.

Source 1

Source 2

[/quote]

The sad part is we have people out there that take any shred of "evidence" and turn it into a story to garner page hits. It happens to both Nvidia and AMD.

And Kepler losing performance is also a myth as well. In Techpowerup's test suite, there are 4 games that were tested at 980 launch and are still tested today. 3 show Kepler performance being better and 1 shows a 1 fps loss. Not to mention Hardware Canucks went through their entire suite and showed the exact same thing. No loss.

However, that pales in comparison to AMD who have done a marvelous job improving GCN performance.
Upstairs Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
4770k Asus Maximus VI Hero evga 1080 Ti with Hybrid mod Corsair Vengeance Pro 2133 mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 500gb WD Caviar Black Corsair h100i GTX Windows 8.1 64bit 
MonitorPowerCase
xb280hk EVGA Supernova 1000 G2 Corsair Carbide Air 540 
  hide details  
Reply
Upstairs Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
4770k Asus Maximus VI Hero evga 1080 Ti with Hybrid mod Corsair Vengeance Pro 2133 mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 500gb WD Caviar Black Corsair h100i GTX Windows 8.1 64bit 
MonitorPowerCase
xb280hk EVGA Supernova 1000 G2 Corsair Carbide Air 540 
  hide details  
Reply
post #392 of 451
So, based on what I've been reading in this thread, even if I upgraded to a 6600k, which I plan on doing next month, it still wouldn't solve much of the CPU overhead or would it be good enough for the CPU overhead issue to not make too much of an impact?
post #393 of 451
Quote:
Originally Posted by PsYcHo29388 View Post

So, based on what I've been reading in this thread, even if I upgraded to a 6600k, which I plan on doing next month, it still wouldn't solve much of the CPU overhead or would it be good enough for the CPU overhead issue to not make too much of an impact?
Since the plot is Fall Out 4, let me share some insight from TechSpot:
The Machine
(14 items)
 
Nexus 7 2013
(11 items)
 
 
CPUMotherboardGraphicsRAM
A10 6800K Asus F2A85-V MSI 6870 Hawx, VTX3D 5770, AMD HD6950(RIP), Sap... G.skill Ripjaws PC12800 6-8-6-24 
Hard DriveOptical DriveOSMonitor
Seagate 7200.5 1TB NEC 3540 Dvd-Rom Windows 7 x32 Ultimate Samsung P2350 23" 1080p 
PowerCaseMouseAudio
Seasonic s12-600w CoolerMaster Centurion 5 Logitech G600 Auzen X-Fi Raider 
CPUMotherboardGraphicsRAM
Quad Krait 300 at 1.5Ghz Qualcomm APQ8064-1AA SOC Adreno 320 at 400mhz 2GB DDR3L-1600 
Hard DriveOSMonitorKeyboard
32GB Internal NAND Android 5.0 7" 1920X1200 103% sRGB & 572 cd/m2 LTPS IPS Microsoft Wedge Mobile Keyboard 
PowerAudio
3950mAh/15.01mAh Battery Stereo Speakers 
  hide details  
Reply
The Machine
(14 items)
 
Nexus 7 2013
(11 items)
 
 
CPUMotherboardGraphicsRAM
A10 6800K Asus F2A85-V MSI 6870 Hawx, VTX3D 5770, AMD HD6950(RIP), Sap... G.skill Ripjaws PC12800 6-8-6-24 
Hard DriveOptical DriveOSMonitor
Seagate 7200.5 1TB NEC 3540 Dvd-Rom Windows 7 x32 Ultimate Samsung P2350 23" 1080p 
PowerCaseMouseAudio
Seasonic s12-600w CoolerMaster Centurion 5 Logitech G600 Auzen X-Fi Raider 
CPUMotherboardGraphicsRAM
Quad Krait 300 at 1.5Ghz Qualcomm APQ8064-1AA SOC Adreno 320 at 400mhz 2GB DDR3L-1600 
Hard DriveOSMonitorKeyboard
32GB Internal NAND Android 5.0 7" 1920X1200 103% sRGB & 572 cd/m2 LTPS IPS Microsoft Wedge Mobile Keyboard 
PowerAudio
3950mAh/15.01mAh Battery Stereo Speakers 
  hide details  
Reply
post #394 of 451
yes you will fix alot of cpu overhead issues smile.gif singlethread performance is about 30-50% better clock for clock
but after DX12 arrive FX-8350 will keep with i5 as in multithread their performance is about same

with NVIDIA GPU difference between AMD cpus and Intel CPUs is much lower because of DX11 multithreaded driver
for AMD GPUS there is bigger difference between AMD/INTEL cpus as it have only single threaded DX11 driver
Edited by OneB1t - 2/3/16 at 4:30pm
FX
(7 items)
 
