{"id":"1ca4afaa-40ae-4bbd-b12f-f384c398a11c","arxiv_id":"2501.02317","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Kratos is a new GPU-optimized astrophysical hydrodynamics framework using mixed precision and stream-aware communication, verified on standard shock and instability benchmarks.","lead":"A new GPU-based astrophysical simulation framework called Kratos is described, using mixed precision to run hydrodynamics fast on consumer graphics cards while aiming to keep conservation laws accurate. The paper reports standard benchmark tests showing accuracy and speed, but the code is not yet publicly released.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Machine-accuracy conservation claim in §3.2.2 and the abstract is asserted, not demonstrated: no conservation-error diagnostic appears in §4, and the <1e-3 mixed-vs-double error is explicitly marked as not shown.","rationale":"The reader correctly identified that the 10^-3 mixed-precision error bound is asserted but not shown and that code unavailability blocks independent verification. My stress-test focuses more narrowly on the conservation half of the central claim: the paper never presents any conserved-quantity drift diagnostic, even though 'machine accuracy' for conservation laws is the property advertised in the abstract. This is a missing-evidence concern rather than an internal inconsistency: the algorithms described are standard, the benchmark results are broadly consistent with an established Godunov scheme, and the performance tables are plausible, so the framework itself is not suspect. The concrete check proposed would settle the issue by producing the missing diagnostic; if it passes, the conditional acceptance can be upgraded, and if it fails, the headline claim must be weakened. Because this concern reinforces the reader's CONDITIONAL verdict rather than overturning it, I recommend no change to the verdict.","tokens_in":786,"tokens_out":2004,"duration_ms":49149,"concrete_test":"Instrument the Kratos hydrodynamics module to track total mass, total momentum, and total energy (including all blocks and refinement-boundary flux contributions) at every step, and compute the relative drift from initial values for the Sod shock tube, double Mach reflection, Kelvin-Helmholtz, Rayleigh-Taylor, and outflow tests in both mixed and full double precision. If the relative drift is not at the ~1e-12 level characteristic of double-precision accumulation, or if it grows secularly rather than oscillating at roundoff level, the machine-accuracy conservation claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's headline property is that the mixed-precision hydrodynamics module holds conservation laws to machine accuracy. In a finite-volume Godunov scheme, exact discrete conservation requires the same interface flux to enter adjacent cell updates with opposite signs, and at refinement boundaries the fine-side fluxes to be summed onto the coarse side consistently (§2.3.2, step 4). The paper states that conservative variables are stored as double-precision float2_t while fluxes are computed as single-precision float_t (§3.2.2), but it never reports a direct measurement of total mass, momentum, or energy drift. The phrase 'holds the conservation laws to machine accuracy' is therefore unsupported by any shown diagnostic. Furthermore, the quantitative accuracy claim in §3.2.2 — relative error below 10^-3 versus double precision over the relevant astrophysical parameter space — is explicitly accompanied by '(not shown in this paper)'. The only full-solution mixed-versus-double comparison shown, the double Mach reflection in Figure 9, reports a maximum relative density discrepancy of about 10^-2, an order of magnitude larger than 10^-3, even if chaotic evolution may amplify differences. Finally, the code is not released, so this central advertised behavior cannot be independently checked. Collectively, the leading numerical claim of the paper rests on an unshown diagnostic and an unrunable code.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents Kratos, a GPU-oriented framework for astrophysical simulations, and describes its fundamental infrastructure (device abstraction layer, multiprocessing communication model, tree-based mesh management with refinement, and module container) together with a hydrodynamics module built on PLM reconstruction, the HLLC Riemann solver, and a second-order Heun time integrator. The hydrodynamics module uses a mixed-precision strategy in which conservative variables are stored in double precision while reconstruction and flux calculations are carried out in single precision. The paper verifies the implementation with convergence tests, Sod shock tubes, double Mach reflection, Liska-Wendroff implosion, Kelvin-Helmholtz and Rayleigh-Taylor instabilities, and colliding outflow tests, and reports performance measurements on a range of CPUs and GPUs. The abstract claims that the mixed-precision method maintains conservation laws to