• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Official Statement from AMD on the PCI-Express Overcurrent Issue

Joined
Oct 9, 2009
Messages
706 (0.18/day)
Location
Finland
System Name :P~
Processor Intel Core i7-5930K (ES)
Motherboard Asus Rampage V Extreme/3.1
Cooling Phanteks PH-TC14PE
Memory 32GB Corsair Vengeance LPX 2400 MHz
Video Card(s) Asus GTX 1080 Strix
Storage 400GB Intel 750 PCI-E SSD, 512GB Crucial MX100 SSD, 3TB WD RED HDD
Display(s) QNIX QX2710LED OC @ 96 Hz 27"
Case Corsair Obsidian 750D
Audio Device(s) Audioquest Dragon Red + Sennheiser HD 650
Power Supply Corsair HX1000i + Cablemod sleeved cables kit
Mouse Logitech G500s
Keyboard Logitech Ultra X Flat Premium
Software Windows 10 64-bit
Why is this an issue? Both the GTX 750 Ti and the GTX 950 both drew significantly more than 75W from the PCI bus! In short bursts, there is absolutely no problem with brief power spikes. Shitty deal if you have an older/cheap mobo and this is an issue; Seriously, quality motherboards are dirt cheap!

:toast:
None of those breaks the average spec of 75W (actually, for 12V its even less), while RX 480 breaks average spec by significant amount. Its not short bursts only. Huge difference.
 
Joined
Apr 30, 2011
Messages
1,648 (0.49/day)
Location
Greece
Processor AMD Ryzen 5 2600X@95W
Motherboard MSI B450 Tomahawk MAX
Cooling Deepcool Gammaxx 400 Black
Memory 2*8GB PATRIOT PVS416G373C7K@3333MT_C16
Video Card(s) Sapphire Radeon RX 5700 Pulse 8GB
Storage Sandisk SSD 120GB, INTEL 540S SSDSCKKW180H6 180GB, Samsung F1 1TB, Hitachi HUS724040ALE640 4TB
Display(s) AOC 27G2U/BK IPS 144Hz
Case SHARKOON M25-W 7.1 BLACK
Audio Device(s) Realtek 7.1 onboard
Power Supply Zalman Z550
Mouse Sharkoon SHARK Force Black
Keyboard Trust GXT280
Software Win 7 sp1 64bit/Win 10 pro 64bit
Benchmark Scores CB R15 64bit: single core 173p, multicore 1306p
Honestly I think Nvidia pushed AMD with their new gpu's.
The fact that you can barely OC the RX480 makes me believe AMD quickly issued as high a clock out of the box as possible to make the cards look good in the performance section, but that originally they were not meant to run this high.
I guess its best to get a custom designed RX480 from a partner that indeed has an 8pin connector and more cooling capability.
6pin was used just because AMD targets OEMs and they don't usually sell PCs with high end PSUs. So, they could put a cheap PSU into a PC with Polaris.

The custom ones don't need to take this into consideration so they will clock very high (rumors say about +1400 MHz which will reach or even surpass stock 980).

None of those breaks the average spec of 75W (actually, for 12V its even less), while RX 480 breaks average spec by significant amount. Its not short bursts only. Huge difference.
Driver fixable thing for those who don't know how to do it themselves, -10% in power settings for the current driver for those who want to do that now. And not any decrease in performance also. :toast:

As for the custom 480s, they will have alternative BIOS as usual and won't depend on default driver settings as usual.
 
Joined
Oct 9, 2009
Messages
706 (0.18/day)
Location
Finland
System Name :P~
Processor Intel Core i7-5930K (ES)
Motherboard Asus Rampage V Extreme/3.1
Cooling Phanteks PH-TC14PE
Memory 32GB Corsair Vengeance LPX 2400 MHz
Video Card(s) Asus GTX 1080 Strix
Storage 400GB Intel 750 PCI-E SSD, 512GB Crucial MX100 SSD, 3TB WD RED HDD
Display(s) QNIX QX2710LED OC @ 96 Hz 27"
Case Corsair Obsidian 750D
Audio Device(s) Audioquest Dragon Red + Sennheiser HD 650
Power Supply Corsair HX1000i + Cablemod sleeved cables kit
Mouse Logitech G500s
Keyboard Logitech Ultra X Flat Premium
Software Windows 10 64-bit
Driver fixable thing for those who don't know how to do it themselves, -10% in power settings for the current driver for those who want to do that now. And not any decrease in performance also. :toast:
You think they are wasting power for no reason? -10% will change performance, or otherwise AMD would have done it already.

