- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper proposes TiFuse for cloud workload forecasting. It combines a frequency branch that selects energy-dominant DCT components with a temporal branch that selects historical patches from a multiscale representation while retaining recent patches. The resulting tokens are processed by relation-enhanced attention, which adds a learned global relational prior to sample-dependent attention under a binary mask. Experiments cover CPU, memory, and GPU forecasting on five cloud traces, with results averaged over four forecasting horizons. TiFuse reports the lowest average MSE in 15 of 16 dataset–resource settings and competitive training time and memory consumption.
Critical strengths
-
The method addresses a practical forecasting problem and makes a sensible distinction between recurring workload patterns and recent dynamics. In particular, retaining recent patches alongside selected historical patterns reduces the risk of losing information near the forecasting boundary. The DCT branch also provides a compact representation without requiring an inverse transform.
-
The experiments cover several real cloud traces and different resource types. TiFuse performs well against both general forecasting methods and a workload-specific baseline, FP-MLP. The training-time and memory comparisons are also useful. For example, Table 5 shows memory consumption close to PatchTST and PatchMixer, although these models are generally faster. Together, these results suggest a useful balance between forecasting accuracy and computational cost.
Critical weaknesses
-
The supplied implementation does not fully match the stated method and experimental protocol. In DAFC.py, spectral energy is averaged over the batch, variables, and within-patch positions, producing one shared set of frequency indices. This differs from the input-specific selection described in Eqs. (4)–(6), and means that a sample’s representation can change depending on the other samples in its inference batch. The default selection also forces the DC component into the frequency budget even when it is outside the energy top-, so the unconstrained maximum-energy property in Appendix A does not directly apply to this implementation. Separately, Dataset_Custom assigns 70% of the data to training and 20% to testing, leaving approximately 10% for validation, whereas Appendix B specifies 7:2:1 for training, validation, and testing. These differences need to be reconciled with the configurations used to produce the reported results.
-
The central claim of stable fusion is not directly demonstrated. The attention visualizations in the main paper and Appendix C show increased matrix rank and different connection patterns, but neither establishes robustness to noise or workload shifts. Adding a learned matrix can itself increase rank. The Ori-Attn ablation supports a forecasting benefit, but removes the prior and mask together, leaving their individual roles unclear. The shared prior also operates over energy-rank and selection-channel positions whose underlying frequencies and historical locations vary across samples. This design may work, but the current evidence does not explain why it provides stable cross-domain relationships. The prediction plots are useful illustrations, though they do not provide a controlled comparison of this claim.
Questions for authors
-
Was DAFC evaluated using per-sample frequency selection or the batch-and-variable-averaged selection in the released code? If the latter, how sensitive are predictions to inference batch size and batch composition? Was the DC component always forced into the selected set?
-
Which split was used for the reported results: the 7:2:1 split described in Appendix B or the approximately 7:1:2 split implemented in Dataset_Custom? Please provide the exact configurations or commands used for the main experiments.
-
Appendix B states that results are reported as means and sample standard deviations over five runs, with ten runs for the Usage CPU comparison. However, Tables 6–8 show only point estimates. Please provide the standard deviations, especially for the close comparisons with FP-MLP.
-
Can the authors separate the effects of the global prior and the sparse mask, and provide a direct comparison under workload shifts or input perturbations? How does the prior remain useful when a token position corresponds to different frequencies or historical patches across samples?
4 comments
2
The original manuscript incorrectly reported a data split ratio of 7:2:1. Both the code and the experiments used 7:1:2 throughout. We have corrected the typographical error in the revised manuscript and updated the detailed experiment configuration in the code.
3
We thank the reviewer for pointing out this omission. Tables 6–8 in the revised manuscript now report means and sample standard deviations over five runs. The Usage-CPU comparison is based on ten runs.
4
The global prior supplements the estimates of interaction weights derived from the current window alone. It acts as a memory bank learned during training, storing interaction preferences between token positions that are shared across samples. The input-dependent mask determines where the prior is applied. When a connection has a high enough matching score to enter the candidate set for the current window, the model adds the corresponding prior to the base attention score. The current window thus determines which time–frequency connections receive a prior adjustment, and the learned prior modulates their weights during information aggregation.
The global prior encodes learned interaction preferences for pairs of token positions, even when the same positions represent different frequencies or historical segments across samples. For example, the same pair of positions may represent two low-frequency components in sample A and two different historical segments in sample B. The pair shares the same prior value across both samples, while the model recomputes its matching score from the token representations in each window. If the pair enters the candidate set only in sample A, the prior adjusts the attention weight for that connection only in sample A. If the pair enters both candidate sets, the prior adds the same bias in both samples. The final weights can still differ because the base attention scores can differ. The model can therefore apply learned interaction preferences to connections supported by the current input even when the time–frequency content at a given position varies across samples.
