#PDF slides on #FDP, "flexible data placement" to reduce write amplification on #Flash #storage. Apparently this is the thing they're pushing to replace #zoned-storage? It has a stateless API (unlike ZNS on Flash) but requires feedback to guarantee WAF = 1.
on 02025-10-21"f2fs" is the #Flash friendly #filesystem for #zoned-storage. “F2FS is a file system exploiting NAND flash memory-based storage devices, which is based on Log-structured File System (#LFS). The design has been focused on addressing the fundamental issues in LFS, which are snowball effect of wandering tree and high cleaning overhead.”
on 02025-08-24“Zones in #zoned-storage devices must be written sequentially. This is called the sequential write constraint. Each zone in the device address space has a write pointer that keeps track of the position of the next write. Data in a zone cannot be directly overwritten: before being overwritten, the zone must first be erased using a special command (zone reset).” #SSD #Flash #SMR
on 02025-08-21#ZUFS #Flash storage is some sort of vague thing that makes your Flash last longer: “Zoned UFS”
on 02025-08-21#paper on #real-time #garbage-collection of #Flash
on 02016-10-10#paper on #memory #performance using “hybrid main memory” such as a DRAM cache layer over #Flash, as a stopgap until phase-change memory or something like that arrives.
on 02016-10-10#paper on #filesystems with high #performance on NAND #Flash by using multiple channels
on 02016-10-10NOR #Flash transactional #filesystems #paper
on 02016-10-10#hardware for reading FT2232H NAND #Flash for #reverse-engineering
on 02016-10-10“Non-volatile storage: Implications of the Datacenter’s Shifting Center”, an ACM Queue #paper claiming that now #Flash “storage-class memories” of US$3000 to US$5000, such as NVDIMMs, provide durable #memory #hardware that is fast enough to saturate one or more fast CPUs, thus invalidating the assumptions that have underlain our memory hierarchy designs in computer systems ever since semiconductor RAM beat out ferrite cores in the early 1970s. They still only support about 100k IOPS though (or 400k for sequential or read-only workloads). and they still cost 25× as much as spinning rust #disks. (The massive Flash storage speed advantage is basically how #Fastly works.) It argues that now optimizing cost-performance requires us to be willing to let CPUs idle in order to keep the Flash beast fed, not vice versa, since the cost of the Flash rather than the CPU now drives the overall cost of the system. So, for example, maybe you should busywait on your Flash instead of using interrupts.
on 02016-01-06