Superpower Daily
OpenAI’s Codex Shows a Gated Image Upgrade, With Access Still U...Crusoe Signs Reported $13 Billion, Five-Year AI Cloud Deal With...Microsoft Releases MAI-Transcribe-2 at 10 Cents an HourxAI Opens Grok Bot to Enterprises With Controls for Autonomous...Nvidia Sets October Availability for RTX Spark PCs, Still Gives...HUMAIN Releases Arabic AI Preview Built on MiniMax’s M3 LineageOpenAI’s Codex Shows a Gated Image Upgrade, With Access Still U...Crusoe Signs Reported $13 Billion, Five-Year AI Cloud Deal With...Microsoft Releases MAI-Transcribe-2 at 10 Cents an HourxAI Opens Grok Bot to Enterprises With Controls for Autonomous...Nvidia Sets October Availability for RTX Spark PCs, Still Gives...HUMAIN Releases Arabic AI Preview Built on MiniMax’s M3 Lineage
Infrastructurepublished

Nvidia Sets October Availability for RTX Spark PCs, Still Gives No Pricing

New Lenovo systems give Nvidia’s local-AI PC platform more concrete hardware, but a conflict over one model’s memory ceiling and company-supplied performance figures leave key buyer questions open.

By 2 min read
Nvidia Sets October Availability for RTX Spark PCs, Still Gives No Pricing
Nvidia Sets October Availability for RTX Spark PCs, Still Gives No Pricing

Listen to this story

The audio brief

About 1:43
0:001:43
Read transcript
Nvidia now expects its RTX Spark systems to reach buyers in October, turning a June announcement into an actual purchasing decision. The catch is that Nvidia still has not said what these PCs will cost. RTX Spark is built around the N1X, an Arm-based system-on-a-chip that combines a Grace CPU and a Blackwell RTX GPU. Because they share unified memory, the processor and graphics hardware can draw from the same pool—an important design for running AI models locally. There are two versions. The higher-end N1X has 20 CPU cores, 6,144 GPU cores, and between 24 and 128 gigabytes of memory. It is planned for laptops and desktops. The lower tier has 18 CPU cores, 5,120 GPU cores, and 24 to 32 gigabytes, and is initially limited to laptops. Lenovo makes the pitch more concrete with two Yoga systems. Its 15.3-inch Yoga Pro 9n targets local AI, content creation, and software development, with up to 128 gigabytes of memory and four terabytes of storage. But the Yoga 9n 2-in-1 has a basic specification conflict: CNET lists a 32-gigabyte maximum, while Lenovo lists 64. Nvidia also claims one petaflop using the low-precision NVFP4 format, local operation of models with up to 120 billion parameters, and 1440p gaming at 100 frames per second with DLSS 5. Those are company figures, not independent test results. With systems expected from Asus, Dell, MSI, HP, Lenovo, and Microsoft, October’s key unanswered question is whether the high-memory capability has a price buyers will accept.

Story brief

3 key points

Nvidia’s RTX Spark PC platform is moving from its June Computex announcement toward October availability, but the commercial proposition is still incomplete: system pricing has not been disclosed. Lenovo’s Yoga Pro 9n demonstrates the high-memory pitch with up to 128GB of unified memory and 4TB of storage, while a second convertible has conflicting published memory limits of 32GB and 64GB. The N1X’s two tiers target...

  1. 01

    The higher N1X pairs a 20-core Grace CPU with a 6,144-core Blackwell GPU and supports 24GB to 128GB of unified memory.

  2. 02

    The lower tier has 18 CPU cores, 5,120 GPU cores, 24GB to 32GB of memory, and is initially laptop-only.

  3. 03

    Lenovo’s 15.3-inch Yoga Pro 9n offers up to 128GB of unified memory and 4TB of storage for local AI and creator workloads.

Nvidia expects RTX Spark systems to become available in October, putting its AI-focused PC platform on a nearer commercial timetable. Newly detailed Lenovo laptops show the platform reaching portable machines with up to 128GB of unified memory, while Nvidia has yet to disclose system prices and its headline capability figures remain company claims.

The timing follows Nvidia’s June 1 introduction of RTX Spark N1X at Computex. N1X is an Arm-based system-on-chip that combines a Grace CPU and a Blackwell RTX GPU on one piece of silicon. Its unified-memory design lets the CPU and GPU use the same memory pool.

The platform splits into two memory tiers

The higher N1X configuration combines a 20-core Grace CPU with a 6,144-core Blackwell GPU and supports 24GB to 128GB of unified memory. It is intended for laptops and desktops. The lower tier has an 18-core Grace CPU, a 5,120-core Blackwell GPU, and 24GB to 32GB of unified memory; it is initially limited to laptops.

Lenovo adds two concrete laptop designs

Lenovo has announced two RTX Spark Yoga systems. The 15.3-inch Yoga Pro 9n is a creator laptop aimed at local AI development, content creation and software development, with up to 128GB of memory and 4TB of storage. The Yoga 9n 2-in-1 is a 16-inch convertible with an OLED display, RTX Spark, and pen input.

The Yoga 9n 2-in-1’s maximum memory is not yet settled across published specifications. CNET lists up to 32GB and 2TB of storage, while Lenovo specifies up to 64GB of unified memory for the same model.

The performance pitch needs its labels

Nvidia cites one petaflop of GPU performance for RTX Spark using NVFP4, a low-precision data type. Low precision can improve speed and reduce memory overhead, but it has a greater risk of rounding errors than FP16 or FP32. That makes the figure unlike performance results expressed in those higher-precision formats.

Nvidia also says N1X systems can run models with up to 120 billion parameters locally and has promoted 1440p gaming at 100 frames per second with DLSS 5 and up to six-times multi-frame generation, particularly on the higher-end chip. Those are Nvidia’s stated capability and performance figures.

October turns a chip launch into a buying decision

RTX Spark laptops are expected from Asus, Dell, MSI, HP, Lenovo and Microsoft, alongside Acer’s undated desktop plan. Nvidia has not disclosed pricing for RTX Spark systems, leaving the cost of its high-memory local-AI pitch unresolved as partner configurations approach availability.

Sources

  1. cnet.comNvidia RTX Spark Systems Approach the Starting Line - CNET
  2. digitaltrends.comNvidia has two new N1X chips coming with up to 128GB of unified memory and serious AI compute power
  3. news.lenovo.comLenovo Expands the Yoga Portfolio with New AI PCs and Tablets Built for Creators - Lenovo StoryHub

Related coverage

Continue reading