NVIDIA’s DGX Spark is already selling above its official listed price in the US, just weeks before the first RTX Spark systems are expected to arrive.
NVIDIA currently lists the DGX Spark Founders Edition at $4,699, but the system is out of stock through its own marketplace. Available units are more expensive, with one current listing priced at $4,999.99.
That puts the cheapest available system about $300 above NVIDIA’s listed price and roughly 25 percent above the original $3,999 launch figure.
DGX Spark pricing has moved sharply higher
The DGX Spark was initially introduced at $3,999, but the effective market price has climbed since then.
| System | Recent price |
|---|---|
| DGX Spark original announced price | $3,999 |
| NVIDIA current listed price | $4,699 |
| Lowest available unit found | $4,999.99 |
| ASUS Ascent GX10 1TB peak listing | $8,299 |
| ASUS Ascent GX10 2TB peak listing | $9,499 |
| ASUS Ascent GX10 recent price | Around $6,000 |
The increase is not limited to NVIDIA’s own system.
ASUS Ascent GX10 listings also rose sharply, with the 1TB model temporarily reaching $8,299 and the 2TB version reaching $9,499. Those prices have since fallen, with the system now selling for around $6,000.
The temporary increases show how quickly pricing can move when supply is limited and demand remains high.
RTX Spark systems arrive in October
The pricing situation comes shortly before NVIDIA’s RTX Spark platform is due to reach the market.
Compact desktops and laptops based on RTX Spark are expected to begin shipping in October with Windows 11.
The desktop RTX Spark N1X platform combines a 20 core Grace CPU with a Blackwell RTX GPU containing 6,144 CUDA cores.
Higher configurations support up to 128GB of unified LPDDR5X memory, while lower versions can support up to 64GB.
That large unified memory pool is one of the main reasons these systems are being positioned for local AI workloads rather than ordinary desktop use.
Large unified memory could keep prices high
DGX Spark and RTX Spark are both built around the idea of running demanding AI models locally.
A configuration with 128GB of unified memory can handle workloads that would be difficult to run on a conventional consumer PC with separate system RAM and GPU VRAM.
However, that memory design also raises costs.

The current DGX Spark pricing suggests that RTX Spark systems may not be inexpensive when they reach stores, especially if demand is strong during the early launch period.
NVIDIA has not announced final retail prices for RTX Spark systems yet.
That leaves one of the biggest questions unanswered as October approaches.
Windows and Linux support remain separated for now
Another major difference between DGX Spark and RTX Spark is the operating system.
DGX Spark currently uses DGX OS, an Ubuntu based Linux distribution.
RTX Spark systems are officially positioned around Windows 11.
NVIDIA has not announced whether DGX Spark will officially support Windows or whether RTX Spark will receive official Linux support.
That could become an important distinction for developers.
Linux is widely used in AI development environments, while Windows may be more convenient for broader desktop software and commercial applications.
If NVIDIA keeps the two platforms separated by operating system, software support could become just as important as hardware specifications when choosing between them.
AMD and Apple are competing in the same market
NVIDIA is not alone in targeting compact AI workstations with large memory pools.
AMD recently introduced its Ryzen AI Halo Developer Platform using the Ryzen AI Max+ 395.
That system includes 128GB of unified LPDDR5X memory and officially supports both Windows 11 and Linux.
Apple also competes in this segment with the Mac Studio.
The M4 Max version can be configured with up to 128GB of memory, while higher end M3 Ultra configurations support substantially more.
Intel has processors that support large memory capacities, but it currently lacks a direct equivalent built around a high performance integrated GPU and a 128GB unified memory design.
RTX Spark launch pricing will be closely watched
The DGX Spark market gives an early indication of what buyers may face when RTX Spark systems arrive.
NVIDIA’s own listed price has already increased from the original announcement, while available stock is selling above that figure.
Partner systems have shown even larger temporary price increases.
That does not guarantee that RTX Spark PCs will follow the same pattern, but it shows that compact AI workstations with Grace Blackwell hardware and large unified memory pools are already commanding premium prices.
With RTX Spark systems due in October, final retail pricing may determine how broadly the platform can move beyond developers and specialist AI buyers.


Discussion (0)
Be the first to comment.