Sony share price dropping, incorrectly, based on ACTUAL market positions of some of the ‘big’ console players....
Jan 19, 2022 at 10:22 PM Thread Starter Post #1 of 4

whitedragem

500+ Head-Fier
Joined
Feb 26, 2010
Posts
619
Likes
631
G’day (from Australia)

My name is Rene, and I have ‘a few things’ to share on the (market)state of play regarding what will actually play out in the next few years of ‘console warring’.

After hearing the market response to Microsofts’ (NEED) acquiring Activision, it is important that we raise awareness as to ‘what is actually going on’ and so I quickly threw up a video.
Literally ‘threw it up’; it is unscripted and ‘rant-y’, and only touches the surface of a few of the issues (and ends abruptly), but it does bring some ‘reality check’ to the present situation.

Not really trying to start a channel, but these two vids shed some light.. , and https://youtu.be/95VZDKvz8gw

I’d rather not ‘take a stance’ but gaming is BIG BUSINESS and one of the BIGGEST BUSINESSES has started a ‘ball rolling’ to an unsustainable gaming market (in terms of bringing quality products to market in a profitable way to developers).

Sony and Nintendo platforms are what we need if creativity and quality is to remain in console software products; where ‘multiple SKUs’ generally means the ONE ADVANTAGE THE CONSOLE HAS (“optimised, ‘to the metal‘ performance“ squeezing out every last iota of power from the ‘cheap’ console boxes) which happens when there is a set hardware platform that a developer can optimise for.

Given the lion share of console hardware (and software) sales are NOT in Microsofts’ court, the handful(+) of xbox console SKUs that exist will rarely be the lead development platform, and given their ‘weak install base’, will generaly be given straight ports of whichever version launched on Playstation. (like when the XBOX X basically got PS4 Pro games with ‘minor changes to make use of the VASTLY better hardware’)

Due to the PS5 having the VASTLY better hardware (for console gaming), this generation will require Microsoft to release MORE SKUs to survive (a percentage of the customer base will buy based on spec sheets, which are easy to manipulate to look ‘better’).

Anyhow, starting a topic.. have a lot to say.. but wanted to get some awareness out there, cause the media is running in the WRONG DIRECTION with this (and hence misleading us).
 
Jan 21, 2022 at 5:20 PM Post #3 of 4
Good exclusive games has been the draw to choose between Xbox and PS for me. I always choose PS due to the exclusives. At both hardware performances, what's in the library matters more.

I recently got a PS5 because Xbox or PC doesn't have Demon's Souls remake and Bloodborne.
 
Jan 22, 2022 at 12:18 AM Post #4 of 4
customised RDNA (part 1), and ‘real world capabilities’


I did a little ‘informal’ benchmarking this week, of the console (Playstation 5) vs (standard) PC architecture platform; being a ‘great example’ -an i7 5820k is X99 platform and has ‘quad channel ram access’ (ie not a ram bottleneck) using the venerable GTX2070, generally considered ~ the equivalent nvidia card to what the AMD graphics chip in the PS5 would render, in a typical PC. (the Nvidia card is better at Nvidias version of raytracing, but not always quite so strong at the stuff that all games likely depend upon)

Game used is five years+ old, “Planet Coaster”, and whilst it ‘likes’ an Intel i7 4770k for ultimate quality mode, and a i7 5820k typically run more than a 1Ghz LESS (per core)- the issues I was testing had little to do with CPU speed, so wasn’t an issue.

What I had seen, from online clips, was that the ‘wooden coasters’ had a lot of vertices, and would grind the GTX 2070 (using medium games settings @ 1920x1080 resolution) a large challenge.
This isn’t unusual, and is why PowerVR championed, 20 years ago, a ‘tile based’ deferred renderer, that could pre cull graphics information that wouldn’t turn up in the final frame on the ‘users screen’ from the workload to be processed.
These techniques weren’t amazing enough, 20 years ago, to get a fourth place graphics company (VS nvidia/ati(AMD)/3DFX, and the much less serious competition to be found from Matrox Graphics) into everyones’ households; but it has been beautifully implemented into ‘low powered devices’, such as iPads, to get a lot more capable graphics output than what the thermal envelope/total power draw allows in typical graphics tasks.

Sony (and Cerny) had input from Epic games (who have that little thing we know as ‘unreal engine’ at their disposal), and re-engineered the AMD ‘RDNA architecture’ to do things that gamers could actually benefit from (vs PURE RDNA chipset that has some standard functionality that is practical for developers/hollywood perhaps, and not so much ‘gamers’).
Sonys’ highly customised RDNA is the direction that RDNA 3 went in, netting around a 40% improvement over RDNA2, due to ‘early precull‘ not sending useless information down the chain, needlessly burdening the system with ‘overhead’ that could have been ‘cut away’.

