Nvidia DGX Spark Gets $4,999 Cut-Down Model Amid Memory Crunch

What happened

Nvidia has introduced a cheaper variant of its DGX Spark desktop AI computer, priced at $4,999 but fitted with half the memory and storage of the original configuration. The move comes as the company simultaneously raises the price of its existing 128GB DGX Spark to $6,950, a jump of nearly 75% from its original launch price.

According to The Register, the repricing and the launch of the lower-specification model are tied to a broader memory supply crunch affecting the industry. Rather than holding the line on specifications at a fixed price, Nvidia appears to be using a tiered approach: a lower-cost entry point with reduced RAM and storage for buyers who don't need the full configuration, alongside a markedly more expensive top-tier model that reflects the rising cost of memory components.

DGX Spark is positioned as a compact desktop system for developers and researchers to run AI models locally, making it a notable data point for how memory scarcity is now filtering through into AI hardware pricing and product strategy rather than remaining confined to component markets.

Why it matters

This is fundamentally a hardware economics story with direct implications for AI adoption. When the cost of running or building AI infrastructure rises — whether through cloud GPU pricing or, as here, desktop AI workstations — it changes the calculus for which organisations and individual developers can afford to experiment with local AI workloads. A nearly 75% price increase on a flagship configuration, paired with a cut-down budget option, signals that memory constraints are becoming a real constraint on the pace and shape of AI hardware rollout.

For technology and digital transformation leaders, the signal is less about one product and more about planning assumptions. Budgets and procurement timelines built around current AI hardware price points may need revisiting if memory supply remains tight, and organisations evaluating on-premises or edge AI infrastructure should treat sticker prices as more volatile than in a typical hardware cycle.

By the numbers

• $4,999 — price of the new, lower-specification DGX Spark with half the RAM and storage of the original model
• $6,950 — new price of the existing 128GB DGX Spark configuration
• Nearly 75% — increase in the 128GB model's price versus its original launch price

The Renascence take

It's tempting to read this purely as a supply-chain story, but it's also a pricing-and-choice-architecture story. Nvidia hasn't simply raised a price; it has restructured the decision a buyer faces, adding a lower-anchored option that makes the steep increase on the premium tier feel more palatable by comparison.

Most coverage will frame this as a memory shortage squeezing margins, and that's true — but the more interesting move is the choice design. By launching a cheaper, scaled-down model at the same moment the flagship price jumps nearly 75%, Nvidia gives buyers a relative anchor that makes the premium price look like a considered option rather than a shock. Any operator facing input-cost pressure should take note: how you restructure the choice set often matters more to customer perception than the raw number you're asking people to pay. The real test here is whether developers accustomed to the original spec feel they are being asked to pay more for the same thing, or choose less for less — that framing will shape loyalty as much as the price itself.