=Paper=
{{Paper
|id=Vol-3628/paper15
|storemode=property
|title=Evaluating Hashing Algorithms in the Age of ASIC Resistance
|pdfUrl=https://ceur-ws.org/Vol-3628/paper15.pdf
|volume=Vol-3628
|authors=Olexandr Kuznetsov,Yelyzaveta Kuznetsova,Oleksii Smirnov,Oleksii Kostenko,Volodymyr Zvieriev
|dblpUrl=https://dblp.org/rec/conf/ittap/KuznetsovKSKZ23
}}
==Evaluating Hashing Algorithms in the Age of ASIC Resistance==
https://ceur-ws.org/Vol-3628/paper15.pdf
Evaluating Hashing Algorithms in the Age of ASIC Resistance
Oleksandr Kuznetsov 1,2, Yelyzaveta Kuznetsova 2, Oleksii Smirnov 3, Oleksii Kostenko 4 and
Volodymyr Zvieriev 5
1
Department of Political Sciences, Communication and International Relations, University of Macerata, Via
Crescimbeni, 30/32, 62100 Macerata, Italy
2
Department of Information and Communication Systems Security, School of Computer Sciences, V. N. Karazin
Kharkiv National University, 4 Svobody Sq., 61022 Kharkiv, Ukraine
3
Department of cyber security and software, Central Ukrainian National Technical University, 8, University Ave,
25006 Kropyvnytskyi, Ukraine
4
Scientific Laboratory of the Theory of Digital Transformation and Law of the Scientific Center for Digital
Transformation and Law of the State Scientific Institution "Institute of Information Security and Law of the
National Academy of Legal Sciences of Ukraine", P.Orlyk str., 3, 01024, Kyiv, Ukraine
5
Department of Software Engineering and Cybersecurity, State University of Trade and Economics, Kyoto, 19,
02156 Kyiv, Ukraine
Abstract
In the intricate matrix of cryptographic hashing, two contrasting paradigms vie for precedence:
computational swiftness and resilience against specialized hardware, or ASICs. This study
undertakes a meticulous exploration into these dueling priorities, juxtaposing conventional
stalwarts like SHA-2 and KECCAK against the newer, ASIC-resistant X series. Leveraging
detailed performance metrics and visual analytics, we discern the manifest advantage of SHA-
2 and KECCAK in terms of computational alacrity. However, as the paper delves deeper, a
compelling narrative unfolds. The deliberate design intricacies of the X series, despite their
computational latency, emerge as robust bulwarks against potential centralization threats posed
by ASIC miners. These algorithms embody a strategic deceleration, ensuring that the
cryptocurrency terrain remains hospitable to a broad diaspora of miners. Through this
multifaceted lens, we argue that in the cryptographic domain, raw speed is but one facet of a
larger, more nuanced picture. The underlying ethos of decentralization and democratic access
must guide our trajectory, even if it comes at the expense of pure efficiency. In essence, this
paper endeavors to navigate the delicate balance between rapid computation and the
overarching principles of an inclusive digital realm.
Keywords 1
Cryptographic hashing, ASIC resistance, SHA-2, KECCAK, X series, Computational
efficiency, Hashing algorithms 2
1. Introduction
In the dynamically evolving world of blockchain technology, the reliability, security, and
performance of cryptographic hash functions play a paramount role in ensuring the integrity and
stability of blockchain systems [1]. The X-series hash algorithms, starting with X11 and extending to
X14, have emerged as innovative solutions in response to concerns over the potential vulnerabilities of
simpler hash functions [2,3].
Proceedings ITTAP’2023: 3rd International Workshop on Information Technologies: Theoretical and Applied Problems, November 22–24,
2023, Ternopil, Ukraine, Opole, Poland
EMAIL: kuznetsov@karazin.ua (A. 1); elizabet8smidt12@gmail.com (A. 2); dr.SmirnovOA@gmail.com (A. 3); antizuk@gmail.com (A. 4);
zvieriev_vp@knute.edu.ua (A. 5)
ORCID: 0000-0003-2331-6326 (A. 1); 0000-0002-0573-0913 (A. 2); 0000-0001-9543-874X (A. 3); 0000-0002-2131-0281 (A. 4); 0000-
0002-0907-0705 (A. 5)
©️ 2020 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
The primary ethos behind the creation of the X-series was to introduce a multi-algorithmic approach
to hashing. This not only elevates the security barriers but also attempts to democratize the mining
process, making it more resistant to ASIC (Application-Specific Integrated Circuits) domination. ASIC
resistance is fundamental in preventing the monopolization of mining by powerful entities, ensuring
that the power within the blockchain remains decentralized [4]. Furthermore, a multi-algorithmic
approach implies that even if one of the algorithms gets compromised, the entirety of the hash function
remains robust, drawing its strength from the collective security of the other algorithms [5–7].