When this is in effect, the benefit is massive.
Less useful when working with non complex scenery, but when we go ‘natural’ and render lots of trees (huge number of ‘triangles’ to make look NATURAL), the culling of ‘offscreen’ data from the pipeline nets VERY SIGNIFICANT FRAMERATE IMPROVEMENTS.

The best example of this, I have seen, was rendering lots and lots of ‘wooden coasters’ (with huge timber structures underneath them) and panning around the screen and rotating the view port/ doing ‘fly bys’ etc..
Running at a much higher detail level, at a much higher resolution, I was able to avoid slowdowns that even a handful of structures could bring to a ‘standard PC architecture’.

Anyone championing ‘full RDNA 2’ chipset- get your facts right; this isn’t a good thing for anything other than lowering cost and time to bring a console to the market.
Yes, it COULD help microsoft a little by standardising their console to be like ‘standard PC hardware’ (not a good thing for a console generally), and allows their developers to do a lot less to optimise software (an oxymoronic notion in a MS ecosystem) to make use of hardware better across many SKUs (PCs/XBOX Series X/XBOX Series S/XBOX ONE X/XBOX ONE S/XBOX ONE etc).

The highly customised APU in the Playstation 5 ran rings around ‘an expensive PC platform’ in a way a title shouldn’t.
For the record, Frontier Games, who make the Cobra engine (chapioned for its excellent multiplatform scalability) haven‘t had a hard time making their PS4 game work on PS5 (mostly just allowing the larger park sizes to make use of more CPU and RAM that the new generation consoles have), and the game is ‘far from being‘ a second generation title, that generally squeeze better use out of the console hardwares’ maximum capabilities- which generally requires LONGER CONSOLE GENERATIONS and LESS SKUs (such as PS4/PS4 Pro and then PS5(with or without a drive doesn’t effect developers AT ALL)).

The fact is the PS5 basically auto optimises itself (using the early z cull techinique) and will have an easy time with vegetation heavy scenes and titles that render a lot of ‘off screen data’ (eg levels not cleverly engineered based on limitations of a SPECIFIC HARDWARE implementation) (or anything NOT USING UNREAL ENGINE 5s’ NANITE technology).
I was surprised just how well Sonys’ customised RDNA did, due to clever (pre)culling, on the vertex intensive components found in the frame...
This was an example where the PS5 smashed a GTX2070 so hard that it is literally unbelievable.

Titles optimised for PS5 will literally blow people away. Second gen titles even more so...
Second gen titles don’t seem to happen much on the XBOX side of the coin (mostly due to Microsoft handling of software houses and planning, but also due to ‘too many SKUs’), and the optimisation that goes into console titles, to use implemented hardware fully, is where consoles prove GREAT (cheap) vs their counterparts (PCs, and Xboxs). (I say XBoxs there very ‘tongue in cheek’, cause they are generally relatively standard PC architecture- great for previous generation backwards compatibility and ‘cross platform‘ (with PC) and gave them super easy ability to do ‘backwards compatibility’ as a feature (the hardware pretty much ‘the same, but quicker than last years model’)).

Of the informal PLANET COASTER hardware test I just spoke about (testing preculling).. it has nothing to do with other aspects of the Playstation 5 hardware that actually will not be done on PC for a few years yet- certainly if the install base of some ‘future PC hardware’ needs to hit ‘double digit percentage’ of users before any niche software is written to use said features; such as the Playstation 5s’ customised layout, injecting the insanely fast SSD bus directly to demanding hardware (which will not show much use until second gen titles write themselves knowing how to make use of the insane ’never before seen’ (*ANYPLATFORM) throughput/capability.

The Neo Geo console, in the early nineties, had a massive data bus that allowed that console to still be ‘relevant’ (certainly to arcades) over a decade later, BECAUSE IT WAS SO FAR AHEAD OF THE MARKET ‘at the time’.
PS5 has this going for it. Many heads of industry and ‘top tier’ journalists acknowledging that it will be close to the year “2030” (on ‘pre covid’ trajectories) before the PS5 feature set is found in the wider market.
It revolutionises the industry, and whilst the Matrix 5 demo showcases that “something is afoot”, it will be a little while longer before we see more come from the EPIC/Sony collaboration, and more so, the huge amount of independent studios and developers who write for profitable console ecosystems (Sony and Nintendo).