  
CPUMotherboardGraphicsRAM
FX-8320@4.4Ghz M5A99FX PRO R2.0 AMD Radeon R9 290X Patriot Memory  
Hard DriveCoolingMonitor
Samsung 840 Raijintek Ereboss iiyama X4071UHSU (4K) 
  hide details  
Reply
FX
(7 items)
 
  
CPUMotherboardGraphicsRAM
FX-8320@4.4Ghz M5A99FX PRO R2.0 AMD Radeon R9 290X Patriot Memory  
Hard DriveCoolingMonitor
Samsung 840 Raijintek Ereboss iiyama X4071UHSU (4K) 
  hide details  
Reply
post #395 of 451
Quote:
Originally Posted by mtcn77 View Post

Since the plot is Fall Out 4, let me share some insight from TechSpot:
Quote:
yes you will fix alot of cpu overhead issues smile.gif singlethread performance is about 30-50% better clock for clock
but after DX12 arrive FX-8350 will keep with i5 as in multithread their performance is about same

Thanks for the replies, if that techspot graph is correct then I should see crazy performance gains going to a 6600k + 3000Mhz RAM.

Also, I don't disagree that the FX-8350 is pretty close to the i5 when it comes to multi threaded stuff, but the issue is that there are practically no games that will use an 8350 to it's full potential, at least not from what I've seen. The only game I can think of is Battlefield 4 and even with that I am not 100% sure. If DX12 is support to fix that issue like I think you are implying, I'll be long gone from using the FX-8350. In fact, after doing a plethora of research, I don't think I will ever switch back to AMD for CPUs.
post #396 of 451
If Directx12 goes live you will have switched to i5 for naught. Apart from your current gpu making the change, not much of an upgrade, too. A single Fx8350(maybe an Fx6350 as well) is perfectly capable of fulfilling the drawcall budget of a single 290 in directx12 and in directx11, imho.
I think AMD knows enough about this to keep their sales pitch free of pretense.

You'll be having '40 FPS' instead of 20 in Total War, so that is a given, but hopefully better versions are to be expected in the future.
Edited by mtcn77 - 2/3/16 at 5:22pm
The Machine
(14 items)
 
Nexus 7 2013
(11 items)
 
 
CPUMotherboardGraphicsRAM
A10 6800K Asus F2A85-V MSI 6870 Hawx, VTX3D 5770, AMD HD6950(RIP), Sap... G.skill Ripjaws PC12800 6-8-6-24 
Hard DriveOptical DriveOSMonitor
Seagate 7200.5 1TB NEC 3540 Dvd-Rom Windows 7 x32 Ultimate Samsung P2350 23" 1080p 
PowerCaseMouseAudio
Seasonic s12-600w CoolerMaster Centurion 5 Logitech G600 Auzen X-Fi Raider 
CPUMotherboardGraphicsRAM
Quad Krait 300 at 1.5Ghz Qualcomm APQ8064-1AA SOC Adreno 320 at 400mhz 2GB DDR3L-1600 
Hard DriveOSMonitorKeyboard
32GB Internal NAND Android 5.0 7" 1920X1200 103% sRGB & 572 cd/m2 LTPS IPS Microsoft Wedge Mobile Keyboard 
PowerAudio
3950mAh/15.01mAh Battery Stereo Speakers 
  hide details  
Reply
The Machine
(14 items)
 
Nexus 7 2013
(11 items)
 