There is a more technical solution, which does not change power draw or performance but that is still under investigation. Feasibility of that depends whether PCI-E bus and 6-pin are connected to parallel as input to all VREGs or they supply different parts of the VREGs.
 
Joined
Sep 2, 2014
Messages
153 (0.07/day)
Location
Emperor's retreat/Naboo Moenia
System Name Order66
Processor FX-8370
Motherboard ASUS CrosshairV Formula-Z (socketAM3+)
Cooling ArcticFreezer13 CO
Memory 16GB DDR3 1333MHz
Video Card(s) GeForce RTX2060
Storage WD 5003ABYX 500gb SataII
Display(s) Asus VE228HR
Case Corsair Carbide series SPEC-03
Audio Device(s) onboard Realtek
Power Supply EVGA Supernova G3 850
Software Windows10 64bit
Driver fixable thing for those who don't know how to do it themselves, -10% in power settings for the current driver for those who want to do that now. And not any decrease in performance also. :toast:
well , logic suggests that what you say can not be accurate. If there is no decrease when you reduce the power, then why not -20%, or even -50% then?
 

qubit

Overclocked quantum bit
Joined
Dec 6, 2007
Messages
15,925 (3.44/day)
Location
Quantum Well UK
System Name Quantumville™
Processor Intel Core i7-2700K at stock (hits 5 gees+ easily)
Motherboard Asus P8Z68-V PRO/GEN3
Cooling Noctua NH-D14
Memory 16GB (4 x 4GB Corsair Vengeance DDR3 PC3-12800 C9 1600MHz)
Video Card(s) Zotac GTX 1080 AMP! Extreme Edition
Storage Samsung 850 Pro 256GB | WD Green 4TB
Display(s) BenQ XL2720Z | Asus VG278HE (both 27", 144Hz, 3D Vision 2, 1080p)
Case Cooler Master HAF 922
Audio Device(s) Creative Sound Blaster X-Fi Fatal1ty PCIe
Power Supply Corsair HX 850W v1
Software Windows 10 Pro 64-bit
Reading AMD's statement cynically, one could say that they deliberately released the cards like this so they'd look good/better in the benchmarks, because once the power use is brought down the performance is gonna be significantly lower and make their value for money much lower. Be interesting to see exactly how much the performance hit will be.

@W1zzard are we gonna see a quick retest review with a handful of benchmarks with the revised driver to check this out?
 
Joined
Feb 18, 2013
Messages
1,864 (0.68/day)
Location
KL, Malaysia
System Name The Rainbow Puke Machine (Upgrading to Ryzen WIP)
Processor AMD Ryzen 7 3700X*
Motherboard GIGABYTE B550 AORUS PRO AC ATX*
Cooling AMD Wraith Prism w/RGB* + HD120 RGB (3) + SP120 RGB PRO (3) + Commander PRO
Memory Corsair Vengeance RGB PRO 16GB 3200MHz D4 (Black)*
Video Card(s) Zotac RTX2060 Twin Fan 6GB GDDR6 (1.9GHz OC stable via MSI Afterburner)
Storage Samsung 960 EVO 250GB M.2 PCIe SSD + Kingston UV400 480GB 2.5in SATA SSD
Display(s) LG 29WK600-W Ultrawide (21:9) 2560x1080 75Hz IPS w/AMD FreeSync
Case Corsair iCUE 220T RGB Airflow (White) w/Lighting Node CORE + Lighting Node PRO RGB LED Strips (4).
Audio Device(s) Onboard HD Audio (with OP-AMP)* + Corsair VOID Stereo Headphones
Power Supply Corsair RMx 750W 80 Plus Gold Fully Modular
Mouse Corsair M65 RGB FPS Mouse (White)
Keyboard Corsair Strafe RGB Mechanical Keyboard (Cherry MX RGB Red Switches)
Software Windows 10 Professional x64 (Update Build 2004)
If AIB vendors are fixing the "out of spec PCIe limits" with an 8-pin or possibly 8 + 6-pin to reduce excessive power draw from the PCIe slot, then AMD shouldn't even come up with a statement that it will release a driver fix. Limiting the card's power even by a little affects a lot of aspect. @HD64G why not u bench your VGA before limiting it & then u throw a 10% reduction to it via tuning software & then run the same test again? We wanna see if limiting power does not reduce the card's performance, as per what you claimed.
 