To isolate the effects of the components, we added a control that retains the global prior with its input-dependent mask removed. Using the same original inputs, we compared forecasting errors and their increases across the three model variants.
| Dataset | Horizon H | Standard attention MSE | Standard attention MAE | Unmasked prior MSE | Unmasked prior MAE | Full TiFuse MSE | Full TiFuse MAE |
|---|---|---|---|---|---|---|---|
| Alibaba | 1 | 0.09309 | 0.17196 | 0.09271 | 0.16699 | 0.09052 | 0.16157 |
| Alibaba | 5 | 0.27442 | 0.31824 | 0.27325 | 0.31541 | 0.27370 | 0.31328 |
| Alibaba | 10 | 0.41645 | 0.42029 | 0.40992 | 0.41284 | 0.40159 | 0.40563 |
| Alibaba | 30 | 0.61855 | 0.51399 | 0.60864 | 0.51087 | 0.59213 | 0.50775 |
| Google2019 | 1 | 0.00819 | 0.06388 | 0.00799 | 0.06081 | 0.00700 | 0.05774 |
| Google2019 | 5 | 0.03968 | 0.13891 | 0.03859 | 0.13622 | 0.03743 | 0.13352 |
| Google2019 | 10 | 0.07875 | 0.19944 | 0.07739 | 0.19664 | 0.07384 | 0.19195 |
| Google2019 | 30 | 0.20946 | 0.34104 | 0.20859 | 0.33991 | 0.20742 | 0.33692 |
Thank you for clarifying the batch-level frequency selection and the 7:1:2 data split, and for adding the standard deviations and batch sensitivity results. These additions address most of my questions about the experimental setup. The unmasked-prior comparison helps separate the effects of the prior and the mask on forecasting accuracy. My main concern remains the claim of stable fusion. The new ablation uses the original inputs, so it remains unclear whether the mask helps under input perturbations or workload shifts. Given this remaining concern, I am keeping my original rating.
Thank you for raising this concern. We evaluated three attention variants on the Alibaba and Google2019 memory workload datasets at all four forecast horizons. The variants were standard attention, attention with an unmasked relational prior, and Full TiFuse with the input-dependent mask. We added zero-mean Gaussian noise only to the historical observations in each test window. For each window, we applied the same noise realization to all three variants and kept the test windows, inference batch grouping, and forecast targets fixed.
Under Gaussian perturbation, Full TiFuse achieved lower MSE and MAE than the unmasked-prior variant in every evaluated setting. We measured error increases relative to each variant's performance on the original inputs. For both MSE and MAE, Full TiFuse showed a smaller increase in most settings. These results suggest that the input-dependent mask mitigates the degradation in forecasting accuracy under the tested Gaussian perturbation in most settings, with the effect varying across datasets and forecast horizons.
Memory workload forecasting errors after adding Gaussian noise to the historical inputs
| Dataset | Horizon | Standard Attention MSE | Standard Attention MAE | Unmasked Prior MSE | Unmasked Prior MAE | Full TiFuse MSE | Full TiFuse MAE |
|---|---|---|---|---|---|---|---|
| Alibaba | 1 | 0.11487 | 0.20571 | 0.11320 | 0.20275 | 0.11053 | 0.19726 |
| Alibaba | 5 | 0.37054 | 0.39790 | 0.36087 | 0.39324 | 0.35098 | 0.38073 |
| Alibaba | 10 | 0.55761 | 0.54943 | 0.54161 | 0.53339 | 0.53766 | 0.52758 |
| Alibaba | 30 | 0.76112 | 0.68216 | 0.74567 | 0.66667 | 0.72203 | 0.65834 |
| Google2019 | 1 | 0.00908 | 0.07410 | 0.00899 | 0.07317 | 0.00881 | 0.07090 |
| Google2019 | 5 | 0.05203 | 0.17834 | 0.05155 | 0.17426 | 0.04983 | 0.17208 |
| Google2019 | 10 | 0.10245 | 0.25412 | 0.10108 | 0.24690 | 0.09762 | 0.23611 |
| Google2019 | 30 | 0.28952 | 0.46367 | 0.28277 | 0.45240 | 0.27361 | 0.44073 |
1
Thank you for pointing out this discrepancy. The manuscript describes frequency selection for an individual input. The code selects frequencies using energy aggregated across the batch. To simplify the notation and make the computation easier to follow, we first describe DAFC using a single-channel sequence. PatchTST also uses a univariate sequence to illustrate the computation of a multivariate model. Section 3.2 states, “To illustrate the computation of DAFC, we first consider a single-channel historical sequence.” In this setting, the energy definition and frequency selection in Eqs. (4)–(6) are mathematically valid.
In the public implementation, we also aggregate energy across the batch and variable dimensions, extending Eq. (4) as follows.
Eˉq=BCp1∑b=1B∑c=1C∑u=1p∣Zb,c,u,q∣2
We use the aggregated energy to select the frequency indices shared across the entire batch. Here, B, C, and p denote the batch size, the number of variables, and the patch length, respectively. When B=C=1, this definition reduces to Eq. (4). The original manuscript did not fully explain the batch aggregation or the use of shared indices, leaving the distinction between per-sample and batch-level selection unclear. We have clarified this distinction in Appendix A.1.
During model development, we used the experimental option
keep_dc=Trueto compare strategies for handling the DC component. This option was specific to those development experiments. We have since refined and committed the code.To assess the sensitivity of TiFuse to inference batch size and composition, we ran additional memory workload forecasting experiments at a forecast horizon of H=5. Table 11 uses a batch size of 512 as the reference and compares it with batch sizes of 1, 64, and 128. Table 12 compares the original batch grouping with two alternative groupings generated using random seeds 2026 and 2025.
Table 11 (Appendix D). Memory workload forecasting errors at different inference batch sizes (H=5)
Table 12 (Appendix D). Memory workload forecasting errors under different inference batch groupings (H=5)
Under these test conditions, test-set MSE and MAE remained broadly stable across inference batch sizes and groupings.