While solid state drives (SSDs) offer superior read speeds, their design introduces specific performance caveats. First, SSDs store data in blocks (typically $256$ KB or larger) that must be read, erased, and rewritten as a whole, resulting in poor performance for bit-wise random writes. Although modern SSD controllers and firmware have developed algorithms to mitigate this, write speeds can still be much slower, in particular for QLC (quad level cell) SSDs. To optimize write speeds, it is necessary to maintain a queue of operations, prefer reads, and write in large blocks. Second, flash memory cells degrade over time, often wearing out after a few thousand write cycles. While wear-leveling algorithms help spread this degradation, SSDs are not recommended for high-frequency write tasks like swapping files or continuously aggregating large log files. Finally, the massive increase in bandwidth has forced computer designers to attach SSDs directly to the PCIe bus. The drives capable of handling this, referred to as NVMe (Non Volatile Memory enhanced), can use up to $4$ PCIe lanes, achieving up to $8$ GB/s on PCIe $4.0$.