Next Generation Hardware Speculation with a Technical Spin [2018]

[Grall]

I read the 'cartridge style approach' as meaning flash storage provided as blank cartridges you could plug in and download/install to.

Yeah, well, the desire to make proprietary and profitize these cartridges would be enormous, like console makers have done in the past. I don't really see that as a positive development, just another blow to my nads/wallet, honestly...

Even the tiny SD cards are up to 256GB now and lining the side of the box with USB ports or SD card readers wouldn't be difficult.

SD cards are fuggin slow; even the fastest SD cards are very slow compared to M.2 drives especially, and even SATA drives, particularly on small random accesses. (And writes in particular, not that a console would be writing randomly a lot... ) Also, fast SD cards are quite expensive, and cheap SD cards are typically bottom-barrel dredge quality flash chips and have unpredictable (read: bad) performance that will lead to slow loading and poor gaming experience, and may have very bad data retention performance as well. USB flash devices are largely the same; huge spread in quality, performance and price, but generally much worse in the first two categories than proper flash drives.

I seriously doubt a console manufacturer would want all the support headaches from customers using bad flash drives, ending up with really slow loading or data corruption and so on.

i.e., Nintendo wanting NEC in GameCube

Only the fabrication of Flipper was done by NEC, from what I know at least. All the actual IP and other chips came from other companies.

 

[Shortbread]

I could see Sony going more for ARM+Nvidia on PS5 because they're not as invested in backwards compatibility as Microsoft is. I don't think it's likely at all, but if anyone was going to go that way it would be Sony before MS.

If Nvidia offers Sony a banging GPU (17-19TF) at reasonable pricing, then yes, without a doubt no backwards compatibility. Remasters galore!

But this sh** ain't happening ...for so many reasons. Sony seems very content on offering "reasonable wares," without so much overhead and headaches. The new Sony isn't going to rock the boat on making their lives and developers lives harder. In this case, backwards compatibility will keep them competitive with Microsoft when the next-gen rolls around.

 

[ultragpu]

Wouldn't a NV GPU with 17-19 TF be like 20-22 TF of an AMD GPU considering the efficiency? But NV still charges heaps more regardless.

 

[Shifty Geezer]

Furthermore, consider the following other larger strategic business factors that would support an NVIDIA+ARM+Sony PS5 console tie-up.

Not more than the alternatives. You'd have to prove ARM was the best option, and the vast majority of us disagree with that. I don't think it'd be much of a PR win either. nVidia is powering the record-selling Switch, but they aren't making much noise about it. Why? Because they're also powering self-driving cars which is approximately 748 times cooler and more futuristic and worth crazy money. How much can they charge for a SOC used in a $500 console? How much could they charge for the same silicon in a $20,000+ car...

Many years ago, powering the world's most powerful console may have been a bit of a coup. As it is now though, nVidia have the world's most powerful graphics cards that eclipses anything a <$1000 console can ever do, and the world's most powerful super-computers, and the world's most exciting autonomous vehicles. They haven't got the world's most powerful mobile SOC which used to be a big PR win, yet don't give a rainbow-coloured shit any more. Why would they care about a console win if there's no money in it?

The only case I can see is for the fun of it. Raking in the money from elsewhere, nVidia decide to offer a cheap and OP solution to one of the consoles without any view to profit margins, funnelling PR interest back towards their self-driving cars.

"Is it true PS5 can drive my car?"
"PS5 is built on the same technology used by BMW, Audi, and Tesla."

 

[Deleted member 13524]

But then the competition could always counter with 12 TF FP16 as well,

Except they couldn't. If you're talking about the XboneX, it has no 2*FP16 throughput so it has 6 TFLOPs of FP16 throughput.

there's no use for them to be deceptive about it especially in the days of internet, tech sites, YT and social media.

Yet it keeps being done everyday. Especially if it's true in some form (like the FP16 throughput).

Sony really don't have much to worry about right now, not even the best looking X1 enhanced titles could beat most of their first party titles available and upcoming. Like you said Sony thrives on AAA budget no nonsense SP games and that's truly a huge driving point to their success but one should never rest on their laurels at any stage.

Yes, but come 2019 Sony will need to convince PS4 owners to upgrade to PS5.

 

[fehu]

Like the whole cell processor hype (kutaragi)

It worked even too well

 

[milk]

Yes, but come 2019 Sony will need to convince PS4 owners to upgrade to PS5.

Thats why releasing a console when it is actually technologically viable to make one that can produce games the consumer will recognise to be clearly superior is important.