For X11, the journey began with an amalgamation of 11 different cryptographic algorithms [3]:
BLAKE [8–11], BMW [12], GROESTL [13,14], J-H [9,15,16], KECCAK [6,14,15], SKEIN [9,17],
LUFFA [9,18], CUBEHASH [9], SHAVITE [9,19], SIMD [9,11], and ECHO [9,20]. But as the digital
currency ecosystem progressed, the need for additional layers of security became evident. X12
incorporated HAMSI [9], X13 introduced FUGUE [9,21], and X14 brought in SHABAL [9,22] to the
ensemble. Each of these added algorithms has its own strengths and rationale, and their integration
further enhances the resilience of the hash function.
In the context of blockchain systems, the significance of these algorithms cannot be understated. A
blockchain, at its core, is a chain of blocks, and each block contains data that's represented by a hash
[23–25]. If the hashing mechanism is compromised, the integrity of the entire blockchain can be
questioned, jeopardizing trust in the system. Thus, the robustness of the hash function directly correlates
with the trustworthiness of the blockchain.
This paper aims to delve deep into the performance metrics of X11, X12, X13, and X14 hash
algorithms. Through rigorous analysis, we intend to shed light on the efficiency, security, and overall
implications these algorithms have on modern blockchain systems. Given the profound impact and
reliance of blockchain technologies on cryptographic hash functions, understanding the nuances and
intricacies of the X-series algorithms becomes not just an academic pursuit but a necessity for the future
of decentralized systems.
2. Literature Review
The realm of cryptographic hashing has witnessed a plethora of literature, each bringing forth
distinct facets of this multifaceted domain. At its inception, Merkle's revolutionary paper introduced
cryptographic hashing, delineating its foundational principles and applications [26]. This seminal work
laid the groundwork for numerous endeavors in the cryptographic landscape.
The evolution of hashing algorithms has been systematically chronicled by Rogaway and Shrimpton
[27]. Their comprehensive survey provides a diachronic perspective, encompassing early designs to
contemporary cryptographic mechanisms. Particularly, they elucidate the importance of attributes like
collision resistance, pre-image resistance, and second pre-image resistance in cryptographic hashing.
A more specific focus on SHA-2, its design, strengths, and weaknesses, is expertly detailed in the
works of Turner [28] and Dang et al [29]. Their meticulous exploration into the National Institute of
Standards and Technology (NIST) endorsed algorithms affirms their computational efficacy. However,
with the advent of ASICs and their looming threat of centralization, researchers like Harvey-Buschel
and Kisagun began highlighting the potential pitfalls of relying solely on computational efficiency [30].
Evidently, the rise of ASICs necessitated an evolution in the cryptographic hashing paradigm [31].
This brought about the exploration of ASIC-resistant algorithms, a topic elaborately discussed in the
paper by Bex et al [32]. Their discourse emphasizes the need for balanced trade-offs to ensure
decentralization in the cryptocurrency ecosystem.
The Keccak team's treatise on the SHA-3 algorithm offers insights into its design philosophy and
underlying intricacies [33]. While it stands as a testament to efficiency, the cryptocurrency community's
relentless pursuit of ASIC resistance led to the advent of the X series [2,3]. These, while not as
prolifically documented as their predecessors, have been critiqued and discussed in forums and white
papers, emphasizing their role in the broader cryptographic panorama.
In summation, the literature presents a compelling narrative of hashing's evolution, from its nascent
stages to the contemporary challenges and strategic pivots necessitated by the ever-evolving digital
landscape.
Recent our contributions have added depth to this discourse. In our publication, we provided a
rigorous performance analysis of cryptographic hash functions specifically geared towards blockchain
applications [5]. This study offers both empirical and analytical insights, highlighting the balance
between security and efficiency. Our prior work focused on the performance evaluation of hash
algorithms on GPUs for blockchain applicability [6]. With the rise of GPU-intensive tasks and their
potential for parallelization, this research provides valuable insights into how different hashing
algorithms perform in such environments. The importance of this investigation becomes particularly
pertinent given the escalating arms race between ASICs and GPUs in the mining sector. A
complementary our publication approached the domain from a statistical perspective, assessing the
robustness and reliability of blockchain hash algorithms through rigorous testing [7]. Our findings
underscored the importance of continuously vetting and validating the hash functions to ascertain their
resilience against potential vulnerabilities.
Collectively, these contributions have significantly expanded our understanding of the nuances
involved in cryptographic hashing, particularly in the context of blockchain. This current article stands
as a logical extension of this series, delving further into the intricacies of cryptographic hashing in light
of ASIC resistance and the emerging challenges therein.
3. Methods
3.1. System Configuration and Environment
The benchmark testing for the X11, X12, X13, and X14 hashing algorithms was executed on an Intel
Core i9-7980 with a clock speed of 2.60 GHz. The system was complemented by a Windows 10
operating system, ensuring a stable and standardized environment for all tests. This configuration was
chosen to provide a balance between modern computing capabilities and reproducibility for future
comparative studies.
