Overclock.net › Forums › Industry News › Hardware News › [Anand] AMD Comments on GPU Stuttering, Offers Driver Roadmap & Perspective on Benchmarking
New Posts  All Forums:Forum Nav:

[Anand] AMD Comments on GPU Stuttering, Offers Driver Roadmap & Perspective on Benchmarking - Page 8

post #71 of 152
Quote:
Originally Posted by RKTGX95 View Post

well, thanks for the explanation. (everyday you learn something new. at a second read i kinda realized it does sound a bit silly but i already knew before it that it might be rubbish smile.gif )

but still i remember that there is still a problem with NV drivers when installing AMD ones. (as you have seen i'm not the best one at this but IIRC some registry files or something)

Think of that in this way, once the code is in the kernel it has access to the entire system without exploits or any manner of issues. So if Nvidia writes their code to break, or refuse to run some functions ( could get into more detail but it's not really required), it will do so without fault. If they want to find out if AMD has loaded a driver, because they don't want users using their cards as 2nd hand features, they can. It has essentially full control, it's why you see AV software now installing drivers. I don't remember if Steam installs a driver, last had checked it didn't. Some people had made VAC hacks as drivers (ring0 think of it as a rootkit and vice versa), as to be able to read memory without Steam being able to scan for it. =O Dirty little ideas eh? That's what you do when you install a driver, you give it full control. Hence why many people say not to install drivers unless it doesn't work as intended, you give up a huge security risk if you use an unknown source. Now does the MS push for signed drivers sound normal? Having a signed kernel from boot, as to keep any malicious code from being inserted before the OS has set up security features. Then only passing that control onto code that is signed itself, as to ensure their isn't a break in command.

Anyways, I hope that clears the topic up a little. Outside of that, their are more details as to what a driver can and cannot do, for the typical user that isn't really needed.
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
post #72 of 152
Quote:
Originally Posted by mushroomboy View Post

Think of that in this way, once the code is in the kernel it has access to the entire system without exploits or any manner of issues. So if Nvidia writes their code to break, or refuse to run some functions ( could get into more detail but it's not really required), it will do so without fault. If they want to find out if AMD has loaded a driver, because they don't want users using their cards as 2nd hand features, they can. It has essentially full control, it's why you see AV software now installing drivers. I don't remember if Steam installs a driver, last had checked it didn't. Some people had made VAC hacks as drivers (ring0 think of it as a rootkit and vice versa), as to be able to read memory without Steam being able to scan for it. =O Dirty little ideas eh? That's what you do when you install a driver, you give it full control. Hence why many people say not to install drivers unless it doesn't work as intended, you give up a huge security risk if you use an unknown source. Now does the MS push for signed drivers sound normal? Having a signed kernel from boot, as to keep any malicious code from being inserted before the OS has set up security features. Then only passing that control onto code that is signed itself, as to ensure their isn't a break in command.

Anyways, I hope that clears the topic up a little. Outside of that, their are more details as to what a driver can and cannot do, for the typical user that isn't really needed.

thanks for the further explanation thumb.gif

but hypothetically, a badly written driver may cause stuttering (or other problems) right?
"Black & Red"
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5-3570k ASUS Maximus V Formula Sapphire HD 7950 OC 950Mhz Edition G.Skill 16GB (2x8G) 1333Mhz 9-9-9-24 (F3-10666C... 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB SSD Western Digital Caviar Blue 1TB ASUS DRW-24B3ST DVD Burner Antec Kuhler 920 
PowerCase
Corsair AX 750 Corsair C70 
  hide details  
Reply
"Black & Red"
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5-3570k ASUS Maximus V Formula Sapphire HD 7950 OC 950Mhz Edition G.Skill 16GB (2x8G) 1333Mhz 9-9-9-24 (F3-10666C... 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB SSD Western Digital Caviar Blue 1TB ASUS DRW-24B3ST DVD Burner Antec Kuhler 920 
PowerCase
Corsair AX 750 Corsair C70 
  hide details  
Reply
post #73 of 152
Well, the article and this thread has been enlightening. It makes it easy to tell who are the fanboys and who actually reads the article.