 

[Deleted member 13524]

Thats why releasing a console when it is actually technologically viable to make one that can produce games the consumer will recognise to be clearly superior is important.

2019 titles played in the 2019 consoles will be clearly superior to 2013 titles played in the 2013 consoles.
The same probably won't happen between the games that are well optimized for mid-gens and early titles for the 2019/2020 consoles, but that was a compromise I think Sony and Microsoft have already accepted.

 

[Picao84]

2019 titles played in the 2019 consoles will be clearly superior to 2013 titles played in the 2013 consoles.
The same probably won't happen between the games that are well optimized for mid-gens and early titles for the 2019/2020 consoles, but that was a compromise I think Sony and Microsoft have already accepted.

Hopefully that is where PSVR will make an impact. Make it 100% compatible with PS5 (including patches / automatically downscaling from 2K / 4K for better image quality), discount the PSVR hardware to maybe £150 / 150 euros and promote it as a legit new way to enjoy games versus a simply graphics upgrade that will not be as visible.

 

[Anarchist4000]

Yeah, well, the desire to make proprietary and profitize these cartridges would be enormous, like console makers have done in the past. I don't really see that as a positive development, just another blow to my nads/wallet, honestly...

Wouldn't an added revenue source encourage the adoption, for better or worse? The thinking would be to stick the device into a buddies console and log in your account. Yes there would be some considerations, but it may be a more enjoyable experience and bring in some extra cash. Always have collectors edition drives along with merchandising.

SD cards are fuggin slow; even the fastest SD cards are very slow compared to M.2 drives especially, and even SATA drives, particularly on small random accesses. (And writes in particular, not that a console would be writing randomly a lot... ) Also, fast SD cards are quite expensive, and cheap SD cards are typically bottom-barrel dredge quality flash chips and have unpredictable (read: bad) performance that will lead to slow loading and poor gaming experience, and may have very bad data retention performance as well. USB flash devices are largely the same; huge spread in quality, performance and price, but generally much worse in the first two categories than proper flash drives.

In the case of a console, the only writing should be patching. Random reads uncommon depending on implementation. Not unlike DVDs or older models. Speeds have gotten far better in the past year or so and quality would be something to look in to. Branded devices could bypass quality concerns. Always the option of loading the entire image to memory or a small, high performance NVMe drive used as a file cache for the last few titles loaded. Most gamers I wouldn't expect to jump between games rapidly. Even if a device dies, the data could be downloaded again easily enough.

Interesting concept to consider at the very least and performance would improve throughout the consoles life as the tech improves. Two years down the road that SD card may rival relevant current NVMe storage.

 

[Deleted member 13524]

Hopefully that is where PSVR will make an impact. Make it 100% compatible with PS5 (including patches / automatically downscaling from 2K / 4K for better image quality), discount the PSVR hardware to maybe £150 / 150 euros and promote it as a legit new way to enjoy games versus a simply graphics upgrade that will not be as visible.

I hope PS5 is 100% compatible with PSVR, but I also hope there will be a PSVR 2 with significantly higher resolution and wireless if 60GHz modems are cheap enough by then.

 

[RDGoodla]

If Nvidia offers Sony a banging GPU (17-19TF) at reasonable pricing, then yes, without a doubt no backwards compatibility. Remasters galore!

But this sh** ain't happening ...for so many reasons. Sony seems very content on offering "reasonable wares," without so much overhead and headaches. The new Sony isn't going to rock the boat on making their lives and developers lives harder. In this case, backwards compatibility will keep them competitive with Microsoft when the next-gen rolls around.

Why not a x86 CPU and a Nvidia GPU?

For example, AMD ZEN2 and NVIDIA GPU (even a discrete GPU).

Yes it may cost more than a single APU solution, but consumers may also accept a more expensive console with additional 30~40% performance.

Just like x1x vs. ps4pro.

(In best case if AMD+NVIDIA GPU only increase 50 dollars of next-gen price then the solution is worthy )

 

[Allandor]

Why not a x86 CPU and a Nvidia GPU?

For example, AMD ZEN2 and NVIDIA GPU (even a discrete GPU).

Yes it may cost more than a single APU solution, but consumers may also accept a more expensive console with additional 30~40% performance.

Just like x1x vs. ps4pro.

(In best case if AMD+NVIDIA GPU only increase 50 dollars of next-gen price then the solution is worthy )

because costs would go up and compatibility would get screwed.
And don't make the mistake to think nvidia chips would be that much better in a console. The AMD GPUs are really good if the code is optimized for them.