3.2. Data Sets
An array of data lengths was employed to ascertain the speed characteristics and efficiency of the
hashing algorithms over diverse data sizes. The specific byte-lengths chosen were: 1, 2, 4, 8, 16, 32,
and 64 bytes. This expansive range aimed to offer insights into the algorithms' behavior across both
minimal and expansive data sets, simulating real-world scenarios of varying transaction sizes within
blockchain systems.
3.3. Performance Metrics
For each data length, three critical speed characteristics were measured to provide a comprehensive
understanding of the hashing algorithms' performance:
• Cycles per Byte (C/B): This metric denotes the number of CPU cycles required to process each
byte of data, providing insight into the computational intensity of the hash function for different data
sizes.
• Throughput (MB/sec): Illustrating the data processing speed, this parameter conveys the
volume (in Megabytes) processed by the hash function every second. A higher throughput signifies
more efficient data handling.
• Hash Rate (KHash/sec): This measure reflects the number of hash computations executed every
second, expressed in thousands (KiloHashes). It offers a direct lens into the hashing speed, a critical
aspect for blockchain applications, especially in contexts like cryptocurrency mining.
All tests were performed multiple times to account for any potential variability and ensure consistent
results. The average values were subsequently taken for each data length and metric, bolstering the
accuracy and reliability of the findings.
4. The X11 Hashing Algorithms
The X11 hashing scheme, a formidable and intricate construct, amalgamates eleven distinct
cryptographic hash functions to create a multi-layered, robust, and adaptable framework. Each
individual algorithm within the X11 ensemble has been meticulously chosen due to its proven
cryptographic strengths and unique attributes. This section delves deep into each of these algorithms,
shedding light on their mechanisms and relevance within the X11 hashing paradigm.
4.1. BLAKE
• Overview: BLAKE is an algorithm that was one of the finalists in the NIST SHA-3 competition
[34]. It's known for its high speed in software and resistance to differential cryptanalysis.
• Features: Incorporates the HAIFA structure (a variant of the Merkle–Damgård construction)
and leverages components of the ChaCha stream cipher. Its design principles center around
simplicity and security.
• Relevance to X11: Provides a rapid start to the chaining of algorithms within X11 and ensures
initial resistance to various cryptanalytic vulnerabilities.
4.2. BMW (Blue Midnight Wish)
• Overview: Another finalist of the NIST SHA-3 competition [34], BMW stands out due to its
intricate design and high security margins.
• Features: Utilizes a complex structure involving bitwise operations, lookup tables, and modular
arithmetic. Designed for optimal performance on 64-bit platforms.
• Relevance to X11: Offers depth to the chained structure with its unique design, ensuring
diversity in cryptographic techniques used.
4.3. GROESTL
• Overview: A NIST SHA-3 finalist [34], GROESTL is aimed at achieving a high level of
security through its unique construction, which minimizes the risk of certain types of cryptanalytic
attacks.
• Features: Uses a two-fold permutation process and the AES block cipher construct. Provides
high assurance against differential and linear cryptanalysis.
• Relevance to X11: Introduces AES-based operations into the mix, giving the chain enhanced
resilience against potential vulnerabilities.
4.4. J-H
• Overview: A participant of the NIST SHA-3 competition [34], J-H emphasizes cryptographic
robustness and adaptability.
• Features: Built upon the sponge construction, it uses an expansive S-box and an 8x8 matrix for
its cryptographic processes.
• Relevance to X11: The sponge construction of J-H ensures varied cryptographic processing,
further diversifying the X11 structure.
4.5. KECCAK
• Overview: The crowned winner of the NIST SHA-3 competition [34], KECCAK is renowned
for its stellar security and efficiency.
• Features: Employs a permutation-based sponge construction. Renowned for its high-speed
performance and minimalistic design.
• Relevance to X11: Being a core component, KECCAK lends its well-established security
credentials to the overall robustness of the X11 scheme.
4.6. SKEIN
• Overview: A NIST SHA-3 finalist [34], SKEIN is built around the Threefish block cipher and
emphasizes both speed and security.
• Features: Incorporates the Unique Block Iteration (UBI) chaining mode and offers flexibility
in output size.
• Relevance to X11: Its flexible nature and the UBI chaining mode provide a versatile
cryptographic angle to the X11 chain.
4.7. LUFFA
• Overview: A participant in the NIST SHA-3 contest [34], LUFFA is a wide-pipe construction
hash function with a layered design.
• Features: Contains three parallel streams of operations and combines results in a unique
weaving technique.
• Relevance to X11: The wide-pipe structure of LUFFA adds an extra layer of cryptographic
complexity, further solidifying the security parameters of X11.
4.8. CUBEHASH
• Overview: Developed as an entrant to the NIST SHA-3 competition, CUBEHASH offers a
balance of security and efficiency.
• Features: Uses an iterative permutation-based approach and can be tuned for performance on
both hardware and software platforms.
