An Nvidia console seems a no brainer to me. Who else could Console?

[Rangers]

Occasionally you see discussions on video game forums about who might be a candidate join the console business in a big way, if anyone. Apple usually comes up prominently. Which IMO there's zero incentive for an Apple console.

There's one name rarely mentioned in such discussions which to me is arguably the ONLY big player that makes sense to join the console fray...that name is Nvidia.

What brought this to mind again is some internet discussion about the Nvidia Shield I've happened across, at hot deals forums. Despite being virtually abandoned by Nvidia IMO (they have not refreshed the chipset in a while), and overpriced at 199 for the entry level model with only a paltry 16GB of flash storage on board, it still seems quite popular amongst a niche as a relatively open streaming/gaming/emulation/media device.

So I guess you could say technically, Nvidia already has a console. The Shield has a Tegra X1, and it's not underclocked like the Switch, so it's a bit more powerful than the Switch arguably. Which isn't bad. However it's still pretty technically lacking compared to an Xbox One S at the low end of the current core consoles.

I think a console would have pretty huge benefits for Nvidia. It would increase their presence in the gaming market altogether, which would of course be nice for their core PC business.

Anyways with the Shield already existing, the landscape for me seems prime. Iterative modular consoles are now a thing, making it easier than ever to join the console fray. All Nvidia literally has to do is bring the next shield revision up to a level that it can accept 3rd party core ports, and court a few of those ports. So something around 1.5 teraflops on the lower end (and they could go much higher). For Nvidia that would be not even a millimeter sized speed bump obviously.They dont need a giant marketing budget or huge fanfare to launch a bespoke console like in the past. In theory they dont even need a single exclusive software. Just make it at a minimum "The Shield that now has support from Ubisoft, EA, Activision, Take 2, etc"

Of course one might wonder why anybody would buy The Shield console for 3rd party when the Xbox and PS4 have exclusives. Well, that would certainly be Nvidia's challenge, but I dont see it as much of one. They could attack it countless ways. From making it a power attractive box (say 3, 4, or even more than Xbox X at the high end, in FLOPS). So it could gain sales by being one of the better places to play third party games (at least above vanilla Xbox/PS4). They could attack it by being much more open in nature than other consoles, as I feel like the Shield TV is presently. Further, I think the conceit that they'd even need lots of sales might be a false one, I dont get the feeling the current Shield TV is a huge seller, yet it seemingly persists comfortably. On top of all that, I think the impact of exclusives is arguably pretty overrated, excepting Nintendo, and the sales numbers bear that out. Exclusives may get talked about an outized amount, but most people are playing COD/Fifa/Madden/NBA 2k/GTA

Maybe they could even, gasp, offer free online as a differentiator! Or even $20-$30 yearly instead of $50-60. I dont see Nvidia of all companies leaving $ on the table much, but maybe if it was a necessity to gain traction...Quite a few Madden and Fifa fanboys might be enticed by "console that plays FIFA but with cheaper online".

The PS5 might be right around the corner, so what? The iterative landscape has IMO made that irrelevant. In a relatively short time after PS5 release, Nvidia can easily release "Shield TV, power competitive with PS5" version, no problem at all.

Oh and fanboys, fanboys would be a big help in any console publish. Well have no fear, Nvidia has those in spades already. More than almost any other company IMO, at least up there with Sony, Apple.

People might wonder why this post, am I some huge Nvidia fanboy? The reverse if anything! I just find the business discussion of consoles very interesting and this has long been in the back of my mind. I also have always found a more open console that supports android emulators to be an attractive idea.

I guess one more caveat, people might say Nvidia doesn't need in the console business anymore as they supply chips for the Nintendo Switch. That's true to some extent, but it doesn't really invalidate anything IMO. If anything I'm sure Nintendo needs Nvidia much more than vice versa. I dont see it as a dealbreaker or even that Nvidia would stop supplying Nintendo with future chips for say a Switch 2 even, necessarily, for that matter.

