DGX Spark at 64GB: What $4,999 of Local Compute Buys a Small Team

·3 views

TL;DR NVIDIA has put a 64GB configuration of DGX Spark on sale from Acer, ASUS, Dell, Gigabyte, HP and MSI from 23 October, starting at $4,999, with the same GB10 Grace Blackwell superchip, DGX OS and AI software stack as the 128GB model. One unit runs models up to 100 billion parameters entirely on device; two units connect over a QSFP cable into a 200 GbE fabric that pools 128GB and extends support to 200 billion parameters. The vendor's own measurement sets the expectation: in a Qwen 3.8 27B test, two clustered 64GB systems delivered up to 1.7x the performance of one, and the software envir

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Apache NiFi at 6,093 observed hosts: a data flow controller and the credentials it holds

The number A ZoomEye query for app="Apache NiFi" returned 6,093 matching hosts, collected on 2 October 2026 with sub_type=all and a page size of one record. The total is the match count for the fingerprint. A smaller figure than the monitoring or observability products, and one that still deserves attention because of what the software does rather than how many instances run it. NiFi moves data between systems. A flow is a graph of processors that read from a source, transform the payload and write to a destination.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Retrying Failed Jobs in Small Apps — Node.js Queues, DLQs, and 3 Practical Trade-offs

Short answer: for a property-management cleanup that must survive failures, a queue with a dead-letter queue (DLQ) and redrive is usually the better engineering choice than polling a failed_jobs table. The queue gives retries, acknowledgement, and poison-message isolation a visible shape. Database polling can still be the right call for a tiny, low-volume app where one process and one table are worth more than another service. I care about revenue per hour. A retry loop that takes a Friday away from shipping is an expensive feature, even if its cloud bill is small.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How TokenCap Folds Repetitive Imports Without Breaking Code Syntax

When feeding full repositories to AI coding agents, a significant portion of your token budget is wasted on boilerplate: long lists of external imports, standard license headers, and repetitive utility declarations. Simply stripping whitespace is dangerous: Python, YAML, and formatted strings break immediately. In TokenCap, we built src/pack/fold.js to perform format-aware context folding. Import Clustering: The tokenizer detects consecutive import statements and condenses them into a single-line structural fold: // [Folded 18 imports: React, Lucide-React, Lodash] Boilerplate Detection: Repeti

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Speech-to-Text API Timeouts: How to Bound Large Audio Uploads in Node.js

Short answer: treat a long recording as an upload job with explicit admission, transport, and inference stages. Reject an oversized file before opening a connection, put a short deadline around each network attempt, retry only transient responses, and show the user a clear fallback when transcription is unavailable. For supplier-invoice extraction, do not let a failed audio upload quietly become an empty invoice. There is an earlier gate, too. A route-shaped interface is not proof that a transcription capability is ready.

0 comments Read more at DEV Community