Today's systems have diverse needs that are difficult to address using one-size-fits-all commodity DRAM. Unfortunately, although system designers can theoretically adapt commodity DRAM chips to meet their particular design goals (e.g., by reducing access timings to improve performance, implementing system-level RowHammer mitigations), we observe that designers today lack sufficient insight into commodity DRAM chips' reliability characteristics to implement these techniques in practice. In this work, we make a case for DRAM manufacturers to provide increased transparency into key aspects of DRAM reliability (e.g., basic chip design properties, testing strategies). Doing so enables system designers to make informed decisions to better adapt commodity DRAM to meet modern systems' needs while preserving its cost advantages. To support our argument, we study four ways that system designers can adapt commodity DRAM chips to system-specific design goals: (1) improving DRAM reliability; (2) reducing DRAM refresh overheads; (3) reducing DRAM access latency; and (4) mitigating RowHammer attacks. We observe that adopting solutions for any of the four goals requires system designers to make assumptions about a DRAM chip's reliability characteristics. These assumptions discourage system designers from using such solutions in practice due to the difficulty of both making and relying upon the assumption. We identify DRAM standards as the root of the problem: current standards rigidly enforce a fixed operating point with no specifications for how a system designer might explore alternative operating points. To overcome this problem, we introduce a two-step approach that reevaluates DRAM standards with a focus on transparency of DRAM reliability so that system designers are encouraged to make the most of commodity DRAM technology for both current and future DRAM chips.
翻译:不幸的是,尽管系统设计师在理论上能够调整商品 DRAM芯片,以更好地调整商品 DRAM 以适应现代系统的需求,同时维护其成本优势。 为了支持我们的论点,我们研究系统设计师可以使商品 DRAM 芯片适应系统特定设计目标的四种方法:(1) 改进DRAM可靠性;(2) 降低DRAM刷新管理费;(3) 减少DRAM访问问题拉通性;(4) 减轻RawHammer攻击。我们发现,为这四个目标中任何一个目标采用的解决办法需要假设DRAM 的当前操作系统,以便更好地适应现代系统的需求,同时保持其成本优势。为了支持我们的论点,我们研究系统设计师可以使商品 DRAM 芯片适应系统具体设计目标的四种方法:(1) 提高DRAM 可靠性;(2) 降低DRAM 更新管理费;(3) 减少DRAM 问题拉通性;(4) 减缓RawHammer攻击。我们发现,为这四个目标中的任何一个目标采用的解决办法都需要有两种不同的解决方案: DRAM CD RRAM 设定一个当前运行标准,同时采用一种最可靠的标准。