Which will prevail? Xbox Power or Playstation Speed?

[DonFerrari]

Pemalite said:

BraLoD said:

Blast Processing is a myth, it doesn't exist and was a marketing stunt, even if the Mega Drive was indeed faster.

Blast Processing was actually real, it is what happens when the marketing team gets a-hold of a technical specification and runs with it.

Essentially Blast Processing was actually the DMA unit in the VDP graphics processor which expedited memory transfers... It was actually an underutilized feature, but it was there.

sales2099 said:
The more I go outside Vgchartz and see what the internet thinks of the specs between the two....the more I’m convinced multiplats will have a resolution and or FPS disparity between PS5/series x.

The sentiment is that the ssd overall can’t compensate for a weaker cpu and GPU. Too many factors favor the series x for the ssd to make up the gap. And the series x also has a ssd...I mean the only way I can believe the ssd clinging is if Xbox didn’t have one.

Storage subsystems are never going to compensate for a weaker CPU or GPU... Storage generally doesn't do any processing to aid those scenarios.

What a fast storage system can do is compensate for RAM constrained environments.

The Xbox Series X has the technical edge, that cannot be disputed, by how much? We need to wait for the games to see the real-world implications, for simpler games (Platformers etc'), ports from older systems (Xbox One, Playstation 4) the experience will absolutely be identical.

src said:

People need to understand: the entire gaming pipeline is loading data. Everything you see on screen is calculated by your GPU/CPU which loads data from memory.

There are two bottlenecks: how quickly you can load the data and how much you can calculate.

PS5's SSD is a 200% increase over XSX. 

Cerny explained it best. Usually a game loads data from the CD/HDD onto the RAM. Because this loading takes lets say 10 seconds, the game needs have all the data for the next 10 seconds of gameplay on the RAM so you can have stuff coming on the screen while the next batch of data is being loaded onto RAM. With faster SSD, the same amount of RAM can have fewer seconds of gameplay: 8GB of RAM can be used to store 5 seconds of data instead of 10 seconds. This means each second of gameplay can have more data in it, aka more fidelity (such as more animation, larger levels, AI, etc).

It's impossible to say how much the SSD difference means because it's so specific to each game and how it's coded. 

Keep that optimism going.

src said:

Its true, SSD allows a faster data stream into RAM but its the GPU crunching that produces the frames per second. However whats also true is that the CPU/GPU differences are much much smaller than PS4 and XB1. Furthermore there are architectural differences that are still not fully detailed (different compression techniques, additional cores for certain tasks like audio) that could change things.

In summary the FPS and resolution difference should be small. Meanwhile the SSD speed advantage the PS5 is massive. Will the gameplay designs be noticeably different? We will have to wait to see the games.

We don't know every aspect of the GPU yet, the differences could be larger than the current specifications entail.

Ray Tracing is the next-gen buzzword feature, the SSD helps... But the SSD isn't doing all the processing to showcase the pretty pictures on your display... And the Xbox Series X will potentially have additional functional units to possibly handle Ray Tracing better.

chakkra said:

For the life of me I do not understand how Mark Cerny managed to convince people of this BS when we have decades of graphics cards comparisons that prove quite the opposite.

Bigger GPUs with more CUs have ALWAYS performed better than smaller GPUs with higher Clock speeds. ALWAYS.

Not true. Vega 7 has 60 CU's vs Vega 64's, 64 CU's. Vega 7 is significantly faster due to it's bandwidth and clockrate advantages.

DonFerrari said:

Digital Foundry didn't disagree on Mark Cerny.

Also what reason would Sony have to choose less CUs with higher frequency (and much higher than what could be expected, and one that makes dies harder to make and cooling also harder to achieve)? Just for the giggles? I hope you don't come with a they decided to put the boost last week because of Xbox being much stronger or "because they are dumb".

Sony made the best decision for Sony and hitting their goals.

There are two ways to bolstering GPU performance... More functional units or higher clockrates. (If all else is kept equal of course!)

Both have their Pro's and Con's... And contrary to popular belief a smaller chip with higher clockrates isn't always cheaper to manufacture... If a process suffers from terrible power characteristics and the chip is extremely leaky, then yields can be lower than a larger chip.

In short, it's a balancing act... And Sony made the right choice for Sony... And instead put more engineering into the storage subsystems.

Both consoles will be fantastic pieces of kit.

chakkra said:

And there was one particular comment they did that people either missed or chose to ignore: " It's a fascinating idea - and entirely at odds with Microsoft's design decisions for Xbox Series X - and what this likely means is that developers will need to be mindful of potential power consumption spikes that could impact clocks and lower performance."

Microsoft has already confirmed that the Xbox Series X can maintain it's clockrates without throttling in it's reveal.

Which is in stark contrast to Sony as it is relying on mobile technology to share thermal/power.

Drakrami said:

So 6 vs 4.2 Tflops didn't make a difference this gen. What makes you think 12 vs 10 Tflops will make any difference next gen? Most of us throwing around the term Tflop don't even understand half of what the term means. 

There was a massive difference. Just not all developers decided to leverage the differences to the fullest extent... And many differences weren't even down to the computational capabilities... But in part thanks to the increased CPU and RAM capabilities.

Kyuu said:

The general prediction from PS fans was and still is that XSX would be the more capable platform for obvious reasons. But those few unusual and unexpected advantages may play their role. There also seems to be a few missing factors Sony may announce/explain that might reduce that gap further and maybe even give PS5 the upper hand if fully harnessed by 1st party studios. It'll be like X360 vs PS3 in a sense.

It will be nothing like the Xbox 360 vs Playstation 3 scenario.
The Playstation 3 had a multitude of technological advantages, not just one.

BraLoD said:

The Switch success is based on games.

They launched the system with one of the most praised games ever and withing that same year they got another big app killer.

Do you know why the PS4 sold so well at launch?

The Last of Us.

And it wasn't even available there at launch!

Any consoles success isn't based on a singular aspect, that is a 2 dimensional way of looking at things.

The Switch was successful because it was portable, introduced new novel concepts and had amazing games... Even if it's best games are WiiU titles... All at the right price.

The Playstation 4 was successful because it was cheaper and more powerful and had amazing games.

Yep you got what I said, if for the same budget they could get higher CU or higher frequency and chose the second it isn't because they are dumb, but because they thought it was the best decision for their design.

Regarding the cost I have a hard time believing PS5 and XSX cost the same to manufacture. Because if they do and Xbox ended up 15-20% more powerful then yes Sony made a worse decision than MS. Unless the other features are that much better (I don't really think so). So yes I expect at least 50 less to build up to 100.

chakkra said:

DonFerrari said:

At the moment of the design they choose less CUs with higher clock speeds. That have nothing to do with MS. So please explain that if it is impossible to get better result by higher frequency lesser CU count why would they choose these instead of more CUs and lower frequency?

Errr... MS and Sony's decisions are ALWAYS influenced by what the other does or might do. Even if they publicly deny it, I will never believe otherwise. BUT, like I said before, at the moment of designing the console neither of them knew how many CUs and clock the other was going to use, so they both had to make a decision and shoot in the dark, and hope for the best.

Also you are forgetting another VERY important aspect of the decision making process, price. Companies design their products around a certain price point they are trying to meet. If you are planning to sell your device at $399-449 then you have to choose the parts that fit into that price range. Same if you are planning to sell it at $499-549.

Now, we don't really know what price points both MS and Sony were trying to hit but we do have MS on record saying that they were not going to make the same mistake of releasing a weaker console, so at least we know their INTENTION was to shoot as high as possible when it comes to performance. On the other hand, we also have Sony on record saying that they were planning to make PS5 "affordable" so I think is fair to assume that pricing influenced their final decision.

Yes it is influenced by what others might do, but wasn't by what others done.

So if they decided to have higher frequency than CU count it was because they believed it would net the best results for PS5 design.

And about price yes I can believe they gone for something like "to reach 10.2Tflop we could have that using this much CUs at this frequency costing this much or we could save 50 bucks going for higher frequency and lower CUs while accepting it will have some weakness"