Do I see this happening? Well, I dont see any signs of it, at all. So it certainly currently looks unlikely, no matter how much business sense I may think it makes. But I think it would be at least exciting if The Shield 2018 (if there is to be one) went this direction.

 

[manux]

What about geforce now as console substitute? No need for physical box

https://www.theverge.com/2018/1/8/16865480/nvidia-geforce-now-windows-beta-app-gaming-pc-ces-2018

 

[mrcorbo]

What you are suggesting they do is the exact same things they did (nearly point for point) when they made the original Shield, which was a flop. Why would doing the exact same thing the exact same way get a different result this time than it did last time?

 

[Shortbread]

I would like to see a custom Nvidia GPU within Sony's or Microsoft's next machine. But, that ship has already sailed and Sony/MS didn't enjoy the ride. Other than that, no. Nvidia can stay out the console gaming space with any of its wares. Three is enough...

 

[dobwal]

They already have a console, it's called the PC. They don't care about software licensing fees because their hardware margins are absolutely stellar in comparison to consoles.

 

[Deleted member 11852]

They already have a console, it's called the PC. They don't care about software licensing fees because their hardware margins are absolutely stellar in comparison to consoles.

You might want to look at Nvidia's financial's again. Net income hovers around 33% of revenue so $800m profits off $2.6bn in revenue because of massive R&D costs. Go back quarter, over quarter, it's the same story. Most of their revenue is from enterprise and not gaming so having something actually in gaming would be good.

 

[dobwal]

You might want to look at Nvidia's financial's again. Net income hovers around 33% of revenue so $800m profits off $2.6bn in revenue because of massive R&D costs. Go back quarter, over quarter, it's the same story. Most of their revenue is from enterprise and not gaming so having something actually in gaming would be good.

I have look at nvidia earnings.

http://s22.q4cdn.com/364334381/files/doc_financials/quarterly_reports/2018/10Q-2018-Q3.pdf

$2.6 billion in revenue of which $1.5 billion is recorded under "gaming". Their overall GPU business was worth $2.2 billion ($978 million recorded as profit) while the Tegra business is worth ~$400 million in terms of quarterly revenue.

Yes they have massive r&d but $426 million is shared across their entire business. And whatever is recorded as an operating expense for gpu still allows 45% of revenue to be recorded as profit.

Sales of actual console hardware arent producing anywhere near 45% in profit margin for Sony and MS which was my point of argument.

 

[DieH@rd]

They irritated MS with original Xbox, and they irritated Sony with PS3. Nintendo needed a modern solution [which they got by purchasing a plain untouched tegra chip], and they picked Nvidia by default.

I don't expect Nvidia in MS/Sony consoles anytime soon. AMD is not going away from mainline Xbox/PS consoles.

 

[zed]

The main problem is Consoles don't normally sell at a large profit, profit's come from software. nvidia has no studio's thus will be 100% 3rd party. I don't see them taking this gamble.
I read today on artechnica Google maybe getting into the console streaming business, I can't see it being a success though

 

[Goodtwin]

Entering the console space as a new player would be extremely difficult. Nvidia may be a big player in the PC gaming world, but they are far from a household name in the console world. They would be building the brand from the ground up. They arent like Nintendo and Sony with a ton of exclusive first party content that has years of recognition driving their product. Nvidia would basically be offering a third party software machine with little to differentiate themselves from the pack.

 

[Shifty Geezer]

Valve couldn't manage it with Steam Machine; what makes you think nVidia could? The only people who could manage a console now are Apple and Google because they have the ecosystems players would tap into. An Apple console would be an iOS box of course. As a left-field option, someone like Tencent might create a console for the Asian market, and if it does well could make inroads West. nVidia, Intel, Dell, no chance. You need the OS and games and no hardware company is in any position to manage that. Samsung could try throwing money at people - seems to have gotten a bit of traction for Tizen. None of them will go 'high end' (desktop level hardware).

 

[sniffy]

Entering the console space as a new player would be extremely difficult. Nvidia may be a big player in the PC gaming world, but they are far from a household name in the console world. They would be building the brand from the ground up. They arent like Nintendo and Sony with a ton of exclusive first party content that has years of recognition driving their product. Nvidia would basically be offering a third party software machine with little to differentiate themselves from the pack.