They will use a APU for the next iteration and only AMD or Intel (to expensive and the GPU-Part ... well we don't want to discuss this) could deliver a x86er APU. Well there is also VIA (or however they get called today) but they don't have something to even to compete with jaguar cores.
And than there is nvidia with it's tegra plattform. Choosing this plattform would only make sense if you want a total restart of consoles. And I don't think anyone wants that right now.

 

[Shifty Geezer]

Why not a x86 CPU and a Nvidia GPU?

For example, AMD ZEN2 and NVIDIA GPU (even a discrete GPU).

Yes it may cost more than a single APU solution, but consumers may also accept a more expensive console with additional 30~40% performance.

Why spend more for a 30% faster nVidia solution and not spend more on a larger, 30% faster AMD solution? That is, if the options are:

1. $400 console with AMD CPU + 300 mm^2 AMD GPU (SOC or discrete) at 12 TF
2. $500 console with AMD CPU + 300 mm^2 nVidia GPU at 15 TF
3. $500 console with AMD CPU + 390 mm^2 AMD GPU at 15 TF

1 and 3 provide the simplest business deals and probably best economies.

In short, the arguments in favour of nVidia only works if nVidia can be price competitive at a given (high end) performance, if they can offer something AMD can't. That's not true. AMD might run hotter maybe, but it's not like every nVidia card is 30% faster than AMD at the same price and same watts, is it? People are talking here like AMD is a lame dog and a weak booby-prize for the company that can't win nVidia over, and that's just a bit bonkers and seemingly the product of marketing.

 

[Deleted member 13524]

Why not a x86 CPU and a Nvidia GPU?

For example, AMD ZEN2 and NVIDIA GPU (even a discrete GPU).

Yes it may cost more than a single APU solution, but consumers may also accept a more expensive console with additional 30~40% performance.

For starters, because there's no proof a nvidia iGPU from a future architecture would perform 30-40% better than an AMD iGPU from a future architecture at ISO area, power consumption and process/foundry.

Raven Ridge's performance on gaming loads have put the Vega architecture in a completely different light regarding power efficiency.

 

[ultragpu]

I guess this has to be added in the equation now. I for one would totally be OK with a medium to large form factor PS5 if it means to accommodate the necessary TDP requirement for a 15TF APU. In the age of Slim model refreshes and midgen refreshes this could become even more feasible for a launch unit.

 

[Shortbread]

I guess this has to be added in the equation now. I for one would totally be OK with a medium to large form factor PS5 if it means to accommodate the necessary TDP requirement for a 15TF APU. In the age of Slim model refreshes and midgen refreshes this could become even more feasible for a launch unit.

The form factor doesn't necessarily have to be bigger or larger than a original PS4. Sony just needs implement a similar (or better) cooling method as used in the XB1-X (which is quite awesome).

 

[phoenix_chipset]

For starters, because there's no proof a nvidia iGPU from a future architecture would perform 30-40% better than an AMD iGPU from a future architecture at ISO area, power consumption and process/foundry.

Raven Ridge's performance on gaming loads have put the Vega architecture in a completely different light regarding power efficiency.

Yeah, Nvidia probably couldn't do any better than amd when it comes to desktop igpu. Iirc, polaris/vega is far more power efficient at lower clocks, to the point it seems it's not made with high clocks in mind. which is supported when comparing these cards clocks to Nvidias.

 

[DieH@rd]

What are chances that PS5 could switch to 100GB BD discs that are currently used for some 4K UHD movies? That would solve some storage issues [but it would elevate the need to place 1-2TB local storage drive as a base offering].

No matter how strong PS5 is or its exact timing, I am more concerned about the software. A lot of 1st/2nd party projects are slated to hit in 2018/2019, which puts majority of Sony's big guns out of contention for wowing us at PS5's launch.

 

[phoenix_chipset]

I guess this has to be added in the equation now. I for one would totally be OK with a medium to large form factor PS5 if it means to accommodate the necessary TDP requirement for a 15TF APU. In the age of Slim model refreshes and midgen refreshes this could become even more feasible for a launch unit.

It would be the biggest console ever made at 15tf or at least most power hungry. Assuming 2019 release. I mean, 12 is almost as much as the flagship vega 64, and I'll be extremely impressed if the next mid range card can match that. And even if it could, it will likely have cu's disabled and lower clocks for a console. There's nothing to suggest Sony and Ms will aim higher than mid range cards as they've done with their base consoles and mid gen refreshes.