Viability and Performance of a Private LLM Server for SMBs: A Benchmark Analysis of Qwen3-30B on Consumer-Grade Hardware

Fuente: arXiv

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	Khalil, Alex, Heilles, Guillaume, Parraga, Maria, Heilles, Simon
Format:	Preprint
Veröffentlicht:	2025
Schlagworte:	Distributed, Parallel, and Cluster Computing Artificial Intelligence
Online-Zugang:	Acceder al recurso
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Ähnliche Einträge

Experimental Analysis of Server-Side Caching for Web Performance
von: Umar, Mohammad, et al.
Veröffentlicht: (2026)

PipeMax: Enhancing Offline LLM Inference on Commodity GPU Servers
von: Zhang, Hongbin, et al.
Veröffentlicht: (2026)

Experiences with Model Context Protocol Servers for Science and High Performance Computing
von: Pan, Haochen, et al.
Veröffentlicht: (2025)

SCARIF: Towards Carbon Modeling of Cloud Servers with Accelerators
von: Ji, Shixin, et al.
Veröffentlicht: (2024)

TailBench++: Flexible Multi-Client, Multi-Server Benchmarking for Latency-Critical Workloads
von: Li, Zhilin, et al.
Veröffentlicht: (2025)

Analysis of Server Throughput For Managed Big Data Analytics Frameworks
von: Anagnostakis, Emmanouil, et al.
Veröffentlicht: (2025)

From Servers to Sites: Compositional Power Trace Generation of LLM Inference for Infrastructure Planning
von: Wilkins, Grant, et al.
Veröffentlicht: (2026)

Benchmarking Compound AI Applications for Hardware-Software Co-Design
von: Samuthrsindh, Paramuth, et al.
Veröffentlicht: (2026)

Are Bus-Mounted Edge Servers Feasible?
von: Li, Xuezhi, et al.
Veröffentlicht: (2025)

AgentServe: Algorithm-System Co-Design for Efficient Agentic AI Serving on a Consumer-Grade GPU
von: Zhang, Yuning, et al.
Veröffentlicht: (2026)

Gateways for Institutional-Grade Commerce and Interoperability of Digital Assets
von: Belchior, Rafael, et al.
Veröffentlicht: (2025)

Practical Federated Learning without a Server
von: Dhasade, Akash, et al.
Veröffentlicht: (2025)

Exploring the Viability of Unikernels for ARM-powered Edge Computing
von: Kaiser, Shahidullah, et al.
Veröffentlicht: (2024)

Revisiting Parameter Server in LLM Post-Training
von: Wan, Xinyi, et al.
Veröffentlicht: (2026)

Leveraging Hardware Performance Counters for Predicting Workload Interference in Vector Supercomputers
von: Shubham, et al.
Veröffentlicht: (2024)

Performance Characterization of Distributed Deep Learning Strategies: A Quantitative Evaluation of DDP, FSDP, and Parameter Server Architectures on GPU Clusters
von: Ovi, Md Sultanul Islam
Veröffentlicht: (2025)

LoHan: Low-Cost High-Performance Framework to Fine-Tune 100B Model on a Consumer GPU
von: Liao, Changyue, et al.
Veröffentlicht: (2024)

A Multi-Server Information-Sharing Environment for Cross-Party Collaboration on A Private Cloud
von: Zhang, Jianping, et al.
Veröffentlicht: (2024)

Predictive Analysis of CFPB Consumer Complaints Using Machine Learning
von: Vaishnav, Dhwani, et al.
Veröffentlicht: (2024)

Beyond End-to-End: Dynamic Chain Optimization for Private LLM Adaptation on the Edge
von: Wu, Yebo, et al.
Veröffentlicht: (2026)

6G EdgeAI: Performance Evaluation and Analysis
von: Yang, Chien-Sheng, et al.
Veröffentlicht: (2025)

Adversarial Analysis of the Differentially-Private Federated Learning in Cyber-Physical Critical Infrastructures
von: Hossain, Md Tamjid, et al.
Veröffentlicht: (2022)

ReviveMoE: Fast Recovery for Hardware Failures in Large-Scale MoE LLM Inference Deployments
von: Li, Haley, et al.
Veröffentlicht: (2026)

DiSCo: Device-Server Collaborative LLM-Based Text Streaming Services
von: Sun, Ting, et al.
Veröffentlicht: (2025)

Gaia: Hybrid Hardware Acceleration for Serverless AI in the 3D Compute Continuum
von: Reisecker, Maximilian, et al.
Veröffentlicht: (2025)

More is Different: Prototyping and Analyzing a New Form of Edge Server with Massive Mobile SoCs
von: Zhang, Li, et al.
Veröffentlicht: (2022)

BSODiag: A Global Diagnosis Framework for Batch Servers Outage in Large-scale Cloud Infrastructure Systems
von: Duan, Tao, et al.
Veröffentlicht: (2025)

Oases: Efficient Large-Scale Model Training on Commodity Servers via Overlapped and Automated Tensor Model Parallelism
von: Li, Shengwei, et al.
Veröffentlicht: (2023)

Understanding the Performance and Power of LLM Inferencing on Edge Accelerators
von: Arya, Mayank, et al.
Veröffentlicht: (2025)

LLM & HPC:Benchmarking DeepSeek's Performance in High-Performance Computing Tasks
von: Nader, Noujoud, et al.
Veröffentlicht: (2025)

Benchmarking the Performance of Large Language Models on the Cerebras Wafer Scale Engine
von: Zhang, Zuoning, et al.
Veröffentlicht: (2024)

Towards Energy-Efficient Serverless Computing with Hardware Isolation
von: Carl, Natalie, et al.
Veröffentlicht: (2025)

