Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 1909.07166

Exploring the Behavior of Coherent Accelerator Processor Interface (CAPI) on IBM Power8+ Architecture and FlashSystem 900

Kaushik Velusamy·Smriti Prathapan·Milton Halem
View original source ↗Read source document ↗Export metadata ↓

Abstract

The Coherent Accelerator Processor Interface (CAPI) is a general term for the infrastructure that provides high throughput and low latency path to the flash storage connected to the IBM POWER 8+ System. CAPI accelerator card is attached coherently as a peer to the Power8+ processor. This removes the overhead and complexity of the IO subsystem and allows the accelerator to operate as part of an application. In this paper, we present the results of experiments on IBM FlashSystem900 (FS900) with CAPI accelerator card using the "CAPI-Flash IBM Data Engine for NoSQL Software" Library. This library provides the application, a direct access to the underlying flash storage through user space APIs, to manage and access the data in flash. This offloads kernel IO driver functionality to dedicated CAPI FPGA accelerator hardware. We conducted experiments to analyze the performance of FS900 with CAPI accelerator card, using the Key Value Layer APIs, employing NASA's MODIS Land Surface Reflectance dataset as a large dataset use case. We performed Read and Write operations on datasets of size ranging from 1MB to 3TB by varying the number of threads. We then compared this performance with other heterogeneous storage and memory devices such as NVM, SSD and RAM, without using the CAPI Accelerator in synchronous and asynchronous file IO modes of operations. The results indicate that FS900 & CAPI, together with the metadata cache in RAM, delivers the highest IO/s and OP/s for read operations. This was higher than just using RAM, along with utilizing lesser CPU resources. Among FS900, SSD and NVM, FS900 had the highest write IO/s. Another important observation is that, when the size of the input dataset exceeds the capacity of RAM, and when the data access is non-uniform and sparse, FS900 with CAPI would be a cost-effective alternative.

Explore related subjects

cs.DC

Keep this discovery

BibTeXRIS

Kaushik Velusamy, Smriti Prathapan, Milton Halem. 2019-09-12. Exploring the Behavior of Coherent Accelerator Processor Interface (CAPI) on IBM Power8+ Architecture and FlashSystem 900. https://arxiv.org/abs/1909.07166

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Online Treasure Hunt in Vertex-Permuted Dynamic Rings

We study the problem of treasure hunt by a group of $k \geq 1$ agents in vertex-permuted dynamic rings (VP). In this model, the $n$ vertices remain on a ring but are permuted at each time step. We first show that treasure hunt is impossible for any $k \leq n-3$ agents, if there are no restrictions on the sequence of permutations used in the dynamic ring. We then study the $VP(\delta)$ setting, in which for every pair $i, j$ of vertices, the edge $(i, j)$ is guaranteed to appear within $\delta$ steps. We show that the class $VP(\delta)$ is feasible only for $\delta \geq \left\lceil \frac{n-1}{2}\right\rceil$. For the one-agent case, we show a tight bound of $\Theta(\delta n)$ on the worst-case search time as well as competitive ratio of any online algorithm for treasure hunt, provided $\delta \geq 2n$. We then give an optimal algorithm for $k$ agents, thereby showing that $k$ agents can obtain a speedup of $k$ on the worst-case search time. Finally, in the R-VP setting, in which in every step, the vertices are arranged as a ring according to a random permutation, we show that treasure hunt takes expected $\Theta(n)$ steps against an oblivious adversary and $\Theta(n \log n)$ steps against an adaptive adversary.

cs.DC↗
arXiv2026

The Computing Channel: How Modulation Programs the Airwaves

Distributed computing and distributed artificial intelligence require frequent exchanges of intermediate results, although many applications need only an aggregate rather than messages from individual devices. Conventional systems recover each message before computing the aggregate, whereas over-the-air computation (OAC) exploits simultaneous transmission to obtain it directly. However, dominant OAC implementations rely on analog signaling, creating a mismatch with finite-precision data and digital communication procedures. This article presents digital function-oriented communication, in which finite-alphabet symbol representations and receiver decisions are jointly designed so that multiple-access superposition encodes the desired function without recovering individual inputs. We introduce its computational-constellation principle, main design approaches, extensions, and implementation challenges. Federated edge learning illustrates how the framework can reduce user-dependent data-bearing resources while operating directly on quantized model updates.

cs.DC↗
arXiv2026

Can AI Remediate Backend Failures Safely? GuardedAct with Blast-Radius-Aware Sandboxing

Large Language Models (LLMs) have shown promising capabilities in generating remediation actions for microservice failures. However, directly executing AI-generated repair actions in production risks cascading collateral damage. We propose GuardedAct, a sandbox-first remediation framework that interposes a blast-radius-aware verification layer between the LLM action generator and the production environment. GuardedAct operates in four phases: (1) ingesting a diagnosis report together with the live system topology and recent telemetry, (2) prompting an LLM to produce a ranked list of candidate remediation actions, (3) simulating each action in a lightweight digital-twin sandbox that estimates the blast radius and assigns a risk label, and (4) enforcing a rollback-confidence gate that auto-executes only low-risk actions while escalating high-risk ones for human review. We evaluate GuardedAct on five fault scenarios injected into the DeathStarBench social-network application. Experimental results show that GuardedAct achieves an overall recovery rate of 87.4% while reducing collateral damage by 79.7% relative to direct LLM execution (from 25.6% to 5.2%), at the cost of a modest sandbox-induced increase in mean time to recovery (approximately 8 s). Ablation studies confirm that each component contributes meaningfully to the safety-speed trade-off.

cs.DC↗