Pith. sign in

REVIEW 3 major objections 6 minor 19 references

Processing of GASKAP-HI pilot survey data using a commercial supercomputer

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Commercial supercomputer processes radio survey for A$5,250 per field

desk verdict A solid, practical trial of commercial HPC for radio imaging; the feasibility claim holds, but the cost model needs a clarifying edit in Table 4. read the letter →

arxiv 2411.17118 v2 pith:255UU56C submitted 2024-11-26 astro-ph.IM

Ian P. Kemp , Nickolas M. Pingel , Rowan Worth , Justin Wake , Daniel A. Mitchell , Stuart D. Midgely , Steven J. Tingay , James Dempsey
show 10 more authors
Helga Dénes John M. Dickey Steven J. Gibson Kate E. Jameson Callum Lynn Yik Ki Ma Antoine Marchal Naomi M. McClure-Griffiths Snežana Stanimirović Jacco Th. van Loon
This is my paper · ORCID
classification astro-ph.IM
keywords radioastronomyhighperformancecomputingcommercialsuperGASKAP-HIsurveyneutralhydrogenimagingjointdeconvolutionworkflowportingcostestimation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that a commercial supercomputer can run a full-scale radio astronomy imaging workload—processing GASKAP-HI pilot survey data with 108 beams and 2112 spectral channels—and that the optimized workflow is fast and cheap enough for the full survey: about A$5,250 and 11 days per field. The demonstration is actual production: four science-ready image cubes were produced and delivered to the survey team, not just benchmark timings. This matters because it gives astronomers an on-demand alternative to waiting in queues at public supercomputing facilities, with a documented recipe for porting an existing astronomical pipeline to a rented platform. The key to the result is optimization: machine-level tuning (socket pinning, RAM-disk scratch space, concurrent imaging instances) and software-level tuning (clean thresholds, Briggs weighting, disabling model updates) together cut the per-channel imaging time by a factor of 4.5.

What carries the argument

The load-bearing mechanism is the joint deconvolution imaging mode of the WSClean software, which stitches the 108 individual beams into one image before applying CLEAN, preserving the large-scale diffuse emission that would be lost if each beam were deconvolved separately. The argument hinges on the measured performance of this software on the commercial provider's heterogeneous node pool, combined with a set of optimizations: assigning memory-hungry tasks to high-memory nodes, pinning processes to sockets, using a RAM disk for temporary files, running four imaging instances per socket, and suppressing file-system writes by disabling model-column updates. These choices, together with a drip-fed data transfer from the archive, produce the reported cost and elapsed-time figures.

What would settle it

Count the actual number of WSClean imaging jobs in the optimized production run: if it is 2112 rather than 528, the imaging node-hours and the per-field cost would be roughly four times larger, invalidating the A$5,250 estimate.

Watch

Extended reading notes

Core claim

The central claim is that renting commercial high-performance computing is a viable way to produce science-ready image cubes for a large H I survey, at a predictable dollar cost and elapsed time. The paper substantiates this by porting a previously developed joint-deconvolution imaging workflow to the commercial platform, optimizing it, and measuring the production cost on one pilot field: roughly A$4,720 in compute plus A$530 in storage, over 11 days, for a cube spanning 2112 channels at 7-arcsecond resolution across 108 beams. The paper further claims that the optimization gains are substantial and repeatable—imaging time per channel fell from 2.9 hours to 39 minutes—and that the resulting cubes serve both as reference images for an upgrade to the telescope's science data processor and as science products for the survey team.

Load-bearing premise

The per-field cost and elapsed-time estimate assumes that the single benchmark field (SBID 38466) is representative of every full-survey field, and that the node-hour counts in the cost table are internally consistent—including the unexplained difference between the stated 2112 native-resolution channels and the 528 imaging jobs listed.

Editorial extensions

If this is right

  • If the estimate holds, the full GASKAP-HI survey can be processed on demand at a known cost and an 11-day cadence per field, allowing budget planning without queue-time uncertainty.
  • The documented porting and optimization steps give other radio surveys a template for moving existing pipelines to commercial HPC, including the future regional centres for the SKA.
  • The WSClean joint-deconvolution cubes provide independent reference products for validating and refining the ASKAPSoft joint-deconvolution upgrade.
  • The observed factor-of-4.5 speedup suggests similar tuning could reduce turnaround times on public supercomputers even when compute is not billed.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the imaging job count discrepancy is real, the per-field cost may be understated by up to a factor of four; the full-survey budget should not be set until the 528-versus-2112 question is resolved.
  • The optimization techniques (socket pinning, RAM-disk scratch, concurrency tuning) are generic enough to transfer to other radio imaging workloads on shared-memory clusters, whether commercial or academic.
  • The cost model is specific to the GASKAP-HI geometry; scaling to other surveys with different beam counts, channel counts, or weighting schemes would require re-measuring per-channel imaging time rather than reusing A$5,250 directly.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper reports a trial of commercial high-performance computing (HPC) for radio-astronomy imaging, processing ASKAP GASKAP-HI pilot survey data on DUG Technology's commercial supercomputer. The authors ported an existing WSClean-based joint-deconvolution workflow, optimized it for the commercial platform through machine- and software-level changes, and produced four image cubes for the science team. They report a resource estimate of approximately A$5,250 per field and an elapsed time of 11 days per field for future full-survey processing, based on timing data from SBID 38466. The paper also documents lessons learned for researchers considering commercial HPC.

Significance. If the resource estimate is reliable, the paper demonstrates that commercial HPC is a practical and affordable option for large-scale radio-astronomy processing, complementing public facilities and informing plans for SKA regional centres. The feasibility claim is well supported by four successfully produced image cubes, and the paper provides openly available code and an imaging guide. The main weakness is that the quantitative resource estimate, a stated sub-objective, rests on Table 4, whose job counts and node-hour accounting are internally inconsistent and not reproducible from the text; correcting this is essential before the cost estimate can be used by the community.

major comments (3)
  1. [§6.1 and Table 4] The paper states in §6.1 that production runs used the full native resolution of 2112 channels of width 1.157 kHz, and §5 says each spectral channel is imaged separately. However, Table 4 row 08 lists only 528 imaging jobs at 9,419 s each, yielding 1,381.4 node-hr. If 2112 channels are imaged separately, the job count should be 2112, which would raise the imaging node-hours to about 5,526 and increase the per-field cost from roughly A$5,250 to roughly A$15,800, assuming all other rows are unchanged. If instead each of the 528 jobs processes four channels, that grouping must be stated explicitly and reconciled with the text's claim of separate per-channel imaging. This inconsistency directly affects the headline resource estimate and must be resolved.
  2. [Table 4 (node-hour accounting)] The node-hour arithmetic in Table 4 is not reproducible without additional information about concurrency and nodes per job. The text says node-hours are the product of time per job and number of jobs, divided by two for dual-socket nodes without socket pinning, but Step 06 lists 36 jobs at 256,283 s each (2,562.8 node-hr) against an elapsed time of 38.9 hr, which implies each job ran about 71 hr unless jobs used multiple nodes or concurrent node allocations; neither case is specified. Similarly, Step 08 lists 528 jobs at 9,419 s each (1,381.4 node-hr) with 97.6 hr elapsed, and Step 01 lists 108 jobs at 2,443 s (73.3 node-hr) with 6.2 hr elapsed, implying very different concurrency levels that are not explained. Provide, for each step, the number of nodes per job, the wall-clock time per job, and the concurrency, so that the node-hour totals and the A$4,720 cost can be verified and the discrepancy between 2112 and 528 imaging jobs can be resolved.
  3. [§7 (storage-cost assumption)] The per-field estimate of A$5,250 includes A$530 of storage cost, which is derived by assuming that 'multiple workflows can be managed at 3 per month' against a peak storage of ~25 TB per cube. This assumption is stated but not justified; if fields are processed one at a time, the storage cost per field would be approximately A$1,600, raising the total to about A$6,320. The sensitivity of the estimate to the field-processing cadence should be quantified or the assumption should be relaxed, as it is a nontrivial component of the headline figure.
minor comments (6)
  1. [Abstract] The phrase 'processing and storages costs' contains a grammatical error ('storages' should be 'storage').
  2. [§6.1] There is a typo in 'because of the the additional processing capacity available' — delete the duplicated 'the'.
  3. [§3] The word 'esentially' should be 'essentially'.
  4. [Table 5] The lesson numbering skips items 16 and 18; renumber for consistency.
  5. [§6.2] The sentence 'the sidelobe structure becomes more apparent as the weighting transitions from uniform (−2.0) to natural (2.0)' is clear, but the sign convention for Briggs robustness (negative = uniform, positive = natural) is not defined; add a brief definition for readers unfamiliar with WSClean.
  6. [§8 and Table 5] The table lists lessons 1–20 but the body text discusses only some of them; consider cross-referencing each lesson number or adding a sentence noting that unlisted lessons were considered lower priority.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the feasibility result rests on measured production runs and actual image cubes, not on fitted inputs or self-citations.

full rationale

The paper's central claims are empirical demonstrations: four GASKAP-HI pilot image cubes were successfully produced on DUG's commercial HPC platform, and the costs and elapsed times were measured during that production. The A$5,250 per-field full-survey estimate is an arithmetic extrapolation from the measured node-hours in Table 4 plus the measured storage high-water mark, so it is anchored in direct observations rather than being a fitted quantity that is then re-predicted. The WSClean imaging parameters were hand-tuned by experimentation, but the feasibility conclusion does not reduce by construction to those parameters; the parameter choices are inputs to an imaging run whose success is judged by the independently produced cubes. Self-citations to Pingel et al. (2022) document the pre-existing workflow that was being ported, and the porting is evidenced by actual completed processing and runtime logs rather than by the citation itself. The internal inconsistency in Table 4 (528 imaging jobs listed versus 2,112 native-resolution channels stated in Section 6.1, and the uneven application of the dual-socket halving rule) is a genuine reproducibility and cost-estimate concern, but it is an arithmetic and accounting issue rather than circularity: the estimate is still derived from measured per-job runtimes and is not defined so as to equal its own conclusion. No load-bearing argument in the paper reduces to a self-citation or to an equation whose output is identical to its input.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

This engineering report introduces no new physical entities. The ledger captures hand-tuned imaging parameters and domain assumptions about provider representativeness, data identity, and cost-model correctness. The internal inconsistency in channel counting means the cost-model assumption is the most fragile entry.

free parameters (4)
  • Briggs robustness = 1.25 (final), 2.0 (single beam)
    Chosen by hand after experimentation to balance noise and resolution; affects image quality, not the feasibility conclusion.
  • Clean threshold = 0.015 Jy (final), 0.021 Jy (initial)
    Hand-tuned to stop cleaning diverging in outer regions; affects the science products and compute time.
  • Number of clean iterations (niter) = 40,000 (final)
    Selected by hand; influences deconvolution depth and runtime.
  • Clean mask threshold = 15% of peak joint beam response
    Chosen to exclude the noisy outer zone from cleaning; affects stability of the run.
assumptions (4)
  • domain assumption The measured performance and costs at DUG are representative of the commercial HPC market.
    The paper generalizes from a single provider to the commercial sector in the Introduction and Discussion.
  • domain assumption All full-survey fields have processing requirements similar to SBID 38466.
    The resource estimate in Section 7 extrapolates from one field to the full survey.
  • domain assumption The visibility data in CASDA are identical to those used in the earlier ANU processing, so image comparison validates the port.
    Section 5 uses SBID 38215 comparison to establish confidence; relies on identical data.
  • domain assumption The node-hour billing model (per node, with a division rule) is applied correctly in the cost accounting.
    Table 4's note defines the rule, but its application to imaging is not transparent, and the central cost estimate depends on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Processing of GASKAP-HI pilot survey data using a commercial supercomputer." pith.science (2026). https://pith.science/paper/255UU56C

@misc{pith2026241117118,
  author       = {Pith},
  title        = {Pith review of: Processing of GASKAP-HI pilot survey data using a commercial supercomputer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/255UU56C}},
  note         = {Machine review of arXiv:2411.17118}
}
read the original abstract

Modern radio telescopes generate large amounts of data, with the next generation Very Large Array (ngVLA) and the Square Kilometre Array (SKA) expected to feed up to 292 GB of visibilities per second to the science data processor (SDP). However, the continued exponential growth in the power of the world's largest supercomputers suggests that for the foreseeable future there will be sufficient capacity available to provide for astronomers' needs in processing 'science ready' products from the new generation of telescopes, with commercial platforms becoming an option for overflow capacity. The purpose of the current work is to trial the use of commercial high performance computing (HPC) for a large scale processing task in astronomy, in this case processing data from the GASKAP-HI pilot surveys. We delineate a four-step process which can be followed by other researchers wishing to port an existing workflow from a public facility to a commercial provider. We used the process to provide reference images for an ongoing upgrade to ASKAPSoft (the ASKAP SDP software), and to provide science images for the GASKAP collaboration, using the joint deconvolution capability of WSClean. We document the approach to optimising the pipeline to minimise cost and elapsed time at the commercial provider, and give a resource estimate for processing future full survey data. Finally we document advantages, disadvantages, and lessons learned from the project, which will aid other researchers aiming to use commercial supercomputing for radio astronomy imaging. We found the key advantage to be immediate access and high availability, and the main disadvantage to be the need for improved HPC knowledge to take best advantage of the facility.

Figures

Figures reproduced from arXiv: 2411.17118 by the authors.

Figure 1
Figure 1. Fields in the GASKAP-Hi Pilot Surveys. Cyan boxes denote the approximate ASKAP PAF footprint for the pilot phase I fields (20hr integration); yellow boxes are those for the pilot phase II fields (10hr integration) RJS provides a simple method for jobs to be executed in par￾allel, by supplying values for variables in an external schema file. The operating system kernel was CentOS Linux release 7.7.1908, and RJS jobs … view at source ↗
Figure 2
Figure 2. Comparison of dirty images from ASKAPSoft (top row) and WSClean (bottom row) with di [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Portion of the cleaned (but not feathered) image cube from SBID [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Gantt chart for non-trivial workflow steps, SBID 38466. Light blue bars are KNL processing, dark purple bars are CLX processing. [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 10 canonical work pages

  1. [2]

    GASKAP-HI Pilot Survey Science III: An unbiased view of cold gas in the Small Magellanic Cloud

    GASKAP-HI Pilot Survey Science III: An un- biased view of cold gas in the Small Magellanic Cloud. PASA 39, e034. doi:10.1017/pasa.2022.18, arXiv:2204.06285. Dickey, J.M., Dempsey, J.M., Pingel, N.M., McClure-Griffiths, N.M., Jameson, K., Dawson, J.R., D´enes, H., Clark, S.E., Joncas, G., Leahy, D., Lee, M.Y ., Miville-Deschˆenes, M.A., Stanimirovi´c, S., ...

    work page Pith review arXiv 2022

  2. [8]

    URL: https://aussrc.org/ wp-content/uploads/2022/12/DSPFinalReport%5F29112022.pdf

    Australian SKA Regional Cen- tre Design Study Program: Final Report. URL: https://aussrc.org/ wp-content/uploads/2022/12/DSPFinalReport%5F29112022.pdf. McClure-Griffiths, N.M., Pisano, D.J., Calabretta, M.R., Ford, H.A., Lockman, F.J., Staveley-Smith, L., Kalberla, P.M.W., Bailin, J., Dedes, L., Janowiecki, S., Gibson, B.K., Murphy, T., Nakanishi, H., New...

    work page 2022

  3. [10]

    (Eds.), Ground-based and Airborne Telescopes IX, p

    The Square Kilometre Array project up- date, in: Marshall, H.K., Spyromilio, J., Usuda, T. (Eds.), Ground-based and Airborne Telescopes IX, p. 121820Q. doi:10.1117/12.2642184. Nguyen, H., McClure-Gri ffiths, N.M., Dempsey, J., Dickey, J.M., Lee, M.Y ., Lynn, C., Murray, C.E., Stanimirovi ´c, S., Busch, M.P., Clark, S.E., Daw- son, J.R., D´enes, H., Gibson...

    work page doi:10.1117/12.2642184

  4. [11]

    Local HI Absorption towards the Magellanic Cloud foreground using ASKAP

    Local H I absorption towards the magellanic cloud foreground using ASKAP. MNRAS 534, 3478–3497. doi:10.1093/mnras/stae2274, arXiv:2409.20311. Nikolic, B., SDP Consortium, S.,

    work page Pith review arXiv

  5. [14]

    GASKAP-HI Pilot Survey Science I: ASKAP Zoom Observations of HI Emission in the Small Magellanic Cloud

    GASKAP-HI pilot survey science I: ASKAP zoom ob- servations of HI emission in the Small Magellanic Cloud. PASA 39, e005. doi:10.1017/pasa.2021.59, arXiv:2111.05339. Pingel, N.M., Ma, Y .K.,

    work page Pith review arXiv 2021

  6. [15]

    URL: https://github.com/nipingel/GASKAP_Imaging/blob/master/ docs/GASKAP_Imaging.pdf

    The GASKAP-HI Imaging Pipeline. URL: https://github.com/nipingel/GASKAP_Imaging/blob/master/ docs/GASKAP_Imaging.pdf. online; accessed 23-October-2024. Sault, R.J., Teuben, P.J., Wright, M.C.H.,

    work page 2024

  7. [18]

    02083, arXiv:2408.02083

    The Commensal Real-time ASKAP Fast Transient incoherent-sum survey , arXiv:2408.02083doi: 10.48550/arXiv.2408. 02083, arXiv:2408.02083. SKA,

    arXiv doi:10.48550/arxiv.2408.02083

  8. [118]

    Lee-Wadell, K., Pallott, D., German, G., Null, D., Shen, A., Aniruddha, G., Williamson, A., Holmes, K.,

    doi:10.1007/s10509-020-03831-4 , arXiv:2002.07311. Lee-Wadell, K., Pallott, D., German, G., Null, D., Shen, A., Aniruddha, G., Williamson, A., Holmes, K.,

    arXiv 2002

Show all 19 references
  1. [186]

    doi: 10.3847/1538-4357/ ac3a89, arXiv:2111.04545. Dickey, J.M., McClure-Griffiths, N., Gibson, S.J., G´omez, J.F., Imai, H., Jones, P., Stanimirovi´c, S., Van Loon, J.T., Walsh, A., Alberdi, A., Anglada, G., Uscanga, L., Arce, H., Bailey, M., Begum, A., Wakker, B., Bekhti, N.B...

    arXiv

  2. [433]

    astro-ph/0612759, arXiv:astro-ph/0612759

    doi:10.48550/arXiv. astro-ph/0612759, arXiv:astro-ph/0612759. Schinckel, A.E., Bunton, J.D., Cornwell, T.J., Feain, I., Hay, S.G.,

    doi arXiv

  3. [2009]

    Gass: The Parkes Galactic All-Sky Survey. I. Survey Description, Goals, and Initial Data Release. ApJS 181, 398–412. doi: 10.1088/0067-0049/ 181/2/398, arXiv:0901.1159. McMullin, J., Diamond, P., Caiazzo, M., Casson, A., Cheetham, T., Dewdney, P., Laing, R., Lewis, B., Schinck...

    arXiv

  4. [2011]

    Data Challenges for Next-generation Radio Telescopes, in: Sixth IEEE International Conference on eScience, pp. 21–24. doi:10.1109/ eScienceW.2010.13, arXiv:1101.1355. Offringa, A.R., McKinley, B., Hurley-Walker, N., Briggs, F.H., Wayth, R.B., Kaplan, D.L., Bell, M.E., Feng, L....

    2010 arXiv

  5. [2012]

    (Eds.), Ground-based and Airborne Telescopes IV , p

    The Australian SKA Pathfinder, in: Stepp, L.M., Gilmozzi, R., Hall, H.J. (Eds.), Ground-based and Airborne Telescopes IV , p. 84442A. doi: 10.1117/12. 926959. Shannon, R.M., Bannister, K.W., Bera, A., Bhandari, S., Day, C.K., Deller, A.T., Dial, T., Dobie, D., Ekers, R.D., Fon...

    doi

  6. [2013]

    PASA 30, e003

    GASKAP-The Galactic ASKAP Survey. PASA 30, e003. doi: 10.1017/pasa.2012.003, arXiv:1207.0891. Guzman, J.C., Marquarding, M.,

    2012 arXiv

  7. [2014]

    MNRAS 444, 606–619

    WSCLEAN: an implementation of a fast, generic wide-field imager for radio astronomy. MNRAS 444, 606–619. doi:10.1093/mnras/stu1368, arXiv:1407.1943. Pingel, N.M., Dempsey, J., McClure-Gri ffiths, N.M., Dickey, J.M., Jameson, K.E., Arce, H., Anglada, G., Bland-Hawthorn, J., Bre...

    1943 arXiv

  8. [2020]

    Science Bulletin 65, 337–339

    SKA shakes hands with Sum- mit. Science Bulletin 65, 337–339. doi: 10.1016/j.scib.2019.12.016, arXiv:1912.12591. Wieringa, M., Raja, W., Ord, S.,

    2019 arXiv

  9. [2021]

    system description

    Aus- tralian square kilometre array pathfinder: I. system description. PASA 38, e009. doi: 10.1017/pasa.2021.1, arXiv:2102.01870. Huynh, M., Dempsey, J., Whiting, M.T., Ophel, M.,

    2021 arXiv

  10. [2022]

    PASP 134, 114501

    CASA, the Common Astronomy Software Applications for Radio Astronomy. PASP 134, 114501. doi: 10.1088/1538-3873/ac9642, arXiv:2210.02276. Dempsey, J., McClure-Griffiths, N.M., Murray, C., Dickey, J.M., Pingel, N.M., Jameson, K., D ´enes, H., van Loon, J.T., Leahy, D., Lee, M.Y ...

    arXiv

  11. [2024]

    URL: https://github.com/IanKempICRAR/GASKAP-DUG/

    GASKAP-HI DUG Imag- ing Guide. URL: https://github.com/IanKempICRAR/GASKAP-DUG/. online; accessed 23-October-2024. Koribalski, B.S., Staveley-Smith, L., Westmeier, T., Serra, P., Spekkens, K., Wong, O.I., Lee-Waddell, K., Lagos, C.D.P., Obreschkow, D., Ryan-Weber, E.V ., Zwaan...

    2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.