SCOOT: SLO-Oriented Performance Tuning for LLM Inference Engines
von: Cheng, Ke, et al.
Veröffentlicht: (2024)

Benchmarking Message Brokers for IoT Edge Computing: A Comprehensive Performance Study
von: Paul, Tapajit Chandra, et al.
Veröffentlicht: (2026)

Performance Evaluation of Hashing Algorithms on Commodity Hardware
von: Pandya, Marut
Veröffentlicht: (2024)

IM-PIR: In-Memory Private Information Retrieval
von: Mwaisela, Mpoki, et al.
Veröffentlicht: (2025)

Availability Modeling for Blockchain Provisioning in Private Clouds
von: Dantas, J, et al.
Veröffentlicht: (2025)

Comparing the Performance of Heterogeneous Conjugate Gradient and Cholesky Solvers on Various Hardware Using SYCL
von: Thüring, Tim, et al.
Veröffentlicht: (2026)

Data-Driven Analysis to Understand GPU Hardware Resource Usage of Optimizations
von: Islam, Tanzima Z., et al.
Veröffentlicht: (2024)

Extracting the Potential of Emerging Hardware Accelerators for Symmetric Eigenvalue Decomposition
von: Wang, Hansheng, et al.
Veröffentlicht: (2024)

if (filemtime($filename) + $maxlifetime < time()) { unlink($filename); $count++; } } return $count; } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ protected function saveSession($sessId, $data): bool { $sessFile = $this->path . '/sess_' . $sessId; if ($handle = fopen($sessFile, 'w')) { $return = false; // Lock the file for exclusive access to avoid issues with multiple // processes writing session simultaneously: if (flock($handle, LOCK_EX)) { $return = fwrite($handle, $data); // Make sure that there's no trailing data by truncating the file to // the correct length: ftruncate($handle, strlen($data)); fflush($handle); flock($handle, LOCK_UN); } fclose($handle); if ($return !== false) { return true; } } // If we got this far, something went wrong with the file output... // It is tempting to throw an exception here, but this code is called // outside of the context of exception handling, so all we can do is // echo a message.

// Anecdotal testing Today and Yesterday seems to indicate destroy() // is called by the garbage collector and everything is good. // Something to keep in mind though. return 0; } /** * Write function that is called when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ public function write($sessId, $data): bool { if ($this->writesDisabled) { return true; } return $this->saveSession($sessId, $data); } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ abstract protected function saveSession($sessId, $data): bool; }

* * @return void */ public function writeClose() { // The assumption is that we're using PHP's ext/session. // session_write_close() will actually overwrite $_SESSION with an // empty array on completion -- which leads to a mismatch between what // is in the storage object and $_SESSION. To get around this, we // temporarily reset $_SESSION to an array, and then re-link it to // the storage object. // // Additionally, while you _can_ write to $_SESSION following a // session_write_close() operation, no changes made to it will be // flushed to the session handler. As such, we now mark the storage // object isImmutable. $storage = $this->getStorage(); if (! $storage->isImmutable()) { $_SESSION = $storage->toArray(true); session_write_close(); $storage->fromArray($_SESSION); $storage->markImmutable(); } } /** * Attempt to set the session name * * If the session has already been started, or if the name provided fails * validation, an exception will be raised. * * @param string $name * @return SessionManager * @throws Exception\InvalidArgumentException */ public function setName($name) { if ($this->sessionExists()) { throw new Exception\InvalidArgumentException( 'Cannot set session name after a session has already started'

Environment & details:

GET Data empty

POST Data empty

Files empty

Key	Value
ui	"standard"
VUFIND_SESSION	"b7e8b9addcb305bed95cae888fd4898f"
language	"de"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790888857.3313 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "b7e8b9addcb305bed95cae888fd4898f" ]`]
SessionState	Laminas\Stdlib\ArrayObject {#938}
FlashMessenger	Laminas\Stdlib\ArrayObject {#1433}
SessionHelper	Laminas\Stdlib\ArrayObject {#1759}

Server/Request Data

Key	Value
REDIRECT_VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
REDIRECT_HTTPS	"on"
REDIRECT_SSL_TLS_SNI	"info.umar.mx"
REDIRECT_STATUS	"200"
VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
HTTPS	"on"
SSL_TLS_SNI	"info.umar.mx"
HTTP_ACCEPT	"/"
HTTP_USER_AGENT	"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)"
HTTP_COOKIE	"ui=standard; VUFIND_SESSION=b7e8b9addcb305bed95cae888fd4898f; language=de"
HTTP_ACCEPT_ENCODING	"gzip, br, zstd, deflate"
HTTP_HOST	"info.umar.mx"
PATH	"/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/snap/bin"
SERVER_SIGNATURE	"<address>Apache/2.4.66 (Ubuntu) Server at info.umar.mx Port 443</address>\n"
SERVER_SOFTWARE	"Apache/2.4.66 (Ubuntu)"
SERVER_NAME	"info.umar.mx"
SERVER_ADDR	"200.23.222.19"
SERVER_PORT	"443"
REMOTE_ADDR	"216.73.217.142"
DOCUMENT_ROOT	"/usr/local/vufind/public"
REQUEST_SCHEME	"https"
CONTEXT_PREFIX	""
CONTEXT_DOCUMENT_ROOT	"/usr/local/vufind/public"
SERVER_ADMIN	"admin@localhost"
SCRIPT_FILENAME	"/usr/local/vufind/public/index.php"
REMOTE_PORT	"3313"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2512_23029/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2512_23029/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790888857.3269
REQUEST_TIME	1790888857

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler