Why i feel betrayed by my g750jz. And why i hate optimus.

[broadleaf]

Hi guys. Im a professional 3d artist/designer and recently im a gamedeveloper.

Here are some of the problems i encountered with this laptop so far, which make my life miserable.
My main problem is that due to some circumstances im stuck with it, for maybe couple of months ahead for sure.

Why dont i start from the very beginning?

* Bought it, in retail everything looks fine. Brought home - 'W' key falls off. Just like that you press its lower part, it fells off.
No visible damage, if you press it normally you cant tell defect (just works) but once pressed lower part it jumps right off.
Bring it back to vendor - "Sorry its your fault, you broke that" in know i did not.
They want to scam me? Fat chance. Call police, lawyer, customer protection etc.
Next day they begging me to not drag them in court. Okay i don't care just give me working laptop.
Exchanged took money from them for travel time and travel expenses.

* Screen is crap, vertical lines, tho colors and response time are good, ignore that i wont look at that screen so much.

* Start mirgating my software etc, new to windows 8.
Suddenly my life becomes a miserable hunt for hacks to disable all MS horse poop and removing tons of bloatware
carefully provided by ASUS. Hacks to disable charms. To kill Edge UI, to kill Skydrive, to this to that etc.

* Dinamic brightness cant be turn off, but whatever.

* Realise i can't switch completely to Nvidia GPU.
MFW

After 2 months of suffering with all my pro software that you have to manually tweak to work properly disregarding all
optimus attempts to flip me over.
Something happens, that overrides all my registry hacks for windows, rendering months of careful tinkering flush in vain.


Install windows 7.

Realise that windows 8.1 has better performance with optimus.

Hack explorer.exe and dwm.exe to wrok on gpu.
Doesnt change much still pathetic HD4600 is responsible for rendering windows interface which leads to numerous issues
while Aero is off Photoshop cant recognize my GPU no matter what.
While Aero is on Unity 3d drops performance 2 times.
Performance in games is somewhat better sometimes.

i have 3 screens.
In windows 8.1 multiscreen has better performance. Sad to see this

with 3 screens and

with one.

Realize that spitting on my settings for desktop rendering Optimus(tm Pest Technologies) still uses intel for rendering
windows. Everything feels sluggish and response time is bad with Aero on.

Have problems with Unity3d 4.6rc18 completely not being able to recognize nvgpu disregarding anything,
which it should not, because its beta, but would not happen if i wasn't locked on that pathetic intel chip.

No fan control for gpu whatsoever.

Crippled nvidia 880m not working on 100% official nvidia responce is silence check this thread
https://forums.geforce.com/default/topic/754922/geforce-mobile-gpus/nvidia-crippling-the-gtx-880ms-w...

12 pages of people ranging from professional OC's to casuals. Ofc silence like here, because why even bother - they already bought it.

For those 2 months i own it, ive seen only one official response on forums when someone accidentally insulted Asus member and they transitioned from ignore mode, to rage induced SHUTUPORIKILLYOU type of behavior. Which leads to conclusion that they do read this but will act like they don't.

So far optimus caused more nerves, stress and hatred than any other piece of hardware equipment i ever had, ever. Ever.
In my opinion this is the worst decision that was made with this laptop which otherwise (disregarding cheap screen and other minor issues) is good.

And the official silence on the issue is just destroying my trust in this company every day more and more.

Ok i can get that you FAILED miserably by propagating GPU outputs into intel buffer and then it goes to screen.

BUT EXPLAIN TO MY WHY THE HELL IS WINDOWS RENDERED BY THIS PATHETIC PIECE OF JUNK?
Why cant nvidia gpu render EVERYTHING and then just output it to screen through that buffer?
Why?
WHHHHHHHHHHHHHYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY?

Because it will cost more than neural cells of couple of suffering customers? Because you.. can't? Don't want to?
Too hard? Requires admitting you have done something wrong? Even hmscott which tries hard to defend ASUS
agrees that it is WORST decision. Whats gonna happen? Silence. No solution what so ever.

Conclusion is just really, if you are working with 3d/2d gamedev, cg, compositing etc, and you seek for a serious machine
that can carry any load... SADLY THIS IS NOT IT, THANKS TO ASUS INCREDIBLE DESIGN DECISION.
Everything is fine, just everything, but being locked on Intel Chip causes so much compatibility issues with pro software
that you will find yourself in endless hell of seeking solutions to problems that are caused by this, on daily basis problems occur.
This doesnt work right, that throws some exceptions, this lags, that just stupidly refuses to work etc etc etc..

