On the (dis)Advantages of Programmable NICs for Network Security Services
Bibliographic record
Abstract
Emerging programmable network interface cards (a.k.a. SmartNICs) are a viable alternative to reduce the gap between network bandwidths, currently at the scale of multi-hundred Gbps, and the server CPU processing capacity. This has rapidly led to many efforts exploring SmartNICs for offloading or accelerating applications that traditionally run solely on servers (e.g., key-value stores, data analytics). Despite the success of this paradigm, the suitability of SmartNICs for running security applications, specially those that heavily rely on cryptographic operations, still remains largely unstudied. In this paper, we aim at filling this gap and provide the first in-depth analysis of current SmartNICs' crypto capabilities. Our experiments with an ARM-based multi-core SmartNIC show that the device depends heavily on architecture enhancements (e.g., cryptographic instructions and hardware accelerators) to meet server performance on crypto-workloads. Moreover, data movements between the SmartNIC and crypto-hardware accelerator cores can introduce significant overhead and make the latter ineffective, particularly for short living tasks. From a service perspective, SmartNICs can take advantage of their privileged position (i.e., closer to client devices than server CPUs) to speed up crypto-based functions. However, the SmartNIC benefits can be easily outweighed if the application is too much data-intensive or includes several noncrypto tasks.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.008 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".