Yes, this is one chip but it is just half the speed of the PS5 solution.
Even more chips would lead to an even larger interface and larger controllerchip and space requirements on the board.
So I guess there must be a certain point where chips vs error rate vs latency vs interface width vs space requirements vs cost is at the sweet spot.
But my whole point is, now that they even support 5.5gb/s SSDs which should be slower than the internal SSD they are open to support just this in the future to save some costs as the prices for these will go down.
And I am really certain that a slower drive wouldn't change much at best loading screens might be a bit more prominent. Should affect the streaming capabilities as with random access reads the bandwidth will go down and at that point even the read rate of a bit slower and faster SSD should not be that different.
Right, the cheapest method of increasing bandwidth is by increasing the number of chips, versus building a single large chip to do it. No doubts, but in the reverse it's the most expensive way to increase storage. So the discussion about cost in this case has to take into account both. This is like saying everything should be esram or embedded L3 cache because of how much bandwidth it generates, without taking into account how little storage it provides. Both need to be part of the discussion which is what I wanted to point out. The claim I wanted to debate is that what Sony did with PS5 changed the SSD market, and now everyone is following suit. On that particular claim, I would lean that PS5 took advantage of having the realestate to support 6 chips in a surface area much larger than the largest M.2 slot, and access significantly better cooling, such that it made more sense for them to roll their own solution while hard drive manufacturers still needed to figure out how to make it work in a stick format.
I'm fairly certain PS5's internal storage costs are going to be locked for a while at 875 GB until they can secure getting 6x of the next largest size (the current memory size of their chips doesn't seem very popular either). Sony will rely and market their external SSD M2 expansion method as the way to increase storage.
MS should be in an 'ok' position with respect to bringing the cost of the storage down. Their current SSD setup is really just an off the shelf component that fits into that specific m.2 form factor, nothing really more. Like how everyone here is all onboard for 3P SSD support on PS5 looking for a better deal, MS is also looking for a better deal for their internal drives for their Series consoles; the ssd come with controllers and memory chips, swapping them out for another manufacturer is just a matter of ensuring requirements and tolerances are met.
We are very likely soon to see a Series S console at 1TB before either XSX or PS5 move to anything larger.
TLDR; on the subject of cost, since the main debate is around 3P SSD drives being faster and cheaper than the proprietary drive of MS; you should consider carrying that through to the total cost of the system as well. While there is solid guarantee that 3P nvme drives will be much faster, there is no guarantee it will be much cheaper given the chip differential.
Currently the largest mobile phones use 512GB chip for storage, by next generation they will be at 1TB. And as time goes on 1TB chips should be very popular. That means the price of the console and the price of the expansion card should come down together granting larger savings than what is being discussed here. But there is fair commentary that 'today', there is no reason to have shipped a 1TB chip, when perhaps you could have gotten away with 2x512GB chips and still get to 2500MB read speeds for cheaper and those may have been able to fit on the 2240 form factor -- 4x256 would have required chips on both sides which its clear it won't fit.