This is it right here. Why buy a new box that only has third parties on it? It’d have to be more powerful than an X or Pro to bring anything of value to the table at all. Unless they started paying PC only devs to port their stuff over exclusively.

 

[Silent_Buddha]

Not going to happen for a variety of reasons.

NVidia aren't terribly interested in low margin products. They have been increasing their margins for the past few years due to limited competition. Getting into consoles means being willing to take low margins in order to win a contract. Low margins that then impact their financial reports. Which in turn could affect their stock price.

Ah, but Nintendo Switch, some might say. Well, the thing here is that the Switch is using what was basically a failed commercial product. They gained a lot of experience from it, and it helped them gain entry into the automotive market. So it's not a failed project in the sense that it moves them closer to being profitable in that product segment.

As such, since Tegra was already a sunk cost, selling virtually unmodified modules to Nintendo served to not only recoup some of that sunk cost, but also bring in visibility for the platform.

IE - it's not the same as spending a significant amount of R&D to produce a custom low margin part which can only be sold to one customer.

Assuming Nintendo stays with NV for their next console (no guarantees) it is highly unlikely that NV. will make a custom part just for Nintendo unless it included high margins. More likely is that they would be willing to make a deal for Nintendo to just use an existing NV. product. And that's fine since there is already hardware progression from what Nintendo are currently using.

As to making a console? NV. have tried that twice now (Shield handheld and Shield console which became Shield TV) and failed both times. Margins in console hardware is minimal at best. They would need heavy software revenue to make the project attractive. They failed to attract many developers to Shield in both of it's first incarnations. Why would they think that would change with a 3rd incarnation?

Additionally, Shield was a project mostly to show other hardware manufacturer's that Tegra was something they should consider adopting. Shield failed in that as well. Now that NV. are actually making significant inroads in a product segment for the hardware line that Tegra used (automotive and AI) there is little to no reasons for them to make another low margin "console" like device.

So... While you can never say never. I don't see anything that would entice NV. into [1] making custom console hardware for others or [2] making their own console.

Regards,
SB

 

[Pixel]

No Nvidia would fail. The market is too saturated and the existing competitors are too entrenched.
There is a big problem with launching a console. It MUST MUST must be a locked closed platform (or else a massive portion of the install base will install steam/gog etc and thus Nvidia gets no 3rd party royalties no Gold/Plus membership and no 30% digital distribution cut) or you have to sell the hardware at a profit, like Steam machine. (Don't believe me? Around half of the 150 comments on amazon from Steam Machine buyers showed they removed SteamOS and installed Windows.) https://www.amazon.com/Alienware-Ma..._dp_d_acr_sr?ie=UTF8&reviewerType=all_reviews

They would fail because they would need to hit critical mass of users to be profitable on software royalties and services.

1. If you launch a console you need to sell them at slim margins or a loss in order to compete with Sony & MS (Nvidia might make additional slim margins given they are the chipmakers, not more than slim margins though, given AMD itself has officially stated they only makes slim margins on consoles).
2. You need to sink cumulatively hundreds of millions into the platform and advertising, and possibly into 1st party studios. Yearly operating costs with services as competent & competitive as Sony/MS could be in many tens of millions of dollars annually.
3. Nvidia would need tens of millions of users buying software that earns Nvidia royalties and generating similar revenue to MS on digital distribution services.
4. Or have the revenue model to support Nintendo levels of 1st party studios

This means until the userbase grows to tens of millions they will be taking losses. They would need significant edges in certain regards to pull people away from MS & Sony.

 

[Deleted member 11852]

$2.6 billion in revenue of which $1.5 billion is recorded under "gaming". Their overall GPU business was worth $2.2 billion ($978 million recorded as profit) while the Tegra business is worth ~$400 million in terms of quarterly revenue.

