How Microsoft attempted to make the Xbox 360 dashboard load faster

Re:How Microsoft attempted to make the Xbox 360 dashboard lo... by mariusmg on Tuesday January 10, 2023 @12:56PM (#34328748)

Am I the only one which expected that loading from NAND would have been faster than HDD ?

[ Parent | on HN ]

Re:How Microsoft attempted to make the Xbox 360 dashboard lo... by wtallis on Tuesday January 10, 2023 @01:32PM (#34329292)

One NAND chip is slow. NAND-based SSDs can only deliver high throughput by using many NAND chips in parallel. Additionally, the console's 16MB NAND chip is a far cry from today's 64GB chips used in mainstream SSDs.

[ Parent | on HN ]

Re:How Microsoft attempted to make the Xbox 360 dashboard lo... by RyanShook on Tuesday January 10, 2023 @01:17PM (#34329069)

It seems like the slight speed advantage was in uncompressing the files beforehand on the HDD essentially skipping a step that had to be performed on every boot from NAND.

[ Parent | on HN ]

Re:How Microsoft attempted to make the Xbox 360 dashboard lo... by masklinn on Tuesday January 10, 2023 @01:41PM (#34329425)

Seems unlikely, file decompression tends to be quite fast (faster than even relatively fast IO if they're using lz4, but even more expensive compression methods would be faster than an HDD, to say nothing about apparently slow NAND)

[ Parent | on HN ]

Re:How Microsoft attempted to make the Xbox 360 dashboard lo... by TonyTrapp on Tuesday January 10, 2023 @02:09PM (#34329873)

A 7200 RPM drive can easily outperform old flash memory. It wasn't always the case that flash was the fastest storage medium. Flash performance used to be measured in KB/s, not GB/s like on modern NVMe drives.

[ Parent | on HN ]