The DOI matches its Crossref record; 75 of the 76 references it cites were found in a registry or an index; 3 of the 32 sources its claims cite were checked for support.
First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES
Open accessPublished version of a preprint
Abstract
In response to the recent NIST call for a wider variant of the AES algorithm, we developed a fully pipelined, high-throughput FPGA implementation of the 256-bit block size AES, referred to as WAES-256... Additionally, we optimized WAES-256 on GPUs and achieved performance comparable to the best AES-256 results. For instance, we achieved 3053.5 Gbps WAES-256 encryption in counter mode of operation...
Overview
CoChat Summary
TL;DR
The paper presents what it claims is the first FPGA and GPU implementation of the wider AES variant (WAES-256), using fully pipelined FPGA designs and table-based GPU kernels. It reports high throughputs, including 742.63 Gbps for x4-core WAES-256 on Kintex UltraScale+ and 3053.5 Gbps WAES-256 CTR on an RTX 4090, and corrects underreported AES GPU results.
Key findings
- FPGA WAES-256 single-core reached 75.73 Gbps on Spartan-7, 72.32 Gbps on Artix-7, and 206.11 Gbps on Kintex UltraScale+; x2 and x4 cores on Kintex UltraScale+ reached 426.66 Gbps and 742.63 Gbps, respectively.
- GPU WAES-256 CTR achieved 3053.5 Gbps on an RTX 4090; corrected AES-128 CTR reached 943.4 Gbps on an RTX 2070 Super (vs. 878.6 Gbps reported in prior work) and 4441.1 Gbps on an RTX 4090, while corrected AES-256 CTR reached 3062.5 Gbps on an RTX 4090.
- AES-128 implementations achieved the best throughput-per-slice (TPS) ratios reported for fair comparisons on the same FPGA platforms, e.g., 7.13 Mbps/LUT on Kintex UltraScale+ and 7.01 Mbps/LUT on Zynq UltraScale+.
- The transition from AES-256 to WAES-256 resulted in no observable slowdowns on GPUs, and the proposed compact S-box [3] was more area-efficient on Artix-7, Zynq, and Kintex UltraScale+ but less compact than the standard AES S-box on Spartan-7.
Methods
- FPGA: fully pipelined all AES layers, used composite-field arithmetic S-box (Canright) and evaluated the compact S-box from [3]; supported AES-128, AES-256, and WAES-256 via generic parameters; avoided BRAM; ran key schedule and round function in parallel; applied sub-pipelining to Mix-Columns and Sub-Bytes; implemented on Spartan-7 SP701, Artix-7 AC701, Zynq UltraScale+ ZCU106, and Kintex UltraScale+ KCU116.
- GPU: table-based WAES-256 (256-bit block and key) with 32 copies of one table in shared memory (32 KB), deriving other tables via byte permutations at runtime; used CTR mode; tested on MX250, GTX860M, GTX970, RTX 2070 Super, and RTX 4090; re-ran prior AES code to correct underreported results.
- Analysis: compared throughput, TPS, critical delay, and hardware utilization against prior FPGA and GPU AES implementations.
Limitations
- Paper-stated: GPU results may differ due to thermal throttling, especially on mobile GPUs; the parallel key schedule consumes more power; multi-core TPS drops due to routing and critical-path constraints; the compact S-box advantage is less evident in high-resource designs; older FPGA comparisons could not be obtained due to licensing restrictions.
- Evident from the text: the abstract reports 199.46 Gbps WAES-256 on Zynq UltraScale+, while Table 8 lists 191.04 Gbps for the same platform, an internal inconsistency.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 6 Not valid · 29 Not checked · 81 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For WAES-256, our designs reached 75.73 Gbps on Spartan-7, 72.32 Gbps on Artix-7, 199.46 Gbps on Zynq UltraScale+, and 206.11Gbps on Kintex UltraScale+.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Our AES-128 implementations achieved the best throughput-per-slice (TPS) ratios reported in the literature for fair comparisons on the same FPGA platforms.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Additionally, our multi-core parallel WAES-256 designs achieved 426.66 Gbps with x2 cores and 742.63 Gbps with x4 cores on the Kintex UltraScale+ platform, demonstrating the scalability of our approach.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > Abstract
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In the wider variant of AES (256-bit block size), the Mix-Columns layer is implemented with a latency of three clock cycles.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 5.3 Mix-Columns Layer
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For AES-128, the design includes 10 rounds, each requiring 7 cycles, resulting in a total latency of 70 cycles.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 5.4 Key Schedule
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
64 more claims
Create a free account to see all 116 claims and to verify the 1 reference still open.
- It may state two separate claims as one. 36
- The quote only partly states this claim. 3
- Probably not a substantial claim. 11
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“The architectural choice does not preclude future migration to a wider block cipher: GpuFitsCrypt isolates the AES core behind a clear interface, and recent results (Malal and Tezcan, 2026) indicate that migration to WAES-256 / Rijndael-256 produces no observable slowdown (3,053.5 Gbps for WAES-256 versus 3,062.5 Gbps for AES-256 on RTX 4090, a difference of approximately 0.3%).”
From A high-throughput AES-GCM implementation on GPUs for secure, policy-based access to massive astronomical catalogs (2026) · cites this paper as Malal and Tezcan, 2026
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“our design reached 426.66 Gbps with 91204 FFs and 104529 LUTs, resulting in a TPS of 4.68 Mbps/-LUT.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 7 Results and Comparisons
Reported 4.68 · recomputed 0.004082
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > Algorithm 2 Key Expansion Algorithm
“cipherKey of N k bytes Ensure: roundKeys[0 . . . N r ] of N b bytes each”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > AES algorithm is an example of a symmetrickey algorithm, which is based on a substitutionpermutation network (SPN).
“• N b : the block size, in 32-bit words. • N k : the key size, in 32-bit words.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > AES-128 -Layer Pipelined
“An AES implementation on a Virtex-7 FPGA with a latency of 74 clock cycles was developed in [56], achieving a throughput of 278 Mbps.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 8 Conclusion and Future Work
“| AES-128 [56] Virtex-7 | | 1131 FF 4834 LUT 3095 Slices | 0.249 | 109 | 9.17 | 0.051 | 2.75 |”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 7 Results and Comparisons
“Other Artix-7 results, such as those by [55], [70], and [57], show lower throughputs compared to our work.For example, the design by [55] achieved only 1.795 Gbps using 765 slices, while the work by [70] reached 7.67 Gbps using 9175 LUTs, and [57] reported 11.11 Gbps with 1076 FFs and 7498 LUTs.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 8 Conclusion and Future Work
“| AES-128 [57] Artix-7 | | 1095 FF 8772 LUT 2445 Slice | 7.67 | 164.9 | 6.06 | 0.87 | - |”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“our AES-128 design on the AC701 board reached 54.46 Gbps using 18313 FFs and 17414 LUTs, with a critical delay of 2.35 ns and a TPS of 3.17 Mbps/LUT.”
§ First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES > 7 Results and Comparisons
Reported 3.17 · recomputed 0.003127
1 more finding
Create a free account to see every finding of this check.
Citations
Cites
Differential cryptanalysis of DES-like cryptosystems
Biham, E., & Shamir, A. (1991) · Journal of Cryptology · doi:10.1007/bf00630563 · Cited by 2,914 · cited 1×
“The MDS property is critical for maximizing diffusion and resisting differential [7] and linear cryptanalysis [8].”
Linear Cryptanalysis Method for DES Cipher
Matsui, M. (1994) · Lecture Notes in Computer Science · doi:10.1007/3-540-48285-7_33 · Cited by 2,748 · cited 1×
“The MDS property is critical for maximizing diffusion and resisting differential [7] and linear cryptanalysis [8].”
Differential Cryptanalysis of DES-like Cryptosystems
Biham, E., & Shamir, A. (2007) · Lecture Notes in Computer Science · doi:10.1007/3-540-38424-3_1 · Cited by 658
Canright, D. (2005) · Lecture Notes in Computer Science · doi:10.1007/11545262_32 · Cited by 530 · cited 2×
“We used the composite field arithmetic method proposed by [3] for efficient S-box computation.”
High-speed VLSI architectures for the AES algorithm
Zhang, X., & Parhi, K. K. (2004) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2004.832943 · Cited by 448
(2007) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-74735-2 · Cited by 418
The Advanced Encryption Standard Process
Daemen, J., & Rijmen, V. (2002) · Information Security and Cryptography · doi:10.1007/978-3-662-04722-4_1 · Cited by 322
(2003) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-45238-6 · Cited by 286
AES on FPGA from the Fastest to the Smallest
Good, T., & Benaissa, M. (2005) · Lecture Notes in Computer Science · doi:10.1007/11545262_31 · Cited by 233
A 21.54 Gbits/s Fully Pipelined AES Processor on FPGA
Hodjat, A., & Verbauwhede, I. (2004) · 12th Annual IEEE Symposium on Field-Programmable Custom Computing Machines · doi:10.1109/fccm.2004.1 · Cited by 227
Research works printed but not linked: 3
- Kaur, A., Bhardwaj, P., Kumar, N.: Fpga implementation of efficient hardware for the advanced encryption standard. Int. j. innov. technol. explor. eng. 2(3), 186–189 (2013)
- Rais, M.H., Qasim, S.M.: @articleDBLP:journals/cee/AliAHR11, author = Liakot Ali and Ishak Aris and Fakir Sharif Hossain and Niranjan Roy, title = Design of an ultra high speed AES processor for next generation IT security, journal = Comput. Electr. Eng., volume = 37, number = 6, pages = 1160–1170, year = 2011, url = https://doi.org/10.1016/j.compeleceng.2011.06.003, doi = 10.1016/J.COMPELECENG.2011.06.003, timestamp = Mon, 28 Aug 2023 21:33:14 +0200, biburl = https://dblp.org/rec/journals/cee/AliAHR11.bib, bibsource = dblp computer science bibliography, https://dblp.org . International journal of computer science and network security 9(9), 59–63 (2009) · DOI
- Singh, G., Mehra, R.: Fpga based high speed and area efficient aes encryption for data security. Int.l J. Research and Innov. Comput. Eng. 1(2), 53–56 (2011)
Cited by
First-Order Threshold Implementation of AES Against Side-Channel Attacks on FPGA
Çalıkuş, A., Çalışkan, Ş., Örs, B., MALAL, A., Kaan, E. K. E. R., & Aksoy, B. (2025) · doi:10.1109/isctrkiye68593.2025.11224814 · Cited by 2
Self-citationLemes-Perera, S., R. Alarcon, M., Caballero-Gil, P., & Serra-Ricart, M. (2026) · Astronomy and Computing · doi:10.1016/j.ascom.2026.101153 · Cited by 1
Design and Implementation of an AES Hardware Encryption Core
Kumar, A., Mehfuz, S., & Urooj, S. (2026) · Symmetry · doi:10.3390/sym18060897
Metadata
- Authors
- Malal, A., & Tezcan, C.
- Affiliations
- Middle East Technical University (TR)
- Corresponding author
- Malal, Ahmet
- Venue
- Journal of Cryptographic Engineering
- Published
- 2026-01-04
- Volume
- 16
- Issue
- 1
- ISSN
- 2190-8508 (ISSN-L)
- 2190-8516
- OpenAlex
- W7118126257
- Semantic Scholar
- 031d555b898c58f3746e7e1fa6e0366954ac680b
- Cited by
- 4
- References
- 68
- Citations by year
- 2026: 3
- FWCI
- 4.45
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- cc-by
- www.springernature.com/gp/researchers/text-and-data-mining (tdm)
- www.springernature.com/gp/researchers/text-and-data-mining (vor)
- Locations
- Journal of Cryptographic Engineering (publishedVersion)
- Research Square (submittedVersion, cc-by)
- Funders
- The Scientific and Technological Research Council of Türkiye (123N546)
- German Academic Exchange Service
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Cryptographic Implementations and Security
- Physical Sciences › Computer Science › Information Systems › Cryptography and Residue Arithmetic
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Keywords
- AES-256, AES-128, FPGA implementation, GPU acceleration, pipelined hardware architecture, composite field arithmetic, S-box, counter mode, cryptographic coprocessor
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-12, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-12 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES. https://app.cochat.ai/papers/10.1007/s13389-025-00388-2
- The paper
- Malal, A., & Tezcan, C. (2026). First Fully Pipelined High Throughput FPGA Implementation and GPU Optimization of Wider Variant of AES. Journal of Cryptographic Engineering. https://doi.org/10.1007/s13389-025-00388-2
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.