A few extra first party studios, especially the way Microsoft utilises them (never given ONE SKU to optimise for, with second and third wave titles to ‘perfect’ actual hardware utilisation) will not close the advantage that the WHOLE MARKET HAS, when writing for Playstation hardware first.
Small indies not focused on graphics go for Nintendo Switch (user base size and ‘software attachment’ rate ‘per console’); and anyone wanting to push graphics will write for PS5!

The market has fallen back to where we were a few years ago:
Late in Virtual Realitys’ development (of which there were about five -six majors players at the height of the race to “home VR”), it was discovered that for Augemented Reality and avoiding motion sickness in normal VR; the update rate of the displays needed to go to ‘really high refresh rates’ (so the ingame direction matched the one you were facing, and the movement you took to ‘get there’ moved as quickly on the screen(s) as the player experienced the movement).
VR already required rendering a huge overhead vs regular ‘one screen gaming’ (eyes are in ‘stereo’ and required slightly altered perspective), and Sony quickly realised that the base PS4 system wouldn’t have the graphics chops to render games in 90-120hz framerates (x2 screens), so the need for a PS4 Pro was born.
Sony didn’t want to split the console market into multiple SKUs (which kills ability to optimise games for specific hardware (which is a consoles MAIN STRENGTH), or makes it more costly and more time consuming to achieve quality optimised output)) and so put some ‘pro consumer’ policies in place (games have to be written for PS4, and ‘no features‘ special to the PS4 Pro version etc), to encourage gamers to only buy the PS4 Pro version if they wanted the VR.
The market didn’t quite follow suit- a lot of gamers just wanted the ‘faster hardware’ (spec sheet bragging rights and, eventually ‘last generation games working noticably better on the faster hardware’, gave incentives for many NON-VR users to ‘want one’), but the PS4 Pro was for VR to work as planned...
Sony stopped producing the PS4 Pro when the PS5 was coming to market, as the PS5 was going to be the ‘better hardware to push VR forward’ (that early Z cull ”optimsied silicon” will do foveated rendering exceptionally well). Sony still make the PS4 (base edition) console as the market is stupidly strong for the PS4 (well over 100 million consoles SOLD).
Anyhow the VR requirement for a PS4 PRo model has me digress, but we understand the ‘why’ for what actually happened regarding ‘Sony making another SKU’ a little better-

The market at the time was filled with two major competitors (sony and microsoft) with the PS4 being roughtly 40% more grunty (than the base Xbox). this had many early multiplatform titles on xbox running at 20% less resolution @ 20% less framerate.
The market responded by declaring the PS4 the ‘sales champion’, and xbox lost a lot of their market.
As soon as Microsoft had figured out their console hardware was ‘less’ than the competition, they pulled every trick to close the gap. At the time of bringing the XBOX ONE to market, they pulled a few 11th hour moves. They removed kinect from the design (something they had promised developers would be included with EVERY CONSOLE, hence ‘write games to use said features’) as roughly a 10% cpu perf hit would be needed for the ‘Kinect 2’ improved bone mapping feature.
They also upspecced (overclocked?) the CPU a bit... (never let a massive fail rate of consoles, eg the infamous Red Ring of Death, get in the way of beating a competitor to market: consumers want ‘bragging rights’ aka ‘bigger spec numbers’, and standing in line returning failing hardware isn’t a problem if it makes the company money in the long run...)(Miscoroft has ALWAYS been willing to spend a lot of money in order to make more money)
The XBOX ONE S was the result and it closed down the gap, a little, between the base PS4 and the original Xbox One.
But they had seen what a 40% advantage could do to swinging ‘fence sitters’ and ‘regular punters’ into an ecosystem.
It was likely ‘no accident’ that the response to the announcement of a PS4 Pro was to design a console (pretty much) 40% more powerful than it, in a hope to ‘bring the punters back’.
This might have actually worked if the amount of studios that actually write software for the Sony Console wasn’t SO HUGE.

Studios would write the software for the PS4 (lionshare of market, and a requirement by Sony instead of just making PS4 Pro software), and then an enhanced version of the software was generally made for the PS4 Pro console.
When bringing the games to ‘competitors platforms’, very little work/polish was generally done- a deprecated PS4 version went to the Xbox One (sometimes was the same version but just ‘ran worse’), and the PS4 Pro version was generally ported, poorly, to the Xbox One X (not enough user base to demand reworking the game to make use of the extra grunt found in the console hardware).