• Relevance to X11: Its adaptability and iterative nature introduce additional cryptographic
variety into the X11 algorithm.
4.9. SHAVITE
• Overview: Another contestant of the NIST SHA-3 race [34], SHAVITE focuses on providing
high-speed hashing on a variety of platforms.
• Features: Incorporates a fast block cipher in its design and optimizes performance on 32-bit
architectures.
• Relevance to X11: Ensures that X11 remains efficient across diverse hardware architectures.
4.10. SIMD
• Overview: Developed for the NIST SHA-3 competition [34], SIMD stands for "Single
Instruction, Multiple Data" reflecting its parallel processing design.
• Features: Operates on a wide data path and employs a complex sequence of operations for
maximum security.
• Relevance to X11: Its parallel processing approach aids in maximizing the throughput of the
hashing scheme.
4.11. ECHO
• Overview: A NIST SHA-3 competitor [34], ECHO is designed around the AES block cipher,
targeting both efficiency and security.
• Features: Uses a unique double-pipeline structure and optimizes AES operations for hashing.
• Relevance to X11: Contributes AES-based cryptographic strength and introduces an innovative
pipeline design to the X11 algorithm.
In culmination, the X11 hashing scheme stands as a testament to the strengths of amalgamating
diverse cryptographic techniques. By harnessing the unique attributes of each of its constituent
algorithms, X11 ensures a fortified and versatile hashing paradigm, pivotal for blockchain's rigorous
demands.
5. Subsequent Evolutions of the X Hashing Algorithms: X12, X13, and X14
The X series of hashing algorithms has been a hallmark of cryptographic progression, offering multi-
algorithmic chains that provide a robust defense against potential vulnerabilities. The inception of the
X11 algorithm set a precedent in combining 11 individual cryptographic hashes. Its successors, X12,
X13, and X14, further build upon this foundation by incorporating additional hashing mechanisms. This
section highlights the added algorithms in these subsequent versions and provides a detailed
examination of their cryptographic relevance.
5.1. X12 (HAMSI)
Building upon the X11 base, X12 introduces a single additional hashing algorithm (HAMSI):
• Overview: A contender in the NIST SHA-3 competition [34], HAMSI is designed to be compact
and suitable for constrained environments.
• Features: Utilizes a series of S-boxes, linear diffusion layers, and MDS matrices. It exhibits a
strong resistance against differential cryptanalysis.
• Relevance to X12: HAMSI augments X12's resilience by adding its unique cryptographic
structure, making the algorithm more suitable for environments with limited computational
resources.
5.2. X13 (FUGUE)
Expanding further on its predecessor, X13 adds yet another layer to the established hash chain
(FUGUE):
• Overview: Another entrant to the NIST SHA-3 competition [34], FUGUE processes data in a
unique matrix format, providing both speed and security.
• Features: Implements a 3D extension of the AES S-box, combined with a series of rotations,
permutations, and non-linear transformations. This layered approach offers heightened security
assurances.
• Relevance to X13: FUGUE's 3D matrix processing mechanism introduces a novel dimension
to the cryptographic chain, strengthening the diversity of techniques within the X13 algorithm.
5.3. X14 (SHABAL)
The X14 hashing scheme adds one more hash function to the already intricate X13 system
(SHABAL):
• Overview: Participating in the NIST SHA-3 race [34], SHABAL stands out due to its symmetric
structure and ability to process vast amounts of data efficiently.
• Features: Employs a unique combination of bitwise operations and modular arithmetic. Its
design allows for parallel processing, enhancing throughput.
• Relevance to X14: SHABAL's symmetric design and parallel processing capabilities further
diversify the cryptographic techniques within the X series, ensuring that X14 maintains high
efficiency even with its increased complexity.
In summation, the evolution from X11 to X14 signifies the relentless pursuit of cryptographic
perfection. By continuously integrating diverse and proven hash functions, the X series ensures that it
remains at the forefront of cryptographic security, meeting the ever-increasing demands of blockchain
and other cryptographic applications.
6. Results
In the realm of cryptographic hashing, both speed and security are paramount. This section evaluates
the performance metrics of multiple cryptographic hash functions, namely SHA2-256, SHA2-512,
KECCAK-256, KECCAK-512, X11, X12, X13, and X14.
To provide a holistic perspective, we present the results in three distinct tables (Tables 1-3), each
representing a different performance metric: Cycles per Byte, Throughput (MB/sec), and Hash Rate
(KHash/sec). This structured approach enables a clearer comparison among these algorithms.
