Discussion in 'Videocards - AMD Radeon Drivers Section' started by shadow_craft, Dec 9, 2014.
yeah chrome already uses html5 player by default . its still happening randomly tho
I know but no 1080p in HTML5 mode, at least not in firefox and palemoon.
I just spent a week working out my pc, and got a bit excited about my z68, as I got it working perfectly, and was using it as an example incase it help someone else.
Anyway I suppose what I mean is do you have everything installed and flashed from the MSI support page. They mention a sleep fix for your board.
Up until a few years ago sleep was a hit and miss feature of a lot of boards.
It seems a bit daunting, but swapping the PSU for a mates or an old spare only takes half an hour.
Mate my board is AMD and gx-x has Ga not MSI. But thanks anyway.
We already know where the bug is because it's been fixed an fuked by AMD many times. Everyone who has a GCN AMD card knows.
Win raid and the guys on it are very very well known in the advanced user community, Fernando was modding nforce drivers 10 years ago to work on nF3 since Nvidia couldn't be assed.
It's a forum for enthusiasts who know the risks, not for newbies to go and brick their boards then blame the modders. If you want safe and whql go to MSI.
The other member from Guru3D who has WinRaid account is a programmer I believe. As an Engineer I've asked his advice several times, he even wrote some script for me once.
Usually we know what we're doing, but when we don't...well it's an acceptable risk.
I really appreciate the thought of helping and please don't take offense.....just understand Fernando at WinRaid is a good guy.
I understand it is easy to lose sight of, however, I was originally talking to N0bodyOfTheGoat who has the MSI.
Gigabyte explains the other guys problem, I had a Gigabyte in the 90's for a day until I realized that they had left out the 16bit DMA lines to save money, never again. Not many people would notice that in one day, lol. I had an AMD board for a few years and that was a pretty sad time too, I did it when everyone was buying intel, just to be different, but they never got the chipset driver right.
RE win-raid: To give him the benefit of the doubt I have a PCIe2 MB and ASUS have used parts of PCIE2 CPU and PCIE-3 CPU code.
Like I said look at Intel's "FW Bring Up Guide" or "System Tools User Guide" and you will see why it scares me. Also, look for all the people on the internet who have screwed up their BIOS following his advice, there are tons. It is full of sweeping statements, but I tested his advice...
, actually I tested every safe combo, and the event viewer usually filled up with rather scary messages, I doubt people look. Some of his "driver only" packages are missing drivers, most of the full packages are for intel boards with all the features turned on. MEInfo always worked, but MEManuf was usually screwy.
Dude you're not making sense. For starters 16bit support was dropped back when Moses was a baby.
Your making this complicated for yourself because you're all over the place and can't pinpoint the source of the bug.
His PnP monitor has to be redetected. That's all.
You've listed 16bit support, PCI-e bus code?, drivers, BIOS, , corrupt SMBus, and bunch other stuff I didn't read.
So that covers pretty much everything on his system except the parts that are actually affected.
And then u suggested I read Intel system tools. OK. I think Winternals 7 has the OS pretty well covered.
3 things are involved, the GPU, the monitor, and the display driver. That's it.
Now you've got all the other parts out of the way, your job of pinpointing the problem woud much much easier...
Anyway long story short, it's the drivers. So we can move on now to a new topic .
I did say the 90's, I am rambling because I am bored. I don't care about this bug as I have never seen it. When I googled AMD sleep bug, I see it someone fixed his by changing the wake from sleep options in the BIOS, I use the PS/2 space bar and USB mouse.
My problem with sleep was the PSU. You see the PSU controls the power to everything and if it lags some of the rails then it won't work well. The MB chipset... it controls the PCIe bus power. The thing you can solve with a PSU swap is..."Does my MB get power before the card?" i.e. the bus gets power, turns the card on, the card does not get extra power from the PSU in time so does not turn on the monitor... I hope that makes sense.
No, reading "System Tools User Guide", "System Tools for. Intel® 7 Series/C216 Chipset" is a smart thing to do if you every consider flashing a random MEI module into your BIOS. Pretty simple. For some people this may mess their sleep up I suppose, but I was just rambling as you all have other systems.
It works in FF Beta 36. 1080p with 60 FPS
Whatever the deal is with that monitor suspend/sleep bug; it only involves Windows in my experience.
I've seen it happen on a 800-series AMD chipset motherboard, a 990FX board, and a H97 Intel board. Had two different power supplies involved, and about 4 Radeon HD 7850 cards involved (all from MSI though; 2 PEs, 2 OCs). I can't recall there being a time where monitor suspend has worked for extended periods without it failing on Windows (7, 8, 8.1, and 10 TP as of 9926; and it even still occurs with this Omega driver that was said to fix this issue).