After seeing this play out for awhile, Microsoft must have been annoyed. Very few, if anyone outside of first party studios, were optimising for the new ‘super console’.
To change this around they started buying software studios to work specifically on software for the Xbox One X console- to have it actually show itself as ‘more powerful than a PS4 Pro’.

When I alluded, earlier in this ‘stream of consciousness’ piece of prose (I wish), that the market is in a ‘somewhat similar state’ this generation, in this regard it ‘somewhat is’.
Games will be written for PS5 first (by MAJORITY OF THE WORLDS software houses). (hopefully they don’t lean to heavy into vegetation rich stage design :wink: ), with a port then sent off to the Xbox side of the coin.
Reworking the game for two very different SKUs (series S vs series X), the games are unlikely to ever be ‘well optimised’ (unless Unreal Engine 5 which is natively ‘very generous’ to whatever paultry hardware a user has, and scales UP ‘very well’ to use better system capabilities).
It is fair to say that the ‘made for spec sheet looking good‘ “paint by numbers” approach Microsoft bring to console hardware design will rarely ever show a realworld advantage- UNLESS:
Microsoft BUY STUDIOS and have them optimise for a specific SKU. Given they cannot do this (too many SKUs) the best they can do is buy software companies (they become ‘first party’), and then dictate to the studio heads to ‘design levels/features to focus on strengths of whichever product Microsoft is trying to sell at the time. (ie likely to generate profit in a gamespass model, so ‘microtransactions’?!)

If we look at the latest Rainbow Six game out the install size on Playstation 5 is basically identical to that of the other platforms. We know if they had used the Sony tools for file packing (to use the SSD effectively) the install size would be ‘much smaller’ that the other platforms. At which point the loading speeds would become NOTICABLY BETTER. (they are still 50% faster, but this isn’t what I am suggesting here- more about ‘how do you optimise a game to look worse on the platform that naturally gives better performance: little tricks like running anisotropic aliasing higher on the PS5 (we’ve seen this on a few cross gen titles that pull ’only equal performance’ between the consoles’ (or a slight lead to the PS5).

The reality is the PS5 video chip, due to the ‘early Z cull’ will often net 20-40% greater performance than the ‘specifications’ alone imply.
Expect to see a lot of ‘mixed results’ in real world products.
Either platform will have ‘better performance’ (per title) entirely dependant on what the title does.
Given majority of software houses write for the better selling hardware, presently Nintendo and Playstation (due to previous console experience and/or profitability when selling software in the console ecosystem), most of what we have seen in early titles will play out in UNOPTIMISED SOFTWARE.
In unoptimised software there will be MANY MOMENTS where for some ‘unknown’ :wink: reason (cough-=customised RDNA=-cough), the PS5 will appear to be 20% more capable than the Xbox series X.
Then of course we will find developers, who, for whatever reason, favour one console manufacturer and somehow put out software that runs POORLY on the other console (eg the PS5 load times in ‘the medium’ -this takes effort to gimp* a product to perform like this)
*gimp= get in mediocre performance

In the past ‘first party studios‘ were for having software ready for console launch. Due to having their first title ‘out the gate’ so early in the consoles’ life, they are usually good to deliver ‘highly optimised’ software in second and third wave titles... once they have learnt the silicon better, and can squeeze better tricks from it (either closer to 100% hardware utilisation or better game design allowing more organic ‘suspension of disbelief’, where the console fools us it is doing things beyond its capabilities).

The PS5 in a straight race against the xbox series x, will generally prove better, and then once we see actual software designed for it..... (noone knows just how high it will truly punch, but if the cost price suggests its’ “weight class’, then it will be a great fight to watch as the ‘underdog’ competitor (without deep pockets to ‘spend from other business divisions) keeps gracefully outclassing the “20% bigger competitor” (by manipulated representation ‘looking at the spec sheet’, rather than ‘actual size’).

What is clear is that Microsoft cannot move the entire gaming market into its’ camp, so there will always be more software on Playstation showing off the improved silicon. in this regard ‘first parties’ mean much less to the battle than the press might suggest. (especially when Microsofts‘ first parties often skip whole SKUs or fail to utilise a hardware generation etc.. and will likely be focusing on retuning software to turn profit in a gamespass model, that consumers are hesitant to let grow in subscription cost. Rather than get first party developers to focus on quality, focused, output; that on other platforms, ‘sans gamespass‘ that generate actual profit for the platform holder via actual software unit sales with superlative OPTIMISED software of high quality nature

tl: dr
customised RDNA greater than ‘off the shelf’ standard RDNA (unless you are in marketing)
 
Last edited:

Users who are viewing this thread

Back
Top