NVIDIA's Optimus: Will It Come To Linux?

[deanjo]

And OEMs have proven SEVERAL times, that the customer is nothing they care about either (glare displays for example).

Bit of a bad example with the displays. I recall when I was at Apple and the option of glare display was given on the Macbook Pro's at a no cost option. It was about 50/50 split there to which people prefer.

[Hasenpfote]

Who do you think actually drives the requirements? How much does the OEM implementation affect the driver architecture?

Well, it seems that the customer is not the important role in this industry. It's like with a programmer, his program and the user. The user does ALWAYS disturb.

Marketing is either forward looking (how I can sell more widgets), or complementary to a customer (I need to meet this requirement for to help a customer sell more widgets). Note that as a supplier, the customer is really the OEM.

Well, without being too aggressive, the marketing of AMD has proven SEVERAL times, that the customer (=people, not OEM) are nothing they care about (e.g. no netbook-GPU/CPU, no low power consumption CPUs, the 48xx series with incredible idle power consumption, no fast and passive graphic cards (I own a passive 3870. The last one was a 4670, I think, which is a 3870 indeed), etc.).

And OEMs have proven SEVERAL times, that the customer is nothing they care about either (glare displays for example).

So where in this list is the customer?
What can I do as customer tog et what I want? Buying AMD?

[Svartalf]

Who do you think actually drives the requirements? How much does the OEM implementation affect the driver architecture?

Marketing is either forward looking (how I can sell more widgets), or complementary to a customer (I need to meet this requirement for to help a customer sell more widgets). Note that as a supplier, the customer is really the OEM.

In the end, this is why we don't have as much support as we ought to with things in the world. The OEMs are still dancing to the tune of Microsoft's whims- and thereby companies like AMD have to dance to it, even if they don't fully want to. If they don't they don't sell to their customers (we're not their customers...we're HP's, etc. unless we're buying bulk quantities of GPUs from them...HP, Dell, Lenovo, etc. are their customers...) they don't sell much.

So you're going to have to convince the OEMs to work at getting things more straightened out.

As for not supporting Optimus, that was something of a poor decision on NVidia's part- even if I understand the motivations behind them choosing that. They're going with what the OEMs are asking for- the main reason we have drivers is the workstation market and Optimus isn't likely to be used in that space. So, no driver support for it for us.

All this means is more room for AMD to get it's act together on things- and take the market segment from them (and there IS one, it's just not on your OEM's radar...). I WAS going to get a new laptop with NVidia because of the stuff just largely working on Linux (AMD's proprietary drivers have in the past been iffy in varying areas...and rumors of issues with suspend on the mobile parts has me still thinking there's still a bit to be done...) - but if they're going to play that game, I might have to reconsider that decision.

[mtippett]

Sounds easy. Why is it not in the ATi Drivers? Besides the neverseen PowerXpress ("Available on notebook PCs using Windows Vista? OS")

Can anyone please tell the developers at ATi what CUSTOMERS want? And make sure, that the marketing guys are stuck in snow somewhere, so they don't disturbing the devs with useless things.

Who do you think actually drives the requirements? How much does the OEM implementation affect the driver architecture?

Marketing is either forward looking (how I can sell more widgets), or complementary to a customer (I need to meet this requirement for to help a customer sell more widgets). Note that as a supplier, the customer is really the OEM.

[BlackStar]

Quit sad, but well. It's not meant for workstation users anyways .

I'm always wondering why this stuff can't be integrated in one GPU anyways. A simple gpu for the composited desktop and a real 3D unit which can be completely switched off. Or something more integrated: why can't maybe 98% of those processing "cores" be disabled on low workload. There should be a lower idle consumption be possible than AMD's current 5xx0 series has (>15W).

The issue is that you need to run the memory at some relatively high clock in order to feed the framebuffer with data (1920x1200 at 60fps = 527MB/sec) and this takes power.

However, with IGPs being integrated into CPUs right as we speak, the future might look quite different. The output signal could always be routed through the IGP (so no display hardware on the dedicated GPU) and the extra GPU could only be fired up when the load becomes too heavy for the IGP. Something similar to the Voodoo2-era 3d accelerator add-on cards, in fact.

[leidola]

Nvidia: We have no plans to support Optimus on Linux at this time.

Quit sad, but well. It's not meant for workstation users anyways .

I'm always wondering why this stuff can't be integrated in one GPU anyways. A simple gpu for the composited desktop and a real 3D unit which can be completely switched off. Or something more integrated: why can't maybe 98% of those processing "cores" be disabled on low workload. There should be a lower idle consumption be possible than AMD's current 5xx0 series has (>15W).

[macemoneta]

Nvidia: We have no plans to support Optimus on Linux at this time.

http://www.nvnews.net/vbulletin/showthread.php?p=2183477

[BlackStar]

The other thing I expect in this nvidia implementation is that it will only work with nvidia/nvidia pairs. Likely as a result of driver constraints.

Actually, this technology was demoed with a Nvidia/Intel pair. Also, it gave you the option to right click an app and run it with either the Nvidia or the Intel GPU. You could also download application profiles automatically, or even build your own (i.e. always run application foo with Nvidia) - pretty slick if I may say so.

If dual GPU solutions become widespread in the future, I can imagine this coming to the open-source stack at some point. From a previous discussion it seems that X isn't really flexible enough for this right now, so it's mostly a matter of developer interested and manpower.

I also doubt this will come to open- and closed-source driver combinations (e.g. Intel + Nvidia) without very wide cooperation between all competitors (fat chance?) Time will tell.

[droidhacker]

I have no doubt that this, or something functionally equivalent, will be coming to linux at some point. Will nvidia implement it in their blob? I don't know and quite frankly, I don't give a rat's a$$.

To be honest, I really don't want any kind of magical hidden automatic switching... a manual mechanism + an optional daemon monitoring (to monitor for switch conditions and automatically switch it) and switching it will be enough. Let me be in control of it if I happen to want to be in control of it.

The other thing I expect in this nvidia implementation is that it will only work with nvidia/nvidia pairs. Likely as a result of driver constraints. Gallium3D with its softpipe should introduce the possibility of switching *live* between two DIFFERENT GPUs (since all GPUs, in the end, will have equivalent capabilities, albeit at vastly different levels of performance). And this, I think, will be really slick.

[Hasenpfote]

This sounds like new marketing on an existing concept. You can do this already if your system has two GPUs of the same capability level, you just need some extra smarts in the driver. It's basically power selective multi-gpu. Wire one to the displays and then render using the faster or slower one depending on what power mode you are in. When the other one is not in use, ramp it's clocks down.

Sounds easy. Why is it not in the ATi Drivers? Besides the neverseen PowerXpress ("Available on notebook PCs using Windows Vista? OS")

Can anyone please tell the developers at ATi what CUSTOMERS want? And make sure, that the marketing guys are stuck in snow somewhere, so they don't disturbing the devs with useless things.

I vote for PowerXpress on the DESKTOP with high-end graphic card and IGP. But I repeat myself (9 month ago):

http://www.phoronix.com/forums/showpost.php?p=78146&postcount=17

And bridgman did not listen to me