MoHelal

Tech Reviews, PC Enthusiast, Cloud/AI Engineer passionate about Tech, Mini-PCs, and Gaming

[MoHelal @M0hamed__82]

https://t.co/SfksrcyPcR

[MoHelal @M0hamed__82]

6/ Verdict: a great always-on homelab/office box with excellent I/O. Not a local AI machine.

Need more than 24GB? Get the SO-DIMM version (up to 96GB).

@geekompc sent the unit. I wasn’t paid.

Full review: https://t.co/DHITYCV8bP

[MoHelal @M0hamed__82]

1/ @geekompc IT13 Max review: Core Ultra 9 185H, 24GB LPDDR5-5600, dual 2.5GbE and dual USB4 in a 0.9L box.

I ran it as an always-on Docker box for 41 hours and tested local AI on the CPU, Arc iGPU and NPU. Results 🧵

[MoHelal @M0hamed__82]

5/ Local AI, Qwen3-8B generation:
CPU: 9.5 tok/s at 48W
Arc iGPU: 8.0 tok/s at 28W, with 3.6x faster prompt processing
NPU: 8.55 tok/s at 17W (OpenVINO, 4.87s to first token)
A 14B model fits fully on the iGPU and runs at 5.2 tok/s.

[Dario Amodei @DarioAmodei]

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.

Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.

You can read the full post here: https://t.co/OGyPb7yaYt

[Dan Kornas @DanKornas]

Claude Code internals, without copying Claude Code.

Claude Code from Source is an educational book/repo on the architecture, patterns, and internals of Anthropic’s AI coding agent.

It helps you study production agent design by turning source-map research into 18 structured chapters with implementer deep dives, diagrams, and “Apply This” takeaways you can reuse in your own systems.

Key features:

• 18-chapter architecture map – covers foundations, the agent loop, tools, multi-agent orchestration, memory, UI, connectivity, and performance
• Builder-focused chapter format – combines narrative flow, deep dives, and closing takeaways for implementation
• Agent patterns index – highlights ideas like speculative tool execution, concurrent-safe batching, fork agents, context compression, and file-based memory
• Mermaid diagrams – architecture explanations render directly inside GitHub
• Educational-only source policy – README says it includes no Claude Code source code, only original pseudocode for explaining patterns

Free public GitHub repo.

Link in the reply 👇

[MoHelal @M0hamed__82]

@NVIDIAGeForce #RTXPowersPlay

[Roy van Rijn @royvanrijn]

For curious developers 🧠

I built "The Anatomy of an LLM", an interactive explainer showing how text becomes tokens, vectors, attention, transformer blocks, and finally generated text.

https://t.co/fgCeZuQwJf

[Acemagic_global @Acemagic_]

✨ The ultimate Open Claw guide is coming.

We’re putting together a full deployment tutorial so you can get it up and running.

🔥 36% OFF — Code: ACEX5RETRO
💰Final Price: $895.99
🗓️Ends Mar 31 (PST)
👉 https://t.co/PtLQRccDYY
🔔 Follow us, so you don’t miss the tutorial drop.

#Acemagic #minipc #openclaw #openAI

[Acemagic_global @Acemagic_]

The classic reborn 🕹️ACEMAGIC Retro X5
Retro-inspired design meets AMD Ryzen AI 9 HX 370 power.
Go slow… you might miss it.
36% OFF → $895.99 | Code: ACEX5RETRO
PST: 3/6 5PM - 3/31 11:59PM👉 https://t.co/uXO0R6sbaZ
#ACEMAGIC #miniPC #RetroComputing #RetroGaming

[Acemagic_global @Acemagic_]

Dream machine → Reality🚀ACEMAGIC Tank Centre (M1A PRO+ AI MAX+395): tank-inspired AI power
#acemagic #MiniPC #GameNightVibes #SundayFunday

[MoHelal @M0hamed__82]

@troyaitken_ Setup

[Acemagic_global @Acemagic_]

Tank mode activated. Acemagic M1A PRO
RGB glory, uncompromising power, zero limits.
#Acemagic #TechInnovation #MiniPC

[Acemagic_global @Acemagic_]

Acemagic M1A Pro+ AI MAX+ 395 – 126 TOPS AI Mini PC with RTX 4060 & i9-1... https://t.co/IQdxc63OhA via @YouTube

[MoHelal @M0hamed__82]

Check out my Review of AOOSTAR G-Flip AI-370, the best MiniPC with 5" Touch Screen and AMD Ryzen 9 AI HX 370 (Strix Point)!

Thanks to @AMBERL56 for making this review happen, and for the continuous support.

#MiniPCs #AMD #AI #AOOSTAR

https://t.co/R9I0Plsw73

[MoHelal @M0hamed__82]

https://t.co/ZDVsK4Qcon

[MoHelal @M0hamed__82]

https://t.co/ZDVsK4Qcon