PS. i will contact ASUS support and just copy paste this wall of text to them, see what they will respond.

In mean time all asus forum member which are reading this are free to maintain silent mockery, and imply problem does not exist and im just mad for no reason.

[hmscott]

broadleaf wrote:
Yeah got confused those.
Btw hmscott dunno if you know but a little story of quadro series.
Back in the day when they just came out they were giving better performance in 3d apps.
Sure, thats what they were for, right?
Thats why the cost was around 3k and more.
It worked like that - you have quadro, and a driver = more performance.
Suddenly people hacked driver to work with Geforce 9800.
Suddenly you get same EXACT performance from 9800 for 300 as you get from quadro for 3000.
It was proven that quadro and geforce are exact same component and only actual difference was the
driver that detected which one you had and if quadro it switched to high performance mode.
I wont even go into any conclusions. Just lies are tiresome man.

broadleaf, Nvidia develops and supports software/drivers for professional use, and Nvidia needs to recoup those historical and ongoing real costs, so they apply that overhead to the cost of the Quadro Hardware.

Originally it was a software switch that could be flipped, and Nvidia widely published the information needed to flip the switch. Nvidia was cool with people that couldn't afford the high cost of the cards flipping the switch so they could enter their respective professional market, reducing the barrier to entry, helping nurture these new emerging market segments.

Nvidia was helping their customers enter their professional market and get profitable before they needed to bear those high costs. And, most people did just that - they started buying Quadro cards after they got paying work in the field. Both Nvidia and the users were happy.

Nvidia didn't want those sunk costs in development of professional markets to be a barrier to entry for new users that couldn't afford the costs until they got paying work. Nvidia didn't think it was fair to amortize those costs over all their graphics cards because most gamers didn't need and shouldn't have to pay for those features.

Then users began abusing this gentleman's agreement - users kept flipping the switch even though they were now making $$$ and could now afford to help Nvidia cover those continuing sunk costs Nvidia was incurring.

For many well intentioned users, the financial departments of the users business became aware of the software switch, users were forced by them to flip the switch and not pay for the more expensive Quadro cards. They would ask, why are we paying more for these tools that we need to get work done when we can configure the software and make it work for less $$$?

Nvidia was forced to implement the switch detection in hardware - which can still be worked around - but hacking and soldering hardware is a more visible effort to "work around" what are essentially Nvidia's professional software licensing costs.

Now users are able to budget for those higher cost Quadro cards in their paying projects, and the financial people had a more understandable firm barrier to consider than the esoteric conceptualization of a Gentleman's agreement.

The Gentleman's agreement is an esoteric consideration that depends on an understanding of the costs Nvidia is incurring, an appreciation for Nvidia's help in building the software that makes the professional markets possible, and a grasp of the moral limits that one must cross to abuse this Gentleman's agreement.

Nvidia needs to recover their costs, and they can't depend on their end users being able to pay as they go / can, so Nvidia had to implement the current method to insure that they can recover their development and support costs, and keep the niche professional market segments alive.

[broadleaf]

So you're saying that Nvidia was developing a Software and to make people pay for it the developed a whole new GPU branch?
If you're developing software you are charging for software. You don't develop a whole new car to sell your new tires, and then make
the car mandatory. They are just making more money, that's it. Its a pure marketing manipulation, along with programmable obsolescence.

[hmscott]

broadleaf wrote:
So you're saying that Nvidia was developing a Software and to make people pay for it the developed a whole new GPU branch?
If you're developing software you are charging for software. You don't develop a whole new car to sell your new tires, and then make
the car mandatory. They are just making more money, that's it. Its a pure marketing manipulation, along with programmable obsolescence.

broadleaf, it is a conceptual leap to get this, so give it a little more thought before condemning it.

Most of your complaints so far are based in ignorance and anger about that ignorance. Please educate yourself on these issues before jumping to negative conclusions, and condemning others and labeling their actions as immoral or illegal.

A good rule of thumb is if you are quick to anger about a subject, check yourself and stop the resulting reaction from coming out publicly, and then review the information until your response is expressed as a question indicating a desire to understand.

It is a common fault in modern times - there are many complex issues that from the outside seem simple, and there is a clear conclusion to jump to - that is totally incorrect.

It is the failure to conceptualize a situation from others experience and point of view that steers the reaction in the wrong direction.