 
CPUMotherboardGraphicsRAM
A10 6800K Asus F2A85-V MSI 6870 Hawx, VTX3D 5770, AMD HD6950(RIP), Sap... G.skill Ripjaws PC12800 6-8-6-24 
Hard DriveOptical DriveOSMonitor
Seagate 7200.5 1TB NEC 3540 Dvd-Rom Windows 7 x32 Ultimate Samsung P2350 23" 1080p 
PowerCaseMouseAudio
Seasonic s12-600w CoolerMaster Centurion 5 Logitech G600 Auzen X-Fi Raider 
CPUMotherboardGraphicsRAM
Quad Krait 300 at 1.5Ghz Qualcomm APQ8064-1AA SOC Adreno 320 at 400mhz 2GB DDR3L-1600 
Hard DriveOSMonitorKeyboard
32GB Internal NAND Android 5.0 7" 1920X1200 103% sRGB & 572 cd/m2 LTPS IPS Microsoft Wedge Mobile Keyboard 
PowerAudio
3950mAh/15.01mAh Battery Stereo Speakers 
  hide details  
Reply
post #397 of 451
Quote:
Originally Posted by mcg75 View Post

It's actually not interesting at all. It's complete speculation based on data that has since been proven wrong.

The entire part where he speculates about why Fallout 4 "lost" performance because of what some random forum user stated has all turned out to be utter nonsense.

Besides the fact that no actual users reported any performance losses, we now have results from two sources that render the gamegpu results invalid.

The guy clearly says in the video that the results are from different areas in the game and can't be used for a apples to apples comparison. The performance increase from AMD is still pretty good overall.


Quote:
The sad part is we have people out there that take any shred of "evidence" and turn it into a story to garner page hits. It happens to both Nvidia and AMD.

And Kepler losing performance is also a myth as well. In Techpowerup's test suite, there are 4 games that were tested at 980 launch and are still tested today. 3 show Kepler performance being better and 1 shows a 1 fps loss. Not to mention Hardware Canucks went through their entire suite and showed the exact same thing. No loss.

However, that pales in comparison to AMD who have done a marvelous job improving GCN performance.

Not a myth. Once again I had two GTX 670s and I benchmarked the cards myself using newer and older drivers. I clearly saw my performance tank which is why I ran the benchmarks again personally. The so called Kepler fix did little to nothing in my case. I couldn't stay with the older drivers so I HAD to upgrade the drivers for the latest titles. Otherwise I'll be screwed for newer titles. The Nvidia drivers got so bad that my old overclocks wouldn't work and had artifacts on my screen. I thought my cards were crapping out, but after loading the older drivers......everything was 100% fine.

So I performed a apples to apples test using the same programs I had benched before. I was seeing loses in 3DMark FireStrike Performance and Ice Storm Performance and in games like Tomb Raider 2013. In one topic overclock.com mods ACTUALLY DELETED MY POST. Why? For telling the truth and showing my proof.

I can't speak for everyone except myself, but I know why I'm going to be careful the next time it's time to choose the green or red team. I saw the Maxwell prevail and my Keplers fail. I only heard good things from friends using there old HD 7970Ghz with the Omega drivers. Which is why it came down to a good decision between a aftermarket 980 Ti and Fury X. I decided to go with AMD this time around. People can take that however they want to take it. I ran my benchmarked and clearly saw the lost of performance with my very own Kepler cards.

I have no complaints about my Fury X and haven't even thought about overclocking it other than a few benchmarks in my review. Hopefully AMD continues the great work with the drivers overtime and my Fury X doesn't suffer the same fate as my dual GTX 670 SLI.

Tessellation has also been a touchy subject as well.
Edited by Kana-Maru - 2/3/16 at 6:29pm
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
post #398 of 451
Quote:
Originally Posted by Kana-Maru View Post

The guy clearly says in the video that the results are from different areas in the game and can't be used for a apples to apples comparison. The performance increase from AMD is still pretty good overall.

I have his gameplay videos posted in another thread. He tested the game 3 times. The first two were before the blast on the trail beside Vault 111. The third with the 20% performance reduction starts out on that same exact trail next to the vault and walks a little further past the vault and shoots his gun into the ground. The third clip being after the blast, has no foliage on the ground or trees in that area and should have better frame rates because the foliage in the first two creates more shadows. In that 3rd video as well, it shows weapon debris being on when he shoots the gun. If they did not turn debris off, that's your frame rate drop right there because it doesn't work on AMD cards.

Also, I'm sure not many people realize this but the 1.3 patch got an update a few days after realize because something Beth did caused shader problems and long load times. Update was 1.3.47

I'm not interested in playing the blame game here. Sometimes with these sites, things go wrong during testing. But it's very important to prove or disprove what they found because if it were true, I want to And all the evidence shows the results they had were not an accurate representation of the game performance.