It's good to understand why it's happening, and sadly it sounds like it's a problem with the PC gaming platform as a whole. From the sounds of it, it looks like the only way to really fix it is to throw hacks together for separate games in drivers. If you leave Windows in charge all hell breaks loose. A lot of why I'm thinking that this is the problem instead of the driver being completely broken is that Nvidia has historically been much closer with game developers and that if the driver was the problem, we wouldn't see individual games seeing increases, but all DX9/DX10/DX11 titles fixed when the driver was fixed. I'm guessing Nvidia has been patching in stuff for their driver for a while, and AMD just assumed NVidia was having the same problems because it was Windows and the game dev's fault, when in fact there are things the driver teams can do to mitigate the problems.
Tyrant Kuma
(13 items)
 
Starscythe
(13 items)
 
Mobility
(6 items)
 
CPUCPUCPUCPU
Opteron 8431 Opteron 8431 Opteron 8431 Opteron 8431 
MotherboardGraphicsRAMHard Drive
Super Micro h8qme-2+ Sapphire 4870 Samsung DDR2 ECC 667mhz western digital caviar blue 
CoolingCoolingOSPower
FX 8350 stock cooler FX 6300 stock cooler Gentoo Linux 2x Silverstone PSU 
Case
Custom Fabbed Steel Case 
CPUMotherboardGraphicsRAM
AMD FX 8350 gigabyte 990FXA-UD5 7970 Mushkin Enhanced Blackline 16GB 
OSOSMonitorKeyboard
Gentoo Linux Windows 7 x64 Yamakasi Catleap Q270 Corsair K90 
PowerCaseMouseAudio
Silverstone Strider Gold Evolution 850W 80 PLUS... Antec 1200 Logitech G9x Asus Xonar D2X 
Other
XSPC RS360 Raystorm with custom radiator mounts 
CPUGraphicsRAMOS
a4 5000 Radeon HD 8330 8GB G. Skill DDR3 1600 9-9-9 Gentoo Linux 
OSOther
Windows 7 Lenovo Thinkpad x140e 
  hide details  
Reply
Tyrant Kuma
(13 items)
 
Starscythe
(13 items)
 
Mobility
(6 items)
 
CPUCPUCPUCPU
Opteron 8431 Opteron 8431 Opteron 8431 Opteron 8431 
MotherboardGraphicsRAMHard Drive
Super Micro h8qme-2+ Sapphire 4870 Samsung DDR2 ECC 667mhz western digital caviar blue 
CoolingCoolingOSPower
FX 8350 stock cooler FX 6300 stock cooler Gentoo Linux 2x Silverstone PSU 
Case
Custom Fabbed Steel Case 
CPUMotherboardGraphicsRAM
AMD FX 8350 gigabyte 990FXA-UD5 7970 Mushkin Enhanced Blackline 16GB 
OSOSMonitorKeyboard
Gentoo Linux Windows 7 x64 Yamakasi Catleap Q270 Corsair K90 
PowerCaseMouseAudio
Silverstone Strider Gold Evolution 850W 80 PLUS... Antec 1200 Logitech G9x Asus Xonar D2X 
Other
XSPC RS360 Raystorm with custom radiator mounts 
CPUGraphicsRAMOS
a4 5000 Radeon HD 8330 8GB G. Skill DDR3 1600 9-9-9 Gentoo Linux 
OSOther
Windows 7 Lenovo Thinkpad x140e 
  hide details  
Reply
post #74 of 152
Quote:
Originally Posted by RKTGX95 View Post

thanks for the further explanation thumb.gif

but hypothetically, a badly written driver may cause stuttering (or other problems) right?

Poorly written drivers can cause ALL sorts of issues. How did Nvidia squeeze up to 40% on the new Tomb Raider game for some cards? The hardware can only do what the software tells it, if the software fails to do that properly.... Can a drunk person operate machinery as good as a sober person? Excluding the functional alcoholic and those that have state learned muscle memory, no they cannot. State learned muscle memory, think of the drunk who plays pool better drunk than sober. The skateboarder who skates high but sucks sober, that type of thing.