I disable all power-saving tech for the most part (CnQ, Intel's RapidStart thing, C1/6 states, etc). Tried different PCI-E ports on my 990FX board. Tried some Acer S201HL monitors (single with both DVI and HDMI, triple in multimonitor, and triple in Eyefinity; tested with DDC/CI enabled and disabled), and a Gateway VX1120; had the sleep issue occur on all of them.
Meanwhile in Linux, using either the open-source radeon driver or fglrx (AMD proprietary; various versions tested on two different 7850 cards); I have yet to experience the sleep issue, even through various kernel versions (both stable and dailies) and distros.
One other thing you guys could try is, legacy bios sometimes have an option to load video bios into system memory... this may create an artificial delay long enough to wait for the whole system to get power. It is only a millisecond or so.
This issue is probably good windows drivers are too fast, not "bad drivers don't work". Some drivers it is probably accidentally delayed, so it works with a dodgy PSU, but as systems become faster and faster...
Mate you're talking complete nonsense. Seriously, please stop.
LOL so true. No body cares about Far cry anymore but still awaiting Propper crossfire support. LOL.
I given up waiting for AMD drivers to be honest. I like getting new ones though
Dude what are you smoking? Your last posts are totally ridiculous, proof, the heck prooof???
MEI is as buggy shaite and most OEM's are so lazy that they don't move their own arse to update BIOS and their microcodes in order to patch up compatibility with certain hardware. I guess you are also so lazy to actually read the documents about MEI does.
MEI is already uncompressed and active all the time it does not reset with shutdown, it alive on +5SB. What are you telling about system clock? System clock is system clock, and PSU? Drivers too fast, what, so what?
Timings? It is well documented, and usually the GPU vendor is at fault not filling out compatibility settings made by AMI/Phoenix/Award. And Windows/Linux/BSD programing culture is another different story.
All BIOS mods are carried out by BIOS vendor developer tools, they act properly and no reports are ever heard causing such problems, like you shout.
X58SO random BSOD using SSD AHCI = orom upgrade (intel itself didn't give a crap updating their own board)
Fixed Intel ****ty GBE 82579V Lan = manual oprom upgrade.
Fixed TRIM Raid 0 = OROM upgrade
Fixed NEC/Renesas USB3 handling = Firmware Upgrade
Gosh I remember my screwy times with nforce 3500, nvidia never did proper driver releases for certain chipsets at all...
I can tell examples and examples why a FIRMWARE UPDATE IS A MUST.
Stock drivers? Utter rubbish, they are stillborn from a beta test when released still.
Stop spreading POS.
This NEEDS to be fixed in the next driver release. AMD's efforts to optimise games are nonexistent at this point.... WTF is this?
They will provide us with driver update and still the Frametimes will suck in this game. Mantle will be broken for us old gcn1.0 users. The only way to get decent frametimes is by going Nvidia
Its not nonsense, you are just upset. It was my work around for my last graphics card and MB with that dodgy PSU, and it worked. Having the video bios shadowed will slow down the video cards BIOS on modern systems.
The other thing I noticed/learned (after a week or so) was, at the time I did not use a steel case, so the PSU, the monitor and the mother board were not earthed together. This causes all sorts of problems, and the solution was to place a wire from the graphics card case slot metal to the outside of the PSU.... this also stopped whine noise. Many graphics cards leak a lot from the metal plate (from the monitor and card), so ensuring they are all earthed to one point is important. Perspex, aluminum and some painted cases will all have this problem. (and to be extra careful, a good UPS or one power board should be used to one power point to ensure you have one earth at wall as well)
i.e. What I am getting at is people will blame the homemade PC components, but they have just made a whopsy, and they need to work it out (or whinge and swap their PC). If is was a real problem with the cards or drivers then everyone would have the problems. If it still happens after ruling out the PSU then the card or the MB is faulty, or the monitor is old.... I noticed one guy still had a crt.
Do you live in the 80's? I guess you missed the UEFI train? Posting up the PC takes 5 secs, and the most of it is HDD spin up time delay...?
That shadow copy actually causes more trouble than gain... on some lousy boards it is mapped in L2 cache, yes CPU cache is permanently used by crap. Then many kernels are unaware of an used address where that critter resides and thus kernel panic may occur?
Enough? Or you are busy rolling back windows update drivers... :infinity: