AI News Feed
Market watch
AI Chips & Compute

TechRadar compares Nvidia RTX Spark and Apple M6 on price, specs and early performance

TechRadar compares Nvidia RTX Spark and Apple M6 on price, specs and early benchmarks for creators, gamers and AI.

TechRadar reports that neither chip is sold on its own, so system configuration drives the cost. The M6 Mac mini starts at $899/£899/AU$1,449 for a 16GB/256GB model. Upgrading to 24GB of unified memory adds $200/£200/AU$300, while 32GB adds $380/£400/AU$600. A 512GB SSD adds $200/£200/AU$300; 1TB adds $500/£500/AU$750; and 2TB adds $1,000/£1,000/AU$1,500. Nvidia's RTX Spark prices will vary by manufacturer and by N1X configuration, with an entry version using 18 CPU cores and 5,120 CUDA cores and a premium version using 20 CPU cores and 6,144 CUDA cores. Dell, Acer and other partners have announced units, but there is no firm pricing yet. Early indicators put starting prices for RTX Spark laptops and mini PCs at around $2,000/£2,000/AU$3,000. TechRadar gives Apple the price advantage.

In specifications, TechRadar's side-by-side comparison shows Nvidia using TSMC 3nm process technology and Apple using TSMC 2nm. The Nvidia N1X entry and premium chips have 18 and 20 CPU cores, respectively, against 12 CPU cores in the M6. Their maximum CPU clock speeds are 3.90GHz and 4.00GHz, compared with 4.79GHz for the M6. The Nvidia chips have 40 and 48 GPU cores, while the M6 has 12; they also have 160 and 192 AI cores, against 32 in the M6. Maximum memory is 128GB for Nvidia and 32GB for Apple, all using unified LPDDR5X. Nvidia lists 273.2GB/s memory bandwidth and Apple lists 170GB/s. AI performance is quoted at 1,000 TOPS for Nvidia on FP4 and 132 TOPS for Apple on FP16. TechRadar says Nvidia has nearly twice the CPU cores on its premium configuration, almost four times the maximum memory headroom, larger bandwidth and an RTX Blackwell GPU, while Apple's advantage is a maximum clock speed nearly 800MHz higher. It gives Nvidia the specifications advantage.

TechRadar says it has only been able to benchmark the Apple M6 itself, because the new Mac mini went on sale on September 22, so there is no definitive RTX Spark versus M6 performance result. The publication states that the RTX Spark N1X is essentially the Nvidia DGX Spark SoC, though it runs at lower wattage for laptops, roughly 45W to 80W TDP, compared with the desktop DGX Spark's 140W TDP. For RTX Spark desktops and mini PCs, TechRadar expects the N1X TDP to come close to or match 140W. It also cites leaked Geekbench 7 scores for the two RTX Spark N1X SKUs and the M6. In single-core, the Nvidia chips score 2,541 and 2,570, while the M6 scores 4,018. In multi-core, the Nvidia chips score 21,776 and 23,126, while the M6 scores 22,197. In 3DMark Steel Nomad, the Nvidia chips are estimated at about 2,750 and 3,750, against 1,475 for the M6. In 3DMark Solar Bay, they are estimated at about 53,200 and 76,000, against 31,513 for the M6. TechRadar notes that the RTX Spark N1X 3DMark scores are estimated from RTX 5070 and RTX 5070 Ti laptop GPU benchmark results.

TechRadar's comparison does not declare a final winner across all workloads. It states that the M6 has a price advantage and the RTX Spark has a specifications advantage, while performance remains partly based on leaked and estimated figures.

Editor's Summary

According to TechRadar, Nvidia's RTX Spark offers more CPU, GPU and AI cores, higher memory capacity and bandwidth than Apple's M6, while Apple's M6 leads on clock speed and initial Mac mini pricing. The publication cautions that final performance comparisons remain limited because only the M6 has been independently benchmarked and RTX Spark devices are not yet on sale.