Joined
Oct 2, 2004
Messages
13,791 (2.38/day)
Reading AMD's statement cynically, one could say that they deliberately released the cards like this so they'd look good/better in the benchmarks, because once the power use is brought down the performance is gonna be significantly lower and make their value for money much lower. Be interesting to see exactly how much the performance hit will be.

@W1zzard are we gonna see a quick retest review with a handful of benchmarks with the revised driver to check this out?
Except people actually report higher performance when restricting its power...
 
Joined
Oct 9, 2009
Messages
706 (0.18/day)
Location
Finland
System Name :P~
Processor Intel Core i7-5930K (ES)
Motherboard Asus Rampage V Extreme/3.1
Cooling Phanteks PH-TC14PE
Memory 32GB Corsair Vengeance LPX 2400 MHz
Video Card(s) Asus GTX 1080 Strix
Storage 400GB Intel 750 PCI-E SSD, 512GB Crucial MX100 SSD, 3TB WD RED HDD
Display(s) QNIX QX2710LED OC @ 96 Hz 27"
Case Corsair Obsidian 750D
Audio Device(s) Audioquest Dragon Red + Sennheiser HD 650
Power Supply Corsair HX1000i + Cablemod sleeved cables kit
Mouse Logitech G500s
Keyboard Logitech Ultra X Flat Premium
Software Windows 10 64-bit
Except people actually report higher performance when restricting its power...
Reducing operating voltage MAY increase performance in thermally limited situations. Decreasing total power limit without changing anything is going to have opposite effect.
Reducing operating voltage MAY work on some GPUs, but due to obvious negative effect on stability (there is a reason why they put the VID it has now) it is too risky to do on all cards.
 

cadaveca

My name is Dave
Joined
Apr 10, 2006
Messages
17,104 (3.27/day)
I'm wondering why no one thought to consider that the BIOS might be the problem, and that AMD simply gave all cards a BIOS that allowed maximum power draw, over-stepping the driver-based tools to increase that. I found it interesting that power draw was high, and no OC was possible using the driver-based tools to give the GPU more power, and when you put those two together, and then consider the ASUS and MSI GPUs recently reviewed, you get a potential BIOS problem.

Perhaps AMD gave the card a BIOS that allowed it to exceed PCIe spec because it wanted it to be reviewed in the best light, and knew reviewers sometimes do not investigate OC? Given how their clocks are "managed" compared to NVidia's Turbo, this actually seems like the most reasonable explanation for what happened. Not every site has the capability to accurately measure power consumption for PCIe devices, so many sites wouldn't even be able to test such an issue.
 

Alex Rubio

New Member
Joined
Jul 2, 2016
Messages
2 (0.00/day)
Hahaha.... Dead on arrival... People have more and more reason to wait and buy the 1060 now...
Get outta here you fanboy!
As if the gtx 970 or 960 is clean. this is only when you over clock. Try to do some research before commenting you sound like a idiot.
This is just some redit fanboy making a big deal of nothing. I didn't hear a cry when Nvidia came out with the 960 or 970.
 

Alex Rubio

New Member
Joined
Jul 2, 2016
Messages
2 (0.00/day)
None of those breaks the average spec of 75W (actually, for 12V its even less), while RX 480 breaks average spec by significant amount. Its not short bursts only. Huge difference.
Gotta do a bit more research.
Here a video I found to be correct.
 

Maddox

New Member
Joined
Jan 1, 2016
Messages
5 (0.00/day)
Get outta here you fanboy!
As if the gtx 970 or 960 is clean. this is only when you over clock. Try to do some research before commenting you sound like a idiot.
This is just some redit fanboy making a big deal of nothing. I didn't hear a cry when Nvidia came out with the 960 or 970.
Well, for whatever it's worth, I have a habit of checking out user benchmark whenever new hardware is launched, just to check out the performance and popularity of CPU's and GPU's.

