Nvidia's (NVDA) new DGX Spark has half the memory and costs $1,000 more than the original did at launch

Illustration of a desktop monitor showing the Nvidia logo above a backlit keyboard

Key points

  • A 64GB DGX Spark arrives Oct. 23
  • The 128GB model just got a lot pricier
  • The launch comes amid rising memory costs

Nvidia (NVDA) will start selling a DGX Spark with 64GB of memory on Oct. 23, starting at $4,999 through Acer, Asus, Dell, Gigabyte, HP and MSI. That's half the memory of the original DGX Spark, which launched a year ago at $3,999.

The original 128GB model now costs more, too. Nvidia raised its price on Friday to $6,950, nearly 75% above its launch price, according to The Register, which blamed soaring memory prices.

What does the 64GB DGX Spark do?

The DGX Spark is a small desktop computer for running and building AI models locally, without renting cloud servers. The 64GB version keeps the same GB10 Grace Blackwell Superchip and software as the 128GB model, and Nvidia says it can run models with up to 100 billion parameters on the device.

Two units can be connected with a cable to pool their memory to 128GB. Nvidia says a pair can run models with up to 200 billion parameters. In its own test with Qwen 3.8 27B, two clustered 64GB systems delivered up to 1.7 times the performance of one.

Nvidia's Sync app sets up the two-unit cluster automatically. "The cluster assistant feature detects connected units, validates device configuration and configures the ConnectX-7 network," Nvidia said.

Why is the DGX Spark getting more expensive?

Memory chips cost a lot more than they did a year ago. At $4,999, the 64GB model is still 25% more than the 128GB version cost at launch. Nvidia's online store listed the DGX Spark as out of stock on Friday, and Tom's Hardware put the going price for a 128GB GB10 system at roughly $7,000 to $9,000, when you can find one.

It isn't just the Spark. Nvidia also raised the price of its Shield TV Pro streaming box to $299 from $199.99 starting Friday and stopped making the standard Shield TV. "The cost of components, including memory, has increased substantially across the industry," Nvidia said, according to Wccftech.

Meanwhile, memory makers' sales and profits are climbing sharply. Micron (MU) reported fiscal fourth-quarter revenue of $54.2 billion on Wednesday, up from $11.3 billion a year earlier. Samsung Electronics and SK Hynix are expected to report record quarterly profits this month.

How is Nvidia stock doing?

Nvidia shares hit a record intraday high of $237.87 on Friday morning, passing the $236.54 peak from May 14. They were up 3.0% at $237.72 as of 10:02 a.m. ET, above the stock's record close of $235.74, also set on May 14.

Frequently asked questions

How much does the 64GB Nvidia DGX Spark cost?

The 64GB DGX Spark starts at $4,999 and goes on sale October 23, 2026, through Acer, Asus, Dell, Gigabyte, HP and MSI, according to Nvidia (NVDA).

How much does the 128GB DGX Spark cost now?

Nvidia raised the price of the 128GB DGX Spark to $6,950 on October 2, 2026, nearly 75% above its $3,999 launch price, according to The Register, which blamed soaring memory prices.

What is the difference between the 64GB and 128GB DGX Spark?

Both use the GB10 Grace Blackwell Superchip and the same software. The 64GB model has half the memory and, according to Nvidia, runs models with up to 100 billion parameters on the device.

Can two DGX Spark 64GB units work together?

Yes. Two units connected by cable pool their memory to 128GB and can run models with up to 200 billion parameters, Nvidia says. In Nvidia's Qwen 3.8 27B test, two clustered systems delivered up to 1.7 times the performance of one, and the NVIDIA Sync app configures the cluster automatically.

Why are Nvidia products getting more expensive?

Memory costs have risen sharply. The Register attributed the DGX Spark price increase to soaring memory prices, and Nvidia raised its Shield TV Pro to $299 from $199.99 on October 2, 2026, saying the cost of components, including memory, has increased substantially across the industry.

More on NVDA and MU

Dennis Singleton
Dennis Singleton

Dennis Singleton was born in Australia and later moved to the United States. He has spent years following the markets, but what keeps his attention is how AI is built. He writes about the companies behind the technology, from semiconductor designers and advanced packaging to photonics, memory, networking, and the hardware powering modern AI. His approach starts with filings, earnings, and industry research, then translates the important details into clear, straightforward analysis without unnecessary hype.