Paper 2018/205

Static-Memory-Hard Functions, and Modeling the Cost of Space vs. Time

Thaddeus Dryja, Quanquan C. Liu, and Sunoo Park


A series of recent research starting with (Alwen and Serbinenko, STOC 2015) has deepened our understanding of the notion of memory-hardness in cryptography — a useful property of hash functions for deterring large-scale password-cracking attacks — and has shown memory-hardness to have intricate connections with the theory of graph pebbling. Definitions of memory-hardness are not yet unified in the somewhat nascent field of memory-hardness, however, and the guarantees proven to date are with respect to a range of proposed definitions. In this paper, we observe two significant and practical considerations that are not analyzed by existing models of memory-hardness, and propose new models to capture them, accompanied by constructions based on new hard-to-pebble graphs. Our contribution is two-fold, as follows. First, existing measures of memory-hardness only account for dynamic memory usage (i.e., memory read/written at runtime), and do not consider static memory usage (e.g., memory on disk). Among other things, this means that memory requirements considered by prior models are inherently upper-bounded by a hash function’s runtime; in contrast, counting static memory would potentially allow quantification of much larger memory requirements, decoupled from runtime. We propose a new definition of static-memory-hard function (SHF) which takes static memory into account: we model static memory usage by oracle access to a large preprocessed string, which may be considered part of the hash function description. Static memory requirements are complementary to dynamic memory requirements: neither can replace the other, and to deter large-scale password-cracking attacks, a hash function will benefit from being both dynamic memory-hard and static-memory-hard. We give two SHF constructions based on pebbling. To prove static-memory-hardness, we define a new pebble game (“black-magic pebble game”), and new graph constructions with optimal complexity under our proposed measure. Moreover, we provide a prototype implementation of our first SHF construction (which is based on pebbling of a simple “cylinder” graph), providing an initial demonstration of practical feasibility for a limited range of parameter settings. Secondly, existing memory-hardness models implicitly assume that the cost of space and time are more or less on par: they consider only linear ratios between the costs of time and space. We propose a new model to capture nonlinear time-space trade-offs: e.g., how is the adversary impacted when space is quadratically more expensive than time? We prove that nonlinear tradeoffs can in fact cause adversaries to employ different strategies from linear tradeoffs. Finally, as an additional contribution of independent interest, we present an asymptotically tight graph construction that achieves the best possible space complexity up to log log n-factors for an existing memory-hardness measure called cumulative complexity in the sequential pebbling model.

Available format(s)
Publication info
memory-hard functionspebblingspace-time trade-offs
Contact author(s)
quanquan @ mit edu
2018-09-25: last of 3 revisions
2018-02-22: received
See all versions
Short URL
Creative Commons Attribution


      author = {Thaddeus Dryja and Quanquan C.  Liu and Sunoo Park},
      title = {Static-Memory-Hard Functions, and Modeling the Cost of Space vs. Time},
      howpublished = {Cryptology ePrint Archive, Paper 2018/205},
      year = {2018},
      note = {\url{}},
      url = {}
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.