Around midnight of the launch day, going into July 30th, market share ranked 74th. Today it is tied with the R9 290 at 17th and obviously going up. I'm betting it'll -- in terms of sales -- trade blows with the GTX 1070 from here on out and probably slightly surpass it at some point once the 4GB versions hit the market.

Not exactly DOA by my reckoning.

I just think Nvidia fanboys (who are up there with Apple and Nintendo, imo) were spring loaded to fire at any sudden movement after the 970 fiasco.
 

qubit

Overclocked quantum bit
Joined
Dec 6, 2007
Messages
15,925 (3.44/day)
Location
Quantum Well UK
System Name Quantumville™
Processor Intel Core i7-2700K at stock (hits 5 gees+ easily)
Motherboard Asus P8Z68-V PRO/GEN3
Cooling Noctua NH-D14
Memory 16GB (4 x 4GB Corsair Vengeance DDR3 PC3-12800 C9 1600MHz)
Video Card(s) Zotac GTX 1080 AMP! Extreme Edition
Storage Samsung 850 Pro 256GB | WD Green 4TB
Display(s) BenQ XL2720Z | Asus VG278HE (both 27", 144Hz, 3D Vision 2, 1080p)
Case Cooler Master HAF 922
Audio Device(s) Creative Sound Blaster X-Fi Fatal1ty PCIe
Power Supply Corsair HX 850W v1
Software Windows 10 Pro 64-bit
Except people actually report higher performance when restricting its power...
That makes no sense. Perhaps there's something else going on here? Without details one can't say what the true situation is, but simply lowering power consumption isn't gonna increase performance.

Reducing operating voltage MAY increase performance in thermally limited situations. Decreasing total power limit without changing anything is going to have opposite effect.
Reducing operating voltage MAY work on some GPUs, but due to obvious negative effect on stability (there is a reason why they put the VID it has now) it is too risky to do on all cards.
This sounds more plausible.
 
Joined
Oct 2, 2004
Messages
13,791 (2.38/day)
If it's thermal throttling, it can throttle more than it does if it's power limited... this one would probably depend on quality of case cooling...
 

Daiwa Zou

New Member
Joined
Jul 2, 2016
Messages
1 (0.00/day)
Lol. "RIP AMD" and here you are referencing one of the most biased source on seekingalpha.com to prove your point. Not sure if you know Hibben is known for his anti-AMD mentality.

This says a lot about how biased you are as well.
 
Joined
Sep 25, 2007
Messages
5,844 (1.24/day)
Location
New York
Processor Core I7 3770K@4.8Ghz
Motherboard AsRock Z77 Extreme
Cooling Cooler Master Seidon 120M
Memory 12Gb G.Skill Sniper
Video Card(s) EVGA GTX 1070 FTW 2150/2240
Storage Sandisk SSD + 1TB Seagate Barracuda 7200
Display(s) IPS Asus 26inch
Case Antec 300
Audio Device(s) Xonar DG
Power Supply EVGA Supernova 650 G2
Software Windows 10/Windows 7
I've heard cases where the cards are power throttling on the factory overclock, some are doing it on stock also, so increasing the power limit increases the cards performance because its drawing too much power. Some other lower binned ones are running hotter and requiring undervolt or the power limit to be dropped to reduce temps and that increases performance since it won't thermal throttle down the boost clock, its pretty confusing really.
 
Last edited:
Joined
Feb 14, 2012
Messages
2,023 (0.65/day)
System Name msdos
Processor 8086
Motherboard mainboard
Cooling passive
Memory 640KB + 384KB extended
Video Card(s) EGA
Storage 5.25"
Display(s) 80x25
Case plastic
Audio Device(s) modchip
Power Supply 45 watts
Mouse serial
Keyboard yes
Software disk commander
Benchmark Scores still running
If you have a junky motherboard, maybe, but I doubt any mainstream brands don't build in A LITTLE reserve.
I think it has to do more with the connector limitation of the little pins in the slot.
 