Table 1
Comparative Analysis of Algorithms based on Cycles per Byte
Algorithm 1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
SHA2-256 848.68 423.30 211.72 105.80 52.84 26.28 25.74
SHA2-512 945.51 472.36 236.28 118.07 59.05 29.48 14.79
KECCAK-256 1307.78 657.99 328.59 159.08 79.54 39.75 19.88
KECCAK-512 1269.54 633.65 316.76 156.76 77.77 38.83 19.42
X11 35972.07 18012.95 8878.42 4439.23 2220.24 1109.69 554.82
X12 43318.45 21838.60 10907.93 5456.44 2694.58 1347.39 682.47
X13 49509.65 24764.23 12380.19 6191.50 3092.92 1546.94 772.80
X14 51112.37 25668.86 12788.01 6392.37 3189.71 1595.09 797.48
Table 2
Throughput Analysis Across Different Hashing Algorithms
Algorithm 1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
SHA2-256 3.06 6.13 12.24 24.53 49.07 97.98 100.50
SHA2-512 2.74 5.49 10.98 21.94 43.93 87.89 175.35
KECCAK-256 1.97 3.94 7.88 16.29 32.54 65.20 130.53
KECCAK-512 2.04 4.09 8.18 16.44 33.31 66.58 133.64
X11 0.07 0.14 0.29 0.58 1.17 2.33 4.67
X12 0.06 0.12 0.24 0.48 0.96 1.91 3.80
X13 0.05 0.10 0.21 0.42 0.84 1.68 3.35
X14 0.05 0.10 0.20 0.41 0.81 1.62 3.24
Table 3
Hash Rate Efficiency Across Various Hashing Algorithms
Algorithm 1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
SHA2-256 3060.37 3065.65 3059.57 3066.73 3066.73 3061.86 1570.25
SHA2-512 2741.59 2744.53 2744.68 2742.09 2745.54 2746.69 2739.80
KECCAK-256 1974.12 1969.97 1970.41 1970.63 1970.89 1970.56 1970.89
KECCAK-512 2041.02 2045.36 2044.80 2045.45 2044.98 2044.87 2045.01
X11 71.73 72.33 72.98 73.14 73.14 73.12 73.08
X12 59.35 59.34 59.35 59.36 59.36 59.37 59.37
X13 52.36 52.33 52.34 52.34 52.35 52.35 52.36
X14 50.75 50.67 50.67 50.68 50.68 50.69 50.70
The "Cycles per Byte" metric gives an insight into the computational efficiency of a hashing
algorithm. Lower values are indicative of better efficiency. As observed, both SHA2 variants, especially
SHA2-512, and the KECCAK variants show significantly lower cycle counts compared to the X series
(X11 to X14). Amongst the X series, X11 is the most efficient, with X14 being the least. It's crucial to
note that the cycle counts for SHA2 and KECCAK increase (i.e., improve in efficiency) as the input
block size increases, reaching an optimum at 64 bytes for SHA2-256.
Throughput, measured in MB/sec, reflects the data processing speed of the hashing algorithm.
Higher values indicate superior performance. The SHA2-256 algorithm clearly dominates, with its
throughput peaking at 100.50 MB/sec for a 64-byte input block size. Its counterpart, SHA2-512, closely
follows this trend, but with roughly half the throughput for the same block size. The KECCAK variants,
while less efficient than the SHA2 series, substantially outperform the X series in this regard. The
consistently low throughput values for the X series, especially for smaller block sizes, might raise
concerns in scenarios where rapid data processing is paramount.
The hash rate, presented in KHash/sec, provides a measure of how quickly an algorithm can compute
a hash. Similar to throughput, higher values suggest better performance. Notably, SHA2-256 and
SHA2-512 show remarkable efficiency, consistently outperforming other algorithms across varying
block sizes. However, it's intriguing to note a sharp decline in the hash rate of SHA2-256 for a 64-byte
block size. This deviation warrants further investigation. The KECCAK series maintains a steady hash
rate across all block sizes, suggesting consistent performance. The X series algorithms, while lagging
behind their counterparts, show stability in their hash rates.
In summary, the SHA2 series, especially SHA2-256, generally showcases superior efficiency in
terms of cycles per byte, throughput, and hash rate, followed by the KECCAK variants. The X series,
while not as efficient, provides a consistent hash rate, which might be suitable for applications that
prioritize consistency over raw performance. However, it's essential to consider other factors such as
security robustness, application context, and system compatibility when selecting an appropriate
hashing algorithm.
7. Discussion
The recent developments in the field of cryptographic hashing algorithms have reignited the
discourse surrounding efficiency versus security, particularly in the context of Application-Specific
Integrated Circuit (ASIC) resistance. While it's evident from the data that the SHA-2 and KECCAK
families demonstrate superior performance metrics in terms of computational efficiency, throughput,
and hash rates, their dominance raises pivotal concerns in the ever-evolving arms race against ASIC
miners.
ASICs are specialized hardware designed explicitly for a specific computational task. In the
cryptocurrency world, ASICs tailored for certain hashing algorithms can perform mining operations
orders of magnitude faster than general-purpose computing hardware. This advantage not only
centralizes the mining power, often leading to fewer entities controlling the mining process, but also
poses a threat to the security and democratization of the network.
In this light, the X series of algorithms, despite their apparent lag in raw performance metrics,
emerge as potential game-changers. Their intricate sequence of cryptographic functions is tailored to
deter the advantages ASICs traditionally hold. This level of ASIC-resistance means that, despite the
slower speed, there's a leveling of the playing field, making the mining process more accessible and
distributed.
The considerably lower performance metrics of the X series, as observed in the tables, aren't flaws
but deliberate design decisions. ASIC-resistance often comes at the cost of computational efficiency.
By incorporating multiple rounds of hashing with varied algorithms in sequences, the X series ensures
that an ASIC designed for one specific function would not necessarily excel in others. This inherent
complexity acts as a deterrent, making the ASIC design for such algorithms both challenging and
economically unviable.
The slower performance metrics of the X series might raise eyebrows in isolation, but when viewed
from the lens of ASIC-resistance, they manifest as necessary trade-offs. In a rapidly evolving digital
economy, the balance between efficiency and security is paramount, and the X series embodies this
philosophy.
While the lure of superior performance metrics is undeniable, the broader context of network
security and democratization cannot be understated. The rise of ASICs threatens the foundational
principles of many decentralized systems, and while their influence is undeniable, so too is the necessity
for robust countermeasures. The X series, with its deliberative design sacrificing speed for security,
underscores the lengths the cryptographic community is willing to go to ensure a level playing field. As
we forge ahead, it's vital to remember that the most efficient algorithm is not always the most suitable,
especially when the stakes are as high as the security and inclusivity of global decentralized systems.
Discussion on Results with Respect to ASIC Resistance
The Figure 1 illustrates the cycles per byte required for each hashing algorithm across varying input
block sizes. Lower values represent higher efficiency.
A cursory glance at the graph indicates that SHA-2 and KECCAK families outperform the X series
in terms of raw computational efficiency, requiring fewer cycles per byte. The X series, however,
although less efficient, is intentionally designed to mitigate ASIC advantages. The increased
computational requirements can be interpreted as a deliberate sacrifice to deter ASIC dominance.
The Figure 2 provides a comparative representation of the throughput (MB/sec) exhibited by each
algorithm as the input block size increases.
60000
Cycles per Byte SHA2-256
50000 SHA2-512
KECCAK-256
40000
KECCAK-512
30000 X11
X12
20000
X13
X14
10000
Input Block Size
0 (bytes)
1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
Figure 1: Computational Efficiency Across Hashing Algorithms
200
MB/sec SHA2-256
180
SHA2-512
160
KECCAK-256
140
KECCAK-512
120
100 X11
80 X12
60 X13
40
X14
20
Input Block Size
(bytes)
0
1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
Figure 2: Throughput Performance of Hashing Algorithms
The throughput graph reinforces the notion of SHA-2 and KECCAK's superiority in handling larger
data blocks more swiftly. However, the X series' relatively linear scalability showcases its resilience
against input variations. This trait might be instrumental in environments where input data size is
unpredictable.
The Figure 3 depicts the KHash/sec metrics for each algorithm, providing insights into their hashing
speeds.
Once again, the SHA-2 and KECCAK families demonstrate an edge with higher hash rates. The X
series, while trailing, maintains consistent performance across input block sizes. The steadiness of the
X series, even if slower, might be seen as a testament to its robust and ASIC-resistant design.
The graphs offer a concise yet comprehensive overview of the stark contrasts between traditional,
efficient hashing functions and the newer, ASIC-resistant X series. The inherent trade-offs become
more pronounced when visualized. While efficiency is an invaluable metric, the need for a decentralized
and secure network that resists monopolization by powerful ASIC miners is equally crucial.
The visual aids drive home the point that while some algorithms excel in speed, others are tailored
for resilience and security in a constantly evolving digital ecosystem.
10000
KHash/sec
SHA2-256
SHA2-512
1000
KECCAK-256
KECCAK-512
100 X11
X12
X13
10
X14
Input Block Size
(bytes)
1
1 byte 2 bytes 4 bytes 8 bytes 16 bytes 32 bytes 64 bytes
Figure 3: Hash Rate Comparison Across Algorithms
8. Conclusion
In the rapidly evolving landscape of cryptographic hash functions, the quest for efficiency and
security operates in tandem, often at cross purposes. Through this study, we endeavored to elucidate
the performance dynamics of a diverse array of hashing algorithms, juxtaposing renowned ones like
SHA-2 and KECCAK against the ASIC-resistant X series. Our findings underscore a palpable tension
between raw computational efficiency and the deliberate design choices made to thwart potential
centralization by powerful ASIC miners.
The visual representations generated from our performance metrics unequivocally indicate the
superiority of SHA-2 and KECCAK in terms of cycles per byte, throughput, and hash rates. Such
proficiency is advantageous in scenarios demanding swift computational outcomes. However, the
digital world is not merely governed by the swiftness of computations. The principles of
decentralization, equitable access, and resistance to potential monopolization hold paramount
importance in ensuring a sustainable and inclusive digital ecosystem.
