30초 요약
In this September 2024 promotional video, Matthew Berman praises an ASUS ProArt laptop that NVIDIA sent him and mentions its RTX 4070 Laptop GPU, NPU and AI features. He reports about 30 tokens per second on an unspecified Llama configuration. The video description uses the NVIDIA partner tag, and the clip does not provide enough test details to independently compare performance.
- 1A gifted laptopBerman calls the ASUS ProArt machine the most powerful PC laptop he has used and says NVIDIA sent it to him.
- 2GPU and NPUThe clip identifies an RTX 4070 Laptop GPU and a neural processing unit; a Copilot key opens an assistant and does not itself establish where inference runs.
- 3A local model testBerman reports a Llama model running at around 30 tokens per second, with model size and settings unspecified.
- 4RTX softwareHe highlights DLSS, RTX Video Super Resolution, NVIDIA ACE and NVIDIA Broadcast.
핵심 정리
- Berman says NVIDIA sent him the laptop, and the video description identifies an NVIDIA partnership.
- The showcased ProArt configuration has an RTX 4070 Laptop GPU; ASUS offered different specifications across the ProArt line.
- Berman reports around 30 tokens per second for a Llama model, without identifying the model size or test settings.
- NVIDIA’s DLSS, RTX Video Super Resolution and Broadcast are among the features Berman highlights, subject to software and device support.
챕터별로 다듬은 전문 읽기
The laptop and local AI
Berman introduces a gifted ASUS ProArt laptop, its GPU and NPU, and a reported Llama speed. The video does not detail the test configuration.
NVIDIA sent me this ASUS ProArt laptop, which I describe as the most powerful PC laptop I have used. The video is promotional in context: its description uses the #NVIDIAPartner tag. It shows a GeForce RTX 4070 Laptop GPU; [ASUS’s 2024 ProArt announcement](https://press.asus.com/news/press-releases/asus-computex-2024-ai-copilot-pcs-proart-zenbook-vivobook-tuf/) lists configurations up to that GPU, so specifications vary by model.
The machine also has a neural processing unit for supported AI workloads and a dedicated Copilot key. Pressing the key launches the assistant; it does not by itself prove that a particular Copilot request runs locally on the NPU. I say that I ran a Meta Llama model at around 30 tokens per second, but I do not specify its size, quantization or settings, so that observation is not a general benchmark.
RTX features
He points to gaming, video enhancement, digital-human and broadcasting software. Feature availability depends on supported applications and configuration.
I highlight NVIDIA’s DLSS for supported games and RTX Video Super Resolution for video enhancement, which I find impressive. I also mention NVIDIA ACE digital-human technology and NVIDIA Broadcast for video and audio effects. Availability and performance depend on the application, hardware, drivers and configuration; the clip does not independently test each feature.
NVIDIA Broadcast can apply AI-assisted effects to a compatible microphone or camera workflow. I thank NVIDIA for sending the laptop and point viewers to the company’s AI Decoded material for further information. My positive assessment reflects my experience with this particular machine rather than a controlled comparison of all laptops.
같은 채널의 다른 글
34:21EN
Matthew Berman· 2026년 10월 27일Matthew Berman Examines Anthropic’s IPO Filing, AI Growth and GovernanceMatthew Berman, Sam Altman · Forward FutureThe presenter reviews the S-1’s financial metrics, compute strategy and voting structure while questioning the tension between rapid AI development and safety.
0:26ENMatthew Berman· 2026년 10월 25일Sam Altman on Human Values and Problem-Solving in an AI WorldSam Altman · OpenAIIn a short interview excerpt, Altman considers the enjoyment of solving hard problems and the human role in interpreting results.
0:36ENMatthew Berman· 2026년 10월 25일Kimi K3.1 and a Rubik’s Cube Solver DemonstrationMatthew Berman · Forward FutureMatthew Berman shows a scrambled virtual cube being solved and gives the Kimi K3.1 test a favorable assessment.
0:21ENMatthew Berman· 2026년 10월 24일An AI-Generated Kimi Benchmark Chart Used Invented ScoresMatthew Berman · Forward FutureMatthew Berman asks for a graph of Kimi K3.1 Pro benchmarks, then discovers that the displayed figures are illustrative placeholders.
25:50ENMatthew Berman· 2026년 10월 24일Kimi K3.1 Pro Tested: Strong 3D Demos, Uneven Results and Competitive PricingMatthew Berman · Forward FutureMatthew Berman tests the open-weight model on coding, 3D rendering and visual explanations, then compares benchmark results and prices.
32:20ENMatthew Berman· 2026년 10월 23일Who Is Local AI For? The Costs, Trade-offs and a Hybrid FutureMatthew Berman · Forward FutureMatthew Berman weighs data control and independence against hardware cost, setup complexity and inference speed.