Joined
Nov 20, 2013
Messages
4,336 (1.77/day)
Location
Kiev, Ukraine
System Name WS#1337
Processor Ryzen 5 1600X
Motherboard Gigabyte x470 AORUS Ultra Gamin
Cooling Xigmatek Scylla 240 AIO
Memory 2x8GB Team T-Force Vulkan DDR4-3000
Video Card(s) MSI RTX 2060 Super Armor OC
Storage Adata SX8200 Pro 1TB
Display(s) Samsung U24E590D (4K/UHD)
Case Chieftec AL-01B-OP
Audio Device(s) ALC1220
Power Supply SeaSonic SSR-550FX (80+ GOLD)
Mouse Logitech G603
Keyboard Zalman K500 modded (Gateron brown)
Software Windows 10, Ubuntu 18.04 LTS
Why is this an issue? Both the GTX 750 Ti and the GTX 950 both drew significantly more than 75W from the PCI bus! In short bursts, there is absolutely no problem with brief power spikes. Shitty deal if you have an older/cheap mobo and this is an issue; Seriously, quality motherboards are dirt cheap!

:toast:
When did that happen? My MSI GTX750Ti overclocked to 1300 / 1500 never exceeded 65W mark during stress testing (according to sensor readouts). Under normal conditions it stays below 58W.
GTX950 is a 90W card, so under normal circumstances, the only theoretical way it can overdraw power from PCI-E only if it draws no power from 6-pin connector at all (or if you own one of those newer bus-powered cards from ASUS or EVGA).

When it comes to motherboards, you'll be surprised how many shitty products hit the market nowadays. Just because it's high-end does not mean that it won't break.
 

newtekie1

Semi-Retired Folder
Joined
Nov 22, 2005
Messages
27,197 (5.06/day)
Location
Indiana, USA
Processor Intel Core i7 9900K@5.0GHz
Motherboard AsRock Z370 Taichi
Cooling Corsair H115i Pro w/ Noctua NF-A14 Fans
Memory 32GB Corsair DDR4-3000
Video Card(s) ASUS Strix GTX 1080Ti
Storage 500GB SX8200 Pro + 8TB with 1TB SSD Cache
Display(s) QNIX QX2710 1440p@120Hz
Case Fractal Design Define S
Audio Device(s) Onboard is good enough for me
Power Supply eVGA SuperNOVA 1000w G3
Software Windows 10 Pro x64
ven the aforementioned Diablotek could handle powering one of these, paired with a latest-gen Skylake CPU, a couple sticks of RAM and some storage. It would all easily fit into a 250W envelope (absolute peak power draw, realistically less than that), which even the worst of the worst PSUs can manage, at least for a while.

That being said, everyone should have the common sense not to skimp on the PSU. No need to go crazy, a nice $30-or-so PSU from a reputable manufacturer should do fine, as @newtekie1 pointed out.
Sure, but like you pointed out, why skimp on the PSU. And if you have a latest-gen skylake, or even a last gen Haswell, or even an Ivy-Bridge, and your PSU doesn't have an 8-pin connector, seriously go buy a new one before upgrading your graphics card!
 
Joined
Aug 20, 2007
Messages
13,227 (2.79/day)
System Name Pioneer
Processor Intel i9 9900k
Motherboard ASRock Z390 Taichi
Cooling Noctua NH-D15 + A whole lotta Sunon and Corsair Maglev blower fans...
Memory G.SKILL TridentZ Series 32GB (4 x 8GB) DDR4-3200 @ 14-14-14-34-2T
Video Card(s) AMD RX 5700 XT (XFX THICC Ultra III)
Storage Mushkin Pilot-E 2TB NVMe SSD w/ EKWB M.2 Heatsink
Display(s) 55" LG 55" B9 OLED 4K Display
Case Thermaltake Core X31
Audio Device(s) VGA HDMI->Panasonic SC-HTB20/Schiit Modi MB/Asgard 2 DAC/Amp to AKG Pro K7712 Headphones
Power Supply SeaSonic Prime 750W 80Plus Titanium
Mouse ROCCAT Kone EMP
Keyboard WASD CODE 104-Key w/ Cherry MX Green Keyswitches, Doubleshot Vortex PBT White Transluscent Keycaps
Software Windows 10 Enterprise (yes, it's legit.)
I think it has to do more with the connector limitation of the little pins in the slot.
No, that can't be it because otherwise connectors to boost the amperes to the slot (as my board has) would make little sense.
 