Even though it is a modern problem, it is based in the distant past, this reaction to the unknown - the mind needs a clear place to file the information it sees but doesn't comprehend, and it is safer to file it in the "this is bad" file and marked it as "to be avoided" - environmental selection favored those that took that approach - so this reaction is built into our base code.

Nvidia can't charge developers to use their software, and they need developers to build their market. Even Apple doesn't charge for developer tools now. They did in the past, but realized they were limiting their market by setting a barrier to entry. At worst, inexpensive hardware kits are available to developers to reduce the barrier to entry / curiosity.

So Nvidia can't recoup their costs from developers. They can't recoup the costs from licensing their software either - it is included with the hardware and freely downloadable drivers / software. So the only way Nvidia can recoup their costs is to apply the burden of the costs on the items specific to the markete that reap the benefits.

Those that use those optimized libraries, procedures derived from research that cost money to derive in the first place - and continue to cost money to support and improve over time, need to pay for those costs.

A way to accomplish that is through a "hardware dongle" - a piece of hardware required to run the software. In the case of the Nvidia Quadro boards, the hardware part of the dongle is integrated into the card, not a separate piece as is popular for software licenses.

https://en.wikipedia.org/wiki/Dongle

"A dongle is a small piece of hardware that attaches to computer, TV, or other electronic device, and that, when attached, enables additional functions such as copy protection, audio, video, games, data, or other services. These services are available only when the dongle is attached."

So, the Quadro card was born. Hardware license key costs minimized since the simple logic activated through software is built into the board. Only the people that need the special Quadro software support will buy the card, so they bear the burden of the costs.

Nvidia has also in the past blessed those Quadro cards with more memory, better memory bandwidth, and other performance and reliability features.

Over time the gaming and niche markets need for RAM, memory bandwidth, and better cooling for reliability have converged, so just about the only thing left is the dongle/license/software to differentiate them.

But the need to recoup the ongoing development and support costs for the unique market uses for the Quadro still exists.

Is there a point at which there is no longer a need to differentiate the boards, and the overhead can be spread across all markets? And all hardware can run all software? Maybe, but, unlikely - the budgets for the software support for these specialized markets are indexed to how many boards are sold, to keep them self-supporting.

The CUDA niche market is unique in that there are no CUDA only cards specific to that specialized market. But it does automatically bear the cost burden for development by the sale of only the highest end boards into that market segment.

The users cost of servicing each CUDA capable motherboard slot prevents users from using less expensive cards to implement rendering farms - it is more cost effective to use only the highest end Nvidia card with the best CUDA performance.

It makes a lot of sense when you think about it, or get walked through it, or lived through the industry as it was organically grown, but Quadro vs GTX pricing is one of those things that seems like a con-job from first glance.

[broadleaf]

ok i see. ill sell it in some time. will cut 300 from init will make some GAMER happy.
I dont understand these companies. They know they produce junk, yet they do it.

Neither have the cries of thousands of vocal users made any difference.

Never again. I don't support companies that ignore their customers.

[hmscott]

broadleaf wrote:
ok i see. ill sell it in some time. will cut 300 from init will make some GAMER happy.
I dont understand these companies. They know they produce junk, yet they do it.

Never again. I don't support companies that ignore their customers.

broadleaf, so you aren't going to purchase Nvidia, Microsoft, Intel, or Asus products ever again? 🙂

Asus does listen to it's users, that's why Optimus was included in the new G750's - unknowing users demanded Asus enable the Intel GPU for better battery life. Now many users like you and I know what a misery Optimus is, and want it taken out.

Asus isn't ignoring you, you haven't talked with Asus yet, have you?

The Asus responses you see in the forums are from Forum administrators, not product managers, engineers, or other people that can make decisions to implement your desires.

Give them a chance and talk with them via the channels provided.

You can't say they haven't listened to you, yet.

Let us know what they say 🙂

[broadleaf]

sent them email

Hello i want my system to be fully working from NVidia GPU without use of Intel chip to render anything.

Proceed to this thread to fully read what issues i encounter with your bugged Optimus and what awful experience i have with your flawed product.
http://rog.asus.com/forum/showthread.php?51606-Why-i-feel-betrayed-by-my-g750jz.-And-why-i-hate-opti...

I want you to make viable solution that renders everything with Nvidia GPU, and only then puts that 1 image into buffer which is outputted into screen.
Are you able to do such simple thing?

I can say on behalf of all users of ASUS g750jz, - we need you to fix this. This is making our experience with your product miserable.