RTX Spark N1X Beats AMD and Intel, Trails Apple
💡A leaked Geekbench result puts NVIDIA’s RTX Spark N1X ahead of AMD and Intel but behind Apple’s M5 Max.
⚡ 30-Second TL;DR
What Changed
Geekbench 7 lists benchmark results for two RTX Spark N1X configurations.
Why It Matters
The results suggest that NVIDIA’s upcoming ARM-based AI-oriented hardware could be competitive with AMD and Intel in multicore performance, while Apple remains a difficult benchmark target. AI developers should treat the figures as preliminary until retail hardware, drivers, and workloads are independently tested.
What To Do Next
When RTX Spark N1X hardware becomes available, benchmark your inference stack with Geekbench 7 plus representative ONNX or PyTorch workloads before selecting it for deployment.
Key Points
- •Geekbench 7 lists benchmark results for two RTX Spark N1X configurations.
- •The 20-core full-performance model achieved a multicore score of 23,126.
- •Its result was approximately 12% higher than AMD Ryzen AI Max+ 395.
- •It remained approximately 35% behind Apple M5 Max in the reported comparison.
🧠 Deep Insight
AI-generated analysis for this event.
🔑 Enhanced Key Takeaways
- •The RTX Spark N1X utilizes a 20-core CPU configuration split into two clusters of 10 cores each, operating at a base frequency of 4.00 GHz.
- •A secondary, lower-end 18-core variant of the N1X has been identified, featuring a 5,120-core GPU compared to the 6,144-core GPU found in the full-performance model.
- •The N1X chip is built on a 3nm process using 2.5D packaging, integrating a MediaTek-designed ARM v9.2 CPU complex with NVIDIA's Blackwell-class GPU architecture.
- •The platform supports up to 128GB of LPDDR5X-9400 unified memory, connected via NVLink C2C with 600 GB/s of bidirectional bandwidth.
- •Benchmark results were recorded on Windows 11 on Arm, with the 20-core model outperforming several x86 mobile chips, including AMD's Ryzen AI Max+ 395 and Intel's Core Ultra X9 388H, in multi-core tests.
📊 Competitor Analysis▸ Show
| Feature | NVIDIA RTX Spark N1X | Apple M5 Max | AMD Ryzen AI Max+ 395 |
|---|---|---|---|
| Architecture | ARM (MediaTek CPU + Blackwell GPU) | ARM (Fusion Architecture) | x86 (Zen 5 + RDNA 3.5) |
| Process | 3nm (2.5D Packaging) | 3nm (SoIC-mH) | 4nm |
| Max Memory | 128GB LPDDR5X-9400 | 128GB Unified | 128GB LPDDR5X |
| Geekbench 7 Multi-Core | ~23,126 | ~29,233 | ~20,609 |
🛠️ Technical Deep Dive
- CPU Architecture: 20-core (10+10) or 18-core (10+8) ARM v9.2 complex designed by MediaTek.
- GPU Architecture: Blackwell-class GPU with 6,144 or 5,120 CUDA cores.
- Interconnect: NVLink C2C providing 600 GB/s bidirectional bandwidth.
- Memory: Unified memory architecture supporting up to 128GB LPDDR5X-9400.
- Manufacturing: TSMC 3nm process with 2.5D packaging.
🔮 Future ImplicationsAI analysis grounded in cited sources
⏳ Timeline
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: cnBeta (Full RSS) ↗