So yeah, poor drivers can cause any sorts of issues. BSOD/Lockup/Performance problems, it goes on and on. Hence why overclocking, especially on the memory, can be detrimental. Even the card's firmware can cause issues, improper timings can do the same things as overclocked memory. It causes corruption, which can lead to very bad things. It's a machine, the machine has to have all it's parts run smoothly. The more complex the machine, the more delicate the parts. If those parts start to act up, they affect the performance of other parts. Software plays a role in that, majorly.
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
post #75 of 152
Dead Ryzen
(12 items)
 
Penguin Box
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD RYZEN 7 1800X  MSI B350M MORTAR ARCTIC AMD Radeon Vega Frontier Edition CORSAIR Vengeance LPX 32GB (4 x 8GB - 2133 MHz) 
Hard DriveHard DriveCoolingMonitor
SAMSUNG SSD 950 PRO 512GB Crucial MX300 2TB Noctua NH-U12S  BenQ XL2730Z - 27” 2560x1440 
KeyboardPowerCaseMouse
Das Keyboard 4 Professional Soft Tactile (MX Br... Seasonic X-850(SS-850KM3 Active PFC F3) Fractal Design Define Mini C BenQ ZOWIE EC1-A 
CPUMotherboardGraphicsRAM
Intel Core i7-2600K @4.6 GHz P8Z68-V AMD Radeon RX 480 (ref. design) Mushkin Enhanced Stealth 8GB (2 x 4GB) (PC3 17000) 
Hard DriveCoolingOSPower
SanDisk Extreme PRO 480GB Noctua NH-U12 S Arch Linux Seasonic SS-520FL 520W 
Case
Corsair Vengeance Series C70 
  hide details  
Reply
Dead Ryzen
(12 items)
 
Penguin Box
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD RYZEN 7 1800X  MSI B350M MORTAR ARCTIC AMD Radeon Vega Frontier Edition CORSAIR Vengeance LPX 32GB (4 x 8GB - 2133 MHz) 
Hard DriveHard DriveCoolingMonitor
SAMSUNG SSD 950 PRO 512GB Crucial MX300 2TB Noctua NH-U12S  BenQ XL2730Z - 27” 2560x1440 
KeyboardPowerCaseMouse
Das Keyboard 4 Professional Soft Tactile (MX Br... Seasonic X-850(SS-850KM3 Active PFC F3) Fractal Design Define Mini C BenQ ZOWIE EC1-A 
CPUMotherboardGraphicsRAM
Intel Core i7-2600K @4.6 GHz P8Z68-V AMD Radeon RX 480 (ref. design) Mushkin Enhanced Stealth 8GB (2 x 4GB) (PC3 17000) 
Hard DriveCoolingOSPower
SanDisk Extreme PRO 480GB Noctua NH-U12 S Arch Linux Seasonic SS-520FL 520W 
Case
Corsair Vengeance Series C70 
  hide details  
Reply
post #76 of 152
Even with some actual data and some well thought out paragraphs...the fanboys continue to shine.
post #77 of 152
Quote:
Originally Posted by Evil Penguin View Post

Then there's this:
http://blogs.valvesoftware.com/linux/faster-zombies/

Man, that fits the name too. wink.gif I kid! it's a good read though, I'm going through the comments. You see a lot of anti-AMD driver related posts, mainly talking about how performance is currently lacking and it could be better. I fully agree, as well as with the notion that they could be better. I don't doubt these cards have power, the issue is the company behind them. Nvidia has just done better in that aspect, it's changing but will still take time. When AMD first bought out ATi, I said it'll take quite a few years for them to catch up in driver support. Looks like it'll be a bit longer than my estimate, though they are doing better than ATi was doing alone, much better.

[edit] Just found a good quote on the importance of latency.
Code:
As Wolfos and others mention, differencing FPS numbers is meaningless, whereas differencing ms is a great way to measure the cost or improvement of an 
algorithm.

Percentage/ratio comparisons are “correct”, but misleading. If a game is running at 5000fps for instance and I can get it to 10000fps, that’s a 2x increase, so a big 
deal right? In reality I’ve just shaved off 0.1ms… once the game is running at a more reasonable rate, that’s just noise. At 60fps that optimization doesn’t even get me 
to 61fps…

Measuring with time/milliseconds is just the most useful and least misleading way to report performance.