The X series of hashing algorithms, despite their apparent computational sluggishness, are
emblematic of this ideological pivot. Their resilience and consistent performance, even at the cost of
raw speed, represent a concerted effort to ensure that the realm of cryptocurrency remains accessible to
a broad spectrum of miners, not just those wielding the power of ASICs.
In sum, while efficiency is indisputably a coveted attribute, the broader objectives of decentralization
and equitable access necessitate a balanced approach. It's imperative to recognize that in the world of
cryptographic hashing, the fastest isn't always the most fitting. As technology continues to advance, the
ongoing challenge for researchers and developers will be to harmonize efficiency with egalitarian
principles, ensuring that the digital future remains inclusive and secure for all.
References
[1] A. Tiwari, Chapter 14 - Cryptography in blockchain, in: R. Pandey, S. Goundar, S. Fatima (Eds.),
Distributed Computing to Blockchain, Academic Press, 2023: pp. 251–265.
https://doi.org/10.1016/B978-0-323-96146-2.00011-5.
[2] What is Hashing Algorithm?, Changelly.Com. (2020). https://changelly.com/blog/hashing-
algorithms-explained/ (accessed September 7, 2023).
[3] coinguides, X11 Coins - List of Cryptocurrencies based on X 11 hashing algorithm, Coin Guides.
(2019). https://coinguides.org/x11-algorithm-coins/ (accessed September 7, 2023).
[4] coinguides, ASIC resistance explained - Importance of coins with ASIC resistant, Coin Guides.
(2018). https://coinguides.org/asic-resistance-explained/ (accessed September 7, 2023).
[5] A. Kuznetsov, I. Oleshko, V. Tymchenko, K. Lisitsky, M. Rodinko, A. Kolhatin, Performance
analysis of cryptographic hash functions suitable for use in blockchain, Int. J. Computer Netw.
Inf. Security. 13 (2021) 1–15. https://doi.org/10.5815/IJCNIS.2021.02.01.
[6] A. Kuznetsov, K. Shekhanin, A. Kolhatin, D. Kovalchuk, V. Babenko, I. Perevozova,
Performance of Hash Algorithms on GPUs for Use in Blockchain, in: IEEE Int. Conf. Adv. Trends
Inf. Theory, ATIT - Proc., Institute of Electrical and Electronics Engineers Inc., 2019: pp. 166–
170. https://doi.org/10.1109/ATIT49449.2019.9030442.
[7] A. Kuznetsov, M. Lutsenko, K. Kuznetsova, O. Martyniuk, V. Babenko, I. Perevozova, Statistical
testing of blockchain hash algorithms, in: Fedushko S., Gnatyuk S., Peleshchyshyn A., Hu Z.,
Odarchenko R., Korobiichuk I. (Eds.), CEUR Workshop Proc., CEUR-WS, 2019.
https://www.scopus.com/inward/record.uri?eid=2-s2.0-
85083260414&partnerID=40&md5=e65f8fb02e0a57799a5aaea7ccfe195e.
[8] BLAKE2, (n.d.). https://www.blake2.net/ (accessed September 6, 2023).
[9] S. Tillich, M. Feldhofer, M. Kirschbaum, T. Plos, J.-M. Schmidt, A. Szekely, High-Speed
Hardware Implementations of BLAKE, Blue Midnight Wish, CubeHash, ECHO, Fugue, Grøstl,
Hamsi, JH, Keccak, Luffa, Shabal, SHAvite-3, SIMD, and Skein, (2009).
https://eprint.iacr.org/2009/510 (accessed September 7, 2023).
[10] E. Andreeva, A. Luykx, B. Mennink, Provable Security of BLAKE with Non-Ideal Compression
Function, (2011). https://eprint.iacr.org/2011/620 (accessed September 7, 2023).
[11] S. Neves, J.-P. Aumasson, Implementing BLAKE with AVX, AVX2, and XOP, (2012).
https://eprint.iacr.org/2012/275 (accessed September 7, 2023).
[12] M.E. Hadedy, D. Gligoroski, S.J. Knapskog, Single Core Implementation of Blue Midnight Wish
Hash Function on VIRTEX 5 Platform, (2010). https://eprint.iacr.org/2010/571 (accessed
September 7, 2023).
[13] M. Rogawski, K. Gaj, Groestl Tweaks and their Effect on FPGA Results, (2011).
https://eprint.iacr.org/2011/635 (accessed September 7, 2023).
[14] B. Jungk, S. Reith, J. Apfelbeck, On Optimized FPGA Implementations of the SHA-3 Candidate
Groestl, (2009). https://eprint.iacr.org/2009/206 (accessed September 7, 2023).
[15] Hash Function JH; designed by Hongjun Wu, (n.d.).
https://www3.ntu.edu.sg/home/wuhj/research/jh/ (accessed September 7, 2023).
[16] R. Bhattacharyya, A. Mandal, M. Nandi, Security Analysis of the Mode of JH Hash Function, in:
S. Hong, T. Iwata (Eds.), Fast Software Encryption, Springer, Berlin, Heidelberg, 2010: pp. 168–
191. https://doi.org/10.1007/978-3-642-13858-4_10.
[17] J.-P. Aumasson, C. Calik, W. Meier, O. Ozen, R.C.-W. Phan, K. Varici, Improved Cryptanalysis
of Skein, (2009). https://eprint.iacr.org/2009/438 (accessed September 7, 2023).
[18] T. Oliveira, J. López, Improving the performance of Luffa Hash Algorithm, (2010).
https://eprint.iacr.org/2010/457 (accessed September 7, 2023).
[19] O. Dunkelman, E. Biham, The SHAvite-3 - A New Hash Function, in: H. Handschuh, S. Lucks,
B. Preneel, P. Rogaway (Eds.), Symmetric Cryptography, Schloss Dagstuhl – Leibniz-Zentrum
für Informatik, Dagstuhl, Germany, 2009: pp. 1–39.
https://doi.org/10.4230/DagSemProc.09031.18.
[20] J. Jean, M. Naya-Plasencia, M. Schläffer, Improved Analysis of ECHO-256, (2011).
https://eprint.iacr.org/2011/422 (accessed September 7, 2023).
[21] S. Halevi, W.E. Hall, C.S. Jutla, The Hash Function “Fugue,” (2014).
https://eprint.iacr.org/2014/423 (accessed September 7, 2023).
[22] Shabal, (2009). https://web.archive.org/web/20091209170807/http://www.shabal.com/ (accessed
September 6, 2023).
[23] S.-W. Lee, I. Singh, M. Mohammadian, eds., Blockchain Technology for IoT Applications,
Springer, Singapore, 2021. https://doi.org/10.1007/978-981-33-4122-7.
[24] K. Xu, J. Zhu, X. Song, Z. Lu, eds., Blockchain Technology and Application: Third CCF China
Blockchain Conference, CBCC 2020, Jinan, China, December 18-20, 2020, Revised Selected
Papers, Springer, Singapore, 2021. https://doi.org/10.1007/978-981-33-6478-3.
[25] S. Van Hijfte, Bitcoin, in: S. Van Hijfte (Ed.), Blockchain Platforms: A Look at the Underbelly
of Distributed Platforms, Springer International Publishing, Cham, 2020: pp. 71–146.
https://doi.org/10.1007/978-3-031-01804-6_2.
[26] R.C. Merkle, A Digital Signature Based on a Conventional Encryption Function, in: C. Pomerance
(Ed.), Advances in Cryptology — CRYPTO ’87, Springer, Berlin, Heidelberg, 1988: pp. 369–
378. https://doi.org/10.1007/3-540-48184-2_32.
[27] P. Rogaway, T. Shrimpton, Cryptographic Hash-Function Basics: Definitions, Implications, and
Separations for Preimage Resistance, Second-Preimage Resistance, and Collision Resistance, in:
B. Roy, W. Meier (Eds.), Fast Software Encryption, Springer, Berlin, Heidelberg, 2004: pp. 371–
388. https://doi.org/10.1007/978-3-540-25937-4_24.
[28] S. Turner, Using SHA2 Algorithms with Cryptographic Message Syntax, Internet Engineering
Task Force, 2010. https://doi.org/10.17487/RFC5754.
[29] D.R.L. Brown, T. Polk, S. Santesson, K. Moriarty, Q. Dang, Internet X.509 Public Key
Infrastructure: Additional Algorithms and Identifiers for DSA and ECDSA, Internet Engineering
Task Force, 2010. https://doi.org/10.17487/RFC5758.
[30] J. Harvey-Buschel, C. Kisagun, Bitcoin Mining Decentralization via Cost Analysis, (2016).
https://doi.org/10.48550/arXiv.1603.05240.
[31] K. Ivy, The Looming Threat of Centralized Cryptocurrency, iM Intelligent Mining. (2022).
https://medium.com/im-intelligent-mining/the-looming-threat-of-centralized-cryptocurrency-
562dcc460eae (accessed September 7, 2023).
[32] L. Bex, F. Turan, M. Van Beirendonck, I. Verbauwhede, Mining CryptoNight-Haven on the
Varium C1100 Blockchain Accelerator Card, (2022). https://doi.org/10.48550/arXiv.2212.05033.
[33] G. Bertoni, J. Daemen, M. Peeters, G. Van Assche, Keccak, in: T. Johansson, P.Q. Nguyen (Eds.),
Advances in Cryptology – EUROCRYPT 2013, Springer, Berlin, Heidelberg, 2013: pp. 313–314.
https://doi.org/10.1007/978-3-642-38348-9_19.
[34] I.T.L. Computer Security Division, SHA-3 Project - Hash Functions | CSRC | CSRC, CSRC |
NIST. (2017). https://csrc.nist.gov/projects/hash-functions/sha-3-project (accessed September 6,
2023).