machine accuracy, and Section 3.2.2 further claims a relative error below 10^-3 versus double precision across the relevant astrophysical parameter space, with the latter explicitly marked as not shown in this paper.","tokens_in":21675,"tokens_out":5432,"duration_ms":54733,"significance":"If the claims are substantiated, Kratos would be a useful addition to the set of GPU-accelerated astrophysical hydrodynamics codes, and the detailed description of the device abstraction, asynchronous communication, and mixed-precision design could be of value to the community. The paper's benchmarks are standard and mostly well chosen: convergence orders are reported, shock-tube solutions are compared with semi-analytic results, and the double Mach reflection and instability tests follow established references. The performance tables, especially the consumer-GPU comparisons, are informative. However, the two most distinctive advertised properties—machine-accurate conservation and a 10^-3 mixed-versus-double error bound—are not demonstrated by any diagnostic shown in the manuscript, and the code is not publicly released. The paper is therefore a plausible technical description, but it currently overstates its central quantitative claims.","major_comments":[{"comment":"The abstract's headline assertion that the mixed-precision method holds the conservation laws to machine accuracy is not supported by any measurement shown in §4. Discrete conservation in a finite-volume Godunov scheme depends on the flux bookkeeping across cell interfaces and, with refinement, on the fine-to-coarse flux summation described in §2.3.2; this is precisely a property that should be verified numerically. I ask the authors to add a direct conservation diagnostic, such as the time evolution of total mass, momentum, and total energy (absolute or relative drift) for at least one two-dimensional test and one AMR test, comparing mixed and double precision. Without such a measurement, the headline claim is an assertion rather than a verified property.","section":"Abstract and §3.2.2"},{"comment":"The quantitative claim that the mixed-precision HLLC solver has relative error below 10^-3 versus double precision 'across the parameter space relevant to typical astrophysical hydrodynamic simulations' is explicitly labelled '(not shown in this paper)'. This is a load-bearing advertised property and should either be demonstrated (for example, with a parameter sweep over Mach number, density contrast, or a representative set of initial conditions) or replaced by a more limited statement. The only full-solution mixed-versus-double comparison shown, the double Mach reflection in Figure 9, reports a maximum relative density discrepancy of roughly 10^-2, an order of magnitude larger than 10^-3; even if chaotic amplification explains this, the paper should quantify the argument or narrow the claim.","section":"§3.2.2"},{"comment":"The Liska-Wendroff test shows a noticeable departure from reflection symmetry in the mixed-precision run, while the double-precision run preserves the symmetry much better. Since this is a concrete accuracy consequence of the mixed-precision mode, the abstract and §3.2.2 should state this limitation explicitly. Presenting mixed precision only as 'accurate and fast' overstates the method's fidelity, especially because the symmetry breaking appears in a standard test that the paper itself uses to assess the solver's directional symmetry.","section":"§4.4"},{"comment":"The manuscript states that Kratos will be provided only upon request and collaboration, so none of the implementation details or numerical claims can be independently reproduced by a reader. For a methods/software paper whose central quantitative claims are not fully documented in the text, the lack of a public release (or at least a clearly specified archival release plan with benchmark inputs) is a significant obstacle to verification. At minimum, the authors should make the hydrodynamics module and the benchmark configurations available, or explicitly state which parts can be shared.","section":"Code availability"}],"minor_comments":[{"comment":"The manuscript contains numerous typographical errors and misspellings (e.g., 'foundamental', 'architechtures', 'parallerizing', 'descrete', 'primative', 'boudnary', 'ususally', 'handeled', 'accompished', 'stucture', and 'MUSCL-Handcock' in the Ramses-GPU reference); these should be corrected in a revision.","section":"Throughout"},{"comment":"The colorbar labels in the bottom panel appear as '10□6', '10□4', and '10□2', which is likely a rendering artifact; the power-of-ten notation should be formatted properly.","section":"Figure 9"},{"comment":"In the paragraph introducing the van Leer integrator, there is a stray superscript '1' after 'does not'; this appears to be an editing artifact and should be removed.","section":"§3.3"},{"comment":"The informal reference 'H., Yang and L. Wang, in prep.' should be formatted in the standard style, e.g., 'Yang & Wang, in prep.'