NVIDIA Launches 64GB DGX Spark at $4,999 as Memory Costs Drive 128GB Model to $6,950

Key points

• NVIDIA DGX Spark 64GB launches October 23, 2026, at $4,999.
• The 128GB model price has risen to over $6,000, with some reports citing $6,950.
• The 64GB model supports up to 100-billion-parameter models; clustering two units supports 200B+ parameters.
• Two clustered 64GB units deliver up to 1.7x performance and double the memory bandwidth of a single unit.
• The new SKU is available from partners including Acer, ASUS, Dell, Gigabyte, HP, and MSI.
• The price increase is attributed to rising memory and component costs, often referred to as a 'RAMpocalypse'.

Background

The NVIDIA DGX Spark was initially launched at $3,999 but faced multiple price increases, reaching $4,699 and then $4,999 for the 128GB model. Wccftech previously tested the device, noting its capabilities for AI workloads. The current market is experiencing significant memory cost inflation, impacting pricing across AI hardware. AMD’s Ryzen AI Halo and MAX systems are currently priced between $3,500 and $4,500, offering competitive value with higher memory configurations compared to the new NVIDIA pricing.

How outlets are covering it

Wccftech highlights the strategic move to maintain a $4,999 price point by reducing memory, noting that the 128GB model is now significantly more expensive. They emphasize the competitive disadvantage against AMD’s Ryzen AI systems, which offer more memory for less money. NVIDIA’s official blog frames the 64GB launch as an accessible entry point for local AI development, emphasizing the ability to scale via clustering and the inclusion of the full NVIDIA AI software stack. The Register focuses on the 'memory crunch,' noting that the 128GB version has jumped nearly 75% above its launch price to $6,950. Tom’s Hardware describes the move as a 'lifeline' for local AI fans amid the 'RAMpocalypse,' acknowledging that while the 64GB model is cheaper, the 128GB option remains prohibitively expensive for many users. All sources agree that the price hikes are driven by rising memory costs, but they differ in emphasis: NVIDIA focuses on scalability and software, while critics focus on the value proposition compared to AMD.

Why it matters

The rising cost of memory is reshaping the market for local AI hardware. NVIDIA’s decision to release a lower-memory variant at the same price as the previous high-memory model indicates the severity of component cost inflation. This affects developers and researchers who rely on local AI workstations, potentially pushing them toward alternative platforms like AMD’s Ryzen AI systems or cloud-based solutions. The ability to cluster units to restore memory capacity is a key differentiator for NVIDIA, but the initial price point remains a barrier for many users. This trend may influence future hardware design and pricing strategies across the AI industry.

What to watch

The 64GB DGX Spark will be available from partner manufacturers starting October 23, 2026. NVIDIA is also launching the Sync Model Launcher at the end of the month, which will simplify running local AI models like Qwen3.8 27B on single or clustered systems. AMD’s Ryzen AI systems are expected to face similar price pressures due to memory costs, though current pricing remains lower. The market will likely see continued volatility in AI hardware pricing as memory costs remain elevated. Developers may need to evaluate the trade-offs between NVIDIA’s clustering capabilities and AMD’s higher memory configurations at lower price points.