Nvidia attribute all profits derived from consumer chips and cards to gaming even when they are going to server farms and other enterprise facilities. It's just how they report, But most people in my business have a good feel for the quantities of consumer chips being used for non-consumer applications and that excludes miners.

 

[Silent_Buddha]

Nvidia attribute all profits derived from consumer chips and cards to gaming even when they are going to server farms and other enterprise facilities. It's just how they report, But most people in my business have a good feel for the quantities of consumer chips being used for non-consumer applications and that excludes miners.

So, what you are saying is that, excluding crypto-miners, most GeForce branded cards are being used for non-gaming purposes? I know many Titan branded cards are used for non-gaming purposes, but that's a small drop in the bucket of the gaming segment. Also I'm assuming Titan is in the gaming segment in their financial reporting, but I'm not actually certain if it is.

Regards,
SB

 

[Deleted member 13524]

More likely is that they would be willing to make a deal for Nintendo to just use an existing NV. product. And that's fine since there is already hardware progression from what Nintendo are currently using.

There is?
Xavier has a 2x larger GPU but is IMO out of the question because it's a large and power-hungry chip that is now completely distant from Tegra's initial mobile focus.
I guess they could use Parker (which is also old right now), which has better clocks at its Max-Q cTDP, but it would be able to do little more than docked mode while mobile.

But Parker also has a lot of non-gaming stuff and nvidia sells it for a lot of money right now. Way more than what Nintendo is willing to give them. It also has twice the number of LPDDR4 channels but Nintendo would probably just use half of them.

I do wonder if simply developing a tiny chip at 16FF+ that is basically a TX1 stripped of all the stuff that isn't being used (like 30% of the whole chip?) and could clock higher on the GPU would be that much of a hassle to nvidia.

 

[Deleted member 13524]

As for the topic at hand, I have two opinions:

1 - The Switch has horrid hardware. Not only is it using a very old and very downclocked SoC, it's also missing critical features like analog triggers, microphones or the ability to connect bluetooth headsets. Apparently it now supports USB audio devices, which we obviously have to connect through a USB-C -> USB-A adapter, which makes the feature rather useless for mobile usage. Combine that with the fact that the console is on the expensive side and it leaves a lot of room for a PC-like platform to trample it.

2 - nvidia is not the company who would go forward with it. Not only are they the platform developers of the Switch (meaning they probably signed a non-compete), their Tegra line of chips is a failure for mobile and a success for automotive. That said, I don't think nvidia will pursue the mobile route anytime soon.
Which BTW may put Nintendo in an awkward position in the long run. Nintendo was happy to keep nvidia's leftovers at a discount, but nothing indicates that nvidia is willing to make a new chip for the amount of money Nintendo is willing to spend. And now there aren't any new leftovers for Nintendo either.

 

[Goodtwin]

There is?
Xavier has a 2x larger GPU but is IMO out of the question because it's a large and power-hungry chip that is now completely distant from Tegra's initial mobile focus.
I guess they could use Parker (which is also old right now), which has better clocks at its Max-Q cTDP, but it would be able to do little more than docked mode while mobile.

But Parker also has a lot of non-gaming stuff and nvidia sells it for a lot of money right now. Way more than what Nintendo is willing to give them. It also has twice the number of LPDDR4 channels but Nintendo would probably just use half of them.

I do wonder if simply developing a tiny chip at 16FF+ that is basically a TX1 stripped of all the stuff that isn't being used (like 30% of the whole chip?) and could clock higher on the GPU would be that much of a hassle to nvidia.

We are at minimum 5 years away from the Switch successor being released. We could see a "New" Switch released in a couple years rocking a TX1 manufactured on a newer process, but the true successor is many years away. This means the processor that it will likely use hasn't even been created yet. For Nvidia, continuing the Tegra processor R&D makes more sense than ever knowing they likely partner in Nintendo. Unless there is a contract nothing is certain, but if Nvidia wanted to continue the Tegra development anyway, targeting Nintendo with a future chip makes sense.

 

[Recop]

They are among leaders in one of the most promising market : artificial intelligence.

More we move forward, more the console market is becoming pointless for them.