Now take this as YOU will, I take this as AMD has work on their hands. Again, no doubt their hardware (with the right implementation) can go above Nvidia. Their latency issues are a sign of poor optimization. Makes me wonder how much better the AMD hardware is, if it's that good it also makes me wonder how little Nvidia needs to change to keep up. Interesting, I like that explanation.
Edited by mushroomboy - 3/26/13 at 5:32pm
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
Current Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 4.6GHz@1.44v GA-990FXA-UD3 R4.0 HD 7950 (1100/1450) 8G Muskin DDR3 1866@8CLS 
Hard DriveOptical DriveOSMonitor
1TB WD LiteOn DVD-RW DL Linux/Windows 19" Phillips TV 1080p 
PowerCaseMouseMouse Pad
OCZ 600W Generic Junk Logitech MX400 Generic Junk 
Audio
SBL 5.1 
  hide details  
Reply
post #78 of 152
I haven't seen or heard of it, but it would be interesting if they tested Intel's GPUs.
post #79 of 152
Quote:
Originally Posted by mcg75 View Post

Great to see AMD taking the frametime issue so seriously.

Especially when we have a few on OCN that claim it either doesn't exist or isn't important. rolleyes.gif

LOL, ya that never happens. wink.gif
Zybane
(20 items)
 
  
CPUMotherboardGraphicsRAM
i7 6950X - 4.5 GHz Edition 10 Rampage V Extreme Titan Xp  G.Skill 32GB 3200 MHz CL14 
Hard DriveOSMonitorKeyboard
1TB - Samsung SM961 Windows 10 Pro OLED 4K 55" LG C7P Razer BlackWidow TE Stealth Chroma 
PowerMouseAudioAudio
Corsair AX1500i 1500W Titanium Razer Mamba TE Denon AVR-X7200WA Elac BS 312 
AudioAudioAudioOther
Sennheiser HD-700 Headphones Velodyne MiniVee Subwoofer Buttkicker LFE Razer Orbweaver Stealth Chroma 
  hide details  
Reply
Zybane
(20 items)
 
  
CPUMotherboardGraphicsRAM
i7 6950X - 4.5 GHz Edition 10 Rampage V Extreme Titan Xp  G.Skill 32GB 3200 MHz CL14 
Hard DriveOSMonitorKeyboard
1TB - Samsung SM961 Windows 10 Pro OLED 4K 55" LG C7P Razer BlackWidow TE Stealth Chroma 
PowerMouseAudioAudio
Corsair AX1500i 1500W Titanium Razer Mamba TE Denon AVR-X7200WA Elac BS 312 
AudioAudioAudioOther
Sennheiser HD-700 Headphones Velodyne MiniVee Subwoofer Buttkicker LFE Razer Orbweaver Stealth Chroma 
  hide details  
Reply
post #80 of 152
We gamers are all doomed once WW4 starts and our rigs breakdown, back to playing in the sandbox
rig
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7-2700k @ 4.5GHz ASRock z75 Pro3 MSI 980 Ti Gaming and Gigabyte R9 290 OC windforce 8GB 2133 CL9 Gskill Snipper 
Hard DriveHard DriveHard DriveCooling
2x 128GB M4 sata3 3TB Seagate + 640GB WD black M550 512GB sata3 2x140mm SilverArrow 
OSMonitorPowerCase
Win7 64bit QNIX 27" 1440p @96Hz 850W EVGA SuperNova +80Gold HAF 932 
MouseAudioAudio
Razor Abyssus SoundBlaster Z  Logitech X-540 5.1 
  hide details  
Reply
rig
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7-2700k @ 4.5GHz ASRock z75 Pro3 MSI 980 Ti Gaming and Gigabyte R9 290 OC windforce 8GB 2133 CL9 Gskill Snipper 
Hard DriveHard DriveHard DriveCooling
2x 128GB M4 sata3 3TB Seagate + 640GB WD black M550 512GB sata3 2x140mm SilverArrow 
OSMonitorPowerCase
Win7 64bit QNIX 27" 1440p @96Hz 850W EVGA SuperNova +80Gold HAF 932 
MouseAudioAudio
Razor Abyssus SoundBlaster Z  Logitech X-540 5.1 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Anand] AMD Comments on GPU Stuttering, Offers Driver Roadmap & Perspective on Benchmarking