.","section":"§2.4"},{"comment":"The numerical diffusivity scaling η ∝ ∆x^2/∆t is invoked to explain resolution-dependent differences in the instability tests, but no coefficient or reference is given; a quantitative or cited estimate would make the comparison more convincing.","section":"§4.5 and §4.6"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of an astronomical methods/software journal, and the technical infrastructure is described in considerable detail. My main concern is that the two headline numerical properties—machine-accurate conservation and the 10^-3 mixed-versus-double error bound—are not demonstrated in the manuscript, and the second is explicitly marked as not shown. The authors should be asked to supply the missing diagnostics in revision. I would also encourage the editor to treat the code availability statement as a serious issue: a framework paper whose code is available only on request cannot be fully evaluated by reviewers or readers, and some journals would require at least a frozen archival version for the claims to be reproducible."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe Kratos paper is a real code paper, not vaporware. What's genuinely new is the device abstraction layer that sits directly on CUDA/HIP without Kokkos, the stream-aware MPI communication that hides transfers behind computation, and the Hilbert-curve load balancing via L-systems. The mixed-precision split (double for conservative variables, single for fluxes) is a sensible design, and the benchmark suite is standard but thorough: convergence tests, five Toro shock tubes, double Mach reflection, Liska-Wendroff, Kelvin-Helmholtz, Rayleigh-Taylor, and colliding outflows. The convergence rates align with the expected second order, and the mixed-vs-double comparisons in the Mach and RT tests show differences at the few-percent level, which is credible for chaotic flows. The honesty about symmetry breaking in the Liska-Wendroff test (mixed precision breaks reflection symmetry) is a point in their favor.\n\nThe soft spots are exactly where the reader pointed. The abstract and §3.2.2 claim that the mixed-precision method 'holds the conservation laws to machine accuracy,' and the paper asserts a relative error below 1e-3 versus double precision over 'the relevant astrophysical parameter space' — but that quantitative claim is explicitly marked '(not shown in this paper).' No conservation-error diagnostic appears anywhere in §4. The only full mixed-vs-double comparison shown, the double Mach reflection, reports a maximum density discrepancy around 1e-2, an order of magnitude larger than the 1e-3 claim. That doesn't mean the claim is false — chaotic flows amplify small differences — but it means the paper doesn't support its headline. And the code is not released, so nobody can check.\n\nThe performance tables are useful but thin: averages over 102 steps with no variance, no power measurements, no detailed configuration. That is a minor issue relative to the missing conservation diagnostic.\n\nBottom line: the framework is plausible and the tests show a working hydrodynamics code. The paper deserves a serious referee, but the referee should demand either the conservation diagnostic and the 1e-3 figure, or a softened claim. I would engage with the infrastructure ideas regardless, but I would not cite the paper until the code is public and the central claim is backed by a direct measurement.","headline":"A genuine GPU hydrodynamics code paper whose infrastructure ideas deserve attention, but the headline mixed-precision conservation claim is asserted, not demonstrated, and the code is not released.","tokens_in":22205,"tokens_out":1930,"would_cite":false,"duration_ms":19218,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Kratos is a GPU-native astrophysical hydrodynamics framework whose mixed-precision scheme claims $10^{-3}$ flux accuracy while conserving mass, momentum, and energy to machine precision.","keywords":["GPU computing","astrophysical hydrodynamics","mixed precision","finite volume methods","HLLC Riemann solver","adaptive mesh refinement","heterogeneous computing","Godunov scheme"],"falsifier":"Run the mixed-precision solver through a systematic parameter sweep of the HLLC solver, especially strong shocks with $|s_i - u_i| \\ll |u_i|$ where the subtraction-safe contact-speed formula is stressed, and compare cell-by-cell fluxes against the double-precision solver; any regime inside the claimed parameter space with relative error above $10^{-3}$ would falsify the accuracy claim. Likewise, evolve a long-duration test and monitor total mass, momentum, and energy: if any conserved quantity drifts beyond the machine-precision level of the double-precision accumulator, the conservation claim fails.","tokens_in":21193,"feed_emoji":"⚡","tokens_out":10016,"duration_ms":91276,"temperature":0.7,"pith_summary":"Kratos is a GPU-native simulation framework built for astrophysics, and this paper presents its foundational infrastructure plus the hydrodynamics module as the first working example. The core claim is that a mixed-precision scheme—storing the conserved variables in double precision but computing reconstruction and Riemann fluxes in single precision—lets hydrodynamics run at near-single-precision speed on consumer GPUs while keeping mass, momentum, and energy conservation at machine accuracy. The paper reports that the mixed-precision fluxes deviate from full-double-precision results by less than one part in a thousand in relative error across the astrophysical parameter space, and verifies the solver on standard benchmarks from Sod shock tubes to Kelvin–Helmholtz and Rayleigh–Taylor instabilities. If the claim holds, production-grade astrophysical hydrodynamics can be done on ordinary graphics cards instead of expensive compute GPUs, with a roughly five-to-seven-fold speedup over double-precision runs. The framework is designed so that further modules, such as MHD, thermochemistry, self-gravity, and particle solvers, can be attached to the same infrastructure.","feed_headline":"Mixed precision hydro runs 5-7x faster with machine-exact conservation","feed_subtitle":"A new astrophysics code computes fluxes in single precision yet keeps conservation laws at machine accuracy.","key_machinery":"The load-bearing mechanism is the mixed-precision split: conservative variables are stored in a double-precision accumulator type (`float2_t`), while all reconstruction and Riemann-flux work uses a single-precision working type (`float_t`). This means the time-consuming per-cell flux calculation runs at single-precision speed on GPUs, while the conserved quantities updated each step accumulate in double precision, giving conservation to machine accuracy. The companion piece is the subtraction-safe HLLC contact-speed evaluation, which replaces the direct computation of $s_*$ with the always-sign-definite differences $\\Delta s_L \\equiv s_L - u_L = -c_{s,L}q$ and $\\Delta s_R \\equiv s_R - u_R = c_{s,R}q$, preventing the vanishing denominators and catastrophic cancellation that single precision can produce. Around this core, the paper constructs stream-aware asynchronous MPI communication and a $2^d$-tree mesh manager with Hilbert-curve load balancing, which together let the single-precision flux calculations hide communication costs behind computation.","core_discovery":"The central claim is that a Godunov finite-volume hydrodynamics solver can run at single-precision speed without sacrificing the conservation properties that make it trustworthy. Kratos stores conservative variables—density, momentum density, and total energy density—as double-precision `float2_t` quantities, so every update accumulates at machine precision, while the PLM reconstruction and the HLLC Riemann solver operate on single-precision `float_t` data. A subtraction-safe form of the HLLC contact-wave speed, $\\Delta s_L = -c_{s,L}q$ and $\\Delta s_R = c_{s,R}q$, removes the catastrophic-cancellation failure mode that single-precision evaluation of the contact speed would otherwise encounter. The paper reports that extensive testing puts the relative error of these mixed-precision fluxes below $10^{-3}$ compared with full double precision across the parameter space of typical astrophysical hydrodynamics, and the benchmark suite demonstrates that mixed-precision results are visually and quantitatively close to double precision, with relative density differences $\\lesssim 10^{-2}$ in the double Mach reflection test.","pith_inferences":["If the $10^{-3}$ relative-error bound holds everywhere the paper claims, then precision ceases to be a reason to buy double-precision compute GPUs for most astrophysical hydrodynamics, and large parameter scans on consumer cards become practical; this is an editorial inference, since the paper does not show the parameter-space survey.","The paper's own Liska-Wendroff and outflow tests show mixed precision can break discrete reflection symmetry in chaotic flows; a natural follow-up would be to characterize how quickly the symmetry is lost as a function of resolution and perturbation amplitude.","Publishing the omitted error survey and releasing the code would let others verify the conservation-to-machine-accuracy claim on their own hardware; without that, the central claim cannot be independently checked."],"forward_implications":["A single consumer GPU can sustain $10^8$ to $10^9$ cells per second in mixed precision on the double Mach reflection problem, about 5 to 7 times faster than the same device in full double precision.","Because conserved variables remain double precision, the conservation laws are held to machine accuracy even though the flux computation itself is single precision.","The device-abstraction layer and module-container design allow the same infrastructure to host MHD, thermochemistry, self-gravity, and particle modules, with the hydrodynamics module as the template.","On two GPUs, stream-aware communication achieves up to about 90 percent of ideal parallel efficiency when single-device speeds are moderate; the paper reports that scaling degrades when a very fast GPU can no longer hide communication behind computation.","Benchmarks reproduce the analytic shock-tube solutions, the double Mach reflection reference, and second-order convergence for smooth waves, with mixed-precision results nearly identical to full double precision."],"supporting_citations":[{"why":"Supplies the HLLC Riemann solver formulation and the standard shock-tube test cases used to validate the mixed-precision solver against analytic solutions.","marker":"Toro (2009)"},{"why":"Provides the CPU performance baseline (about $3 \\times 10^6$ cells per second per core) and the reference Godunov method whose convergence behavior Kratos is compared against.","marker":"Stone et al. (2020)"},{"why":"Defines the double Mach reflection test and the reference solution the paper reproduces at higher resolution.","marker":"Woodward & Colella (1984)"},{"why":"Defines the implosion test used to assess directional symmetry preservation, including how jet symmetry diagnoses the numerical method.","marker":"Liska & Wendroff (2003)"},{"why":"Supplies the initial-condition setups for the Kelvin-Helmholtz and Rayleigh-Taylor instability tests, including perturbation forms and parameter choices.","marker":"Lecoanet et al. (2016)"},{"why":"Supplies the expected $L_1 \\propto N^{-2}$ convergence behavior for the same finite-volume reconstruction and Riemann-solver family, used as a comparison for Kratos's measured convergence.","marker":"Stone et al. (2008)"}],"fun_headline_variants":["Kratos: GPU hydro at 5-7x speed, conservation intact","Single precision flux, double precision state: 5-7x faster hydro","Mixed-precision hydro: 5-7x speedup, machine-exact conservation","Kratos framework: GPU hydro with exact conservation","Run hydro at 5-7x speed, keep conservation exact"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim rests on the assumption that single-precision HLLC flux calculations have relative error below $10^{-3}$ versus double precision across the full parameter space of typical astrophysical hydrodynamics, and that storing conserved variables in double precision keeps conservation laws at machine accuracy; the paper states the supporting tests are not shown, and the code is not released for independent checking.","fun_headline_variants_meta":{"raw":{"variants":["Kratos: GPU hydro at 5-7x speed, conservation intact","Single precision flux, double precision state: 5-7x faster hydro","Mixed-precision hydro: 5-7x speedup, machine-exact conservation","Kratos framework: GPU hydro with exact conservation","Run hydro at 5-7x speed, keep conservation exact"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000397,"raw_usage":{"total_tokens":2068,"prompt_tokens":926,"completion_tokens":1142,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":542,"completion_tokens_details":{"reasoning_tokens":1047}},"tokens_in":542,"tokens_out":1142,"duration_ms":8514,"temperature":1.0,"reasoning_tokens":1047,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:13:18.025610+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the mixed-precision solver through a systematic parameter sweep of the HLLC solver, especially strong shocks with $|s_i - u_i| \\ll |u_i|$ where the subtraction-safe contact-speed formula is stressed, and compare cell-by-cell fluxes against the double-precision solver; any regime inside the claimed parameter space with relative error above $10^{-3}$ would falsify the accuracy claim. Likewise, evolve a long-duration test and monitor total mass, momentum, and energy: if any conserved quantity drifts beyond the machine-precision level of the double-precision accumulator, the conservation claim fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the HLLC Riemann solver formulation and the standard shock-tube test cases used to validate the mixed-precision solver against analytic solutions."},{"cited_title":"1984, Journal of Computational Physics, 54, 115","cited_arxiv_id":null,"evidence_quote":"Defines the double Mach reflection test and the reference solution the paper reproduces at higher resolution."},{"cited_title":"2003, SIAM Journal on Scientific Computing, 25, 995","cited_arxiv_id":null,"evidence_quote":"Defines the implosion test used to assess directional symmetry preservation, including how jet symmetry diagnoses the numerical method."},{"cited_title":"2016, , 455, 4274","cited_arxiv_id":null,"evidence_quote":"Supplies the initial-condition setups for the Kelvin-Helmholtz and Rayleigh-Taylor instability tests, including perturbation forms and parameter choices."},{"cited_title":"M., Gardiner , T","cited_arxiv_id":null,"evidence_quote":"Supplies the expected $L_1 \\propto N^{-2}$ convergence behavior for the same finite-volume reconstruction and Riemann-solver family, used as a comparison for Kratos's measured convergence."}],"review_version":1}