AMD performance was improved for sure. And I'm sure the next driver will put it even closer. The Fury-X and 980 Ti should be pretty much neck and neck.
Quote:
Originally Posted by Kana-Maru View Post

Not a myth. Once again I had two GTX 670s and I benchmarked the cards myself using newer and older drivers. I clearly saw my performance tank which is why I ran the benchmarks again personally. The so called Kepler fix did little to nothing in my case. I couldn't stay with the older drivers so I HAD to upgrade the drivers for the latest titles. Otherwise I'll be screwed for newer titles. The Nvidia drivers got so bad that my old overclocks wouldn't work and had artifacts on my screen. I thought my cards were crapping out, but after loading the older drivers......everything was 100% fine.

So I performed a apples to apples test using the same programs I had benched before. I was seeing loses in 3DMark FireStrike Performance and Ice Storm Performance and in games like Tomb Raider 2013. In one topic overclock.com mods ACTUALLY DELETED MY POST. Why? For telling the truth and showing my proof.

I can't speak for everyone except myself, but I know why I'm going to be careful the next time it's time to choose the green or red team. I saw the Maxwell prevail and my Keplers fail. I only heard good things from friends using there old HD 7970Ghz with the Omega drivers. Which is why it came down to a good decision between a aftermarket 980 Ti and Fury X. I decided to go with AMD this time around. People can take that however they want to take it. I ran my benchmarked and clearly saw the lost of performance with my very own Kepler cards.

I have no complaints about my Fury X and haven't even thought about overclocking it other than a few benchmarks in my review. Hopefully AMD continues the great work with the drivers overtime and my Fury X doesn't suffer the same fate as my dual GTX 670 SLI.

Tessellation has also been a touchy subject as well.

I had 980 SLI and 780 Ti SLI at the same time. I remember the driver that lost major gpu points in Firestrike and there were several of us here doing tests. It didn't translate to games with the same effect.

Don't get me wrong, I'm not doubting your testing and what it showed at that time. But we have undeniable proof today from multiple published sources that shows the performance is still there.
Upstairs Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
4770k Asus Maximus VI Hero evga 1080 Ti with Hybrid mod Corsair Vengeance Pro 2133 mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 500gb WD Caviar Black Corsair h100i GTX Windows 8.1 64bit 
MonitorPowerCase
xb280hk EVGA Supernova 1000 G2 Corsair Carbide Air 540 
  hide details  
Reply
Upstairs Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
4770k Asus Maximus VI Hero evga 1080 Ti with Hybrid mod Corsair Vengeance Pro 2133 mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 EVO 500gb WD Caviar Black Corsair h100i GTX Windows 8.1 64bit 
MonitorPowerCase
xb280hk EVGA Supernova 1000 G2 Corsair Carbide Air 540 
  hide details  
Reply
post #399 of 451
This may not be 100% on topic because they are referring to an OpenGL game client but I thought it was ironic to see Jagex mention CPU overhead on the AMD side being a performance hurdle in their Runescape client....
Quote:
Originally Posted by Jagex View Post

For AMD CPU users, we have seen generally slower performance than Intel CPUs (which is to be expected from their specs), though for the most part it is not something that you need to be concerned about unless you are aiming for 60 fps on max settings. The impact is lessened if you have an Nvidia graphics card, which we have observed to have a lower CPU overhead than an equivalent AMD graphics card. We are continuing to work on optimising the performance on AMD CPUs and hopefully stable 60 fps on max settings on an AMD CPU + AMD GPU will become a reality.
post #400 of 451
Quote:
Originally Posted by Derp View Post

This may not be 100% on topic because they are referring to an OpenGL game client but I thought it was ironic to see Jagex mention CPU overhead on the AMD side being a performance hurdle in their Runescape client....

Good thing it is not C3 'cause i play that game with a phenom and a 7950.
Second Intel Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
2700 4.5/ 1.28 77 290 (2) 16 / 1866 
Hard DriveCoolingOSMonitor
1000 360/240 10 64 28 2160 
PowerCase
850 540 
  hide details  
Reply
Second Intel Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
2700 4.5/ 1.28 77 290 (2) 16 / 1866 
Hard DriveCoolingOSMonitor
1000 360/240 10 64 28 2160 
PowerCase
850 540 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Overclock.net › Forums › Graphics Cards › AMD/ATI › ATI Drivers and Overclocking Software › AMD GPU Drivers: The Real Truth.