Joined
Apr 29, 2014
Messages
3,850 (1.68/day)
Location
Texas
System Name SnowFire / The Reinforcer / Portable? / New Portable?
Processor i7 8086K 5.0ghz (24/7) / 2x Xeon E52650v2 / i7 7700HQ / AMD Ryzen 5 3600X
Motherboard Asus Z370 Prime-A / Dell Dual Socket (R720) / HP Omen Stock / Asrock X570 ITX
Cooling RX 360mm + 140mm Custom Loop / Dell Stock / HP Stock / Noctua L9i (Yes L9i)
Memory Corsair RGB 16gb DDR4 3000 CL 16 / DDR3 128gb 16 x 8gb / 12gb DDR4 3 x 4gb / Corsair RGB 3200 16gb
Video Card(s) GTX Titan XP (2025mhz) / Asus GTX 950 (No Power Connector) / RX 580 / GTX 970 (Temp)
Storage Samsung 970 1tb NVME and 2tb HDD x4 RAID 5 / 300gb x8 RAID 5 / NVME 240gb SSD and 1tb HDD
Display(s) Acer XG270HU 1440p 144hz Freesync, Acer B286HK 4K UHD Monitor / 1080p 60hz Freesync display
Case Corsair Obsidian 800D / Dell Poweredge R720 Rack Mount Case / HP Omen 17 / Fractal Design Node 202
Audio Device(s) Realtec ALC1150 (On board)
Power Supply Rosewill Lightning 1300Watt / Dell Stock 750 / Brick / Fractal Design 450 Watt Bronze
Mouse Logitech G5
Keyboard Logitech G19S
Software Windows 10 Pro / Windows Server 2008 R2 / Windows 10 Home / Windows 10 Pro
Sure, but like you pointed out, why skimp on the PSU. And if you have a latest-gen skylake, or even a last gen Haswell, or even an Ivy-Bridge, and your PSU doesn't have an 8-pin connector, seriously go buy a new one before upgrading your graphics card!
Yea, I mean even the most entry level PSU is $50 bucks comes with both an 8 and 6 pin (EVGA) so I don't see the logic in not having it and instead opting for the 6 pin.

When did that happen? My MSI GTX750Ti overclocked to 1300 / 1500 never exceeded 65W mark during stress testing (according to sensor readouts). Under normal conditions it stays below 58W.
GTX950 is a 90W card, so under normal circumstances, the only theoretical way it can overdraw power from PCI-E only if it draws no power from 6-pin connector at all (or if you own one of those newer bus-powered cards from ASUS or EVGA).

When it comes to motherboards, you'll be surprised how many shitty products hit the market nowadays. Just because it's high-end does not mean that it won't break.
Never checked my 950, but its the 0 power connector from Asus. Would be curious how much it draws under load.
 

newtekie1

Semi-Retired Folder
Joined
Nov 22, 2005
Messages
27,197 (5.06/day)
Location
Indiana, USA
Processor Intel Core i7 9900K@5.0GHz
Motherboard AsRock Z370 Taichi
Cooling Corsair H115i Pro w/ Noctua NF-A14 Fans
Memory 32GB Corsair DDR4-3000
Video Card(s) ASUS Strix GTX 1080Ti
Storage 500GB SX8200 Pro + 8TB with 1TB SSD Cache
Display(s) QNIX QX2710 1440p@120Hz
Case Fractal Design Define S
Audio Device(s) Onboard is good enough for me
Power Supply eVGA SuperNOVA 1000w G3
Software Windows 10 Pro x64
Yea, I mean even the most entry level PSU is $50 bucks comes with both an 8 and 6 pin (EVGA) so I don't see the logic in not having it and instead opting for the 6 pin.

Never checked my 950, but its the 0 power connector from Asus. Would be curious how much it draws under load.

Even the $30 430w eVGA has an 8-pin.

And the ASUS gtx950 with no power connector pulls a maximum of 76w from the slot. So they managed to keep it right at the limit.

 
Top