RTX Spark looks like a Windows mistake, but NVIDIA may be playing a much bigger game. Try out ChatLLM - http://chatllm.abacus.ai/ltf and Abacus AI DeepAgent - http://deepagent.abacus.ai/ltf 🛒 Gear Links 🛒 👀 2 400Gbps switch: https://bhpho.to/4r9vJi1 👀👀 4 400Gbps switch: https://bhpho.to/4qQqOlI 👀 MSI EdgeExpert now available here: https://amzn.to/40luUHG 👀 ASUS GX10: https://amzn.to/4kROLb8 👀 DGX Spark on SALE: https://amzn.to/4kOnczp 🪛🪛Highly rated precision driver kit: https://amzn.to/4fkMVfg 💻☕ Favorite 15" display with magnet: https://amzn.to/3zD1DhQ 🎧⚡ Great 40Gbps T4 enclosure: https://amzn.to/3JNwBGW 🛠️🚀 My nvme ssd: https://amzn.to/3YLEySo 📦🎮 My gear: https://www.amazon.com/shop/alexziskind 🎥 Related Videos 🎥 🧳🧰 Mini PC portable setup - https://youtu.be/4RYmsrarOSw 🍎💻 Dev setup on Mac - https://youtu.be/KiKUN4i1SeU 💸🧠 Cheap mini runs a 70B LLM 🤯 - https://youtu.be/xyKEQjUzfAk 🧪🔥 RAM torture test on Mac - https://youtu.be/l3zIwPgan7M 🍏⚡ FREE Local LLMs on Apple Silicon | FAST! - https://youtu.be/bp2eev21Qfo 🧠📉 REALITY vs Apple’s Memory Claims | vs RTX4090m - https://youtu.be/fdvzQAWXU7A 🧬🐍 Set up Conda - https://youtu.be/2Acht_5_HTo ⚡💥 Thunderbolt 5 BREAKS Apple’s Upcharge - https://youtu.be/nHqrvxcRc7o 🧠🚀 INSANE Machine Learning on Neural Engine - https://youtu.be/Y2FOUg_jo7k 🧱🖥️ Mac Mini Cluster - https://youtu.be/GBR6pHZ68Ho * 🛠️ Developer productivity Playlist - https://www.youtube.com/playlist?list=PLPwbI_iIX3aQCRdFGM7j4TY_7STfv2aXX 🔗 AI for Coding Playlist: 📚 - https://www.youtube.com/playlist?list=PLPwbI_iIX3aSlUmRtYPfbQHt4n0YaX0qw — — — — — — — — — ❤️ SUBSCRIBE TO MY YOUTUBE CHANNEL 📺 Click here to subscribe: https://www.youtube.com/@AZisk?sub_confirmation=1 — — — — — — — — — Join this channel to get access to perks: https://www.youtube.com/channel/UCajiMK_CY9icRhLepS8_3ug/join — — — — — — — — — 📱 ALEX on X: https://x.com/digitalix #rtxspark #gdxspark #llm
ADVERTISEMENT
A thin powerful gaming laptop sounds nice, huge ram for creative is also nice, run some hobby mini llm also seems nice. But the cost and window on arm is a huge question. Consumer definally won't spend more than 2000 dollar for that.
I use strix halo as my development PCs running windows and it’s super useful to have one machine ‘do it all’. I have a normal mini desktop and the rog flow - but it’s missing CUDA. this is a really good concept particularly if it brings windows on arm out of the doldrums
and the price?
Hope windows on arm able to make steam users and games move to arms chip with good translation software.
I've been wanting to test out the snapdragon's but they're too expensive for not having full application and Linux support and strangely used ones hold their value which to me is weird.🤔
I think this will be interesting once it's released. The question will become will it enable people like steroids but for computer effectiveness of a user *OR* will it be agentic laptops no longer needing humans to do most stuff. 🤔 Plus I'm curious about the ram speed and pricing.
can you do a video or post spec of that desktop you have on your desk please? Looks fireeeee
Wait a sec.... Wasn't the last generation of this hardware run on dgx OS which is a Linux or Ubuntu type os? Or is it only windows on purpose? Can you add an external GPU? If they go into desktop format you should be able to add any rtx card I would think.
Just forget Adobe! Black Magic Designs' DaVinci Resolve is my video editing choice. ComfyUI must work on this or it's not a good value proposition for me.
I’m curious how N1X stands up to 5090 for wan 2.2 considering that it may fit the whole model in ram without any risk of hitting limit but how much will the 128gb models be … I also think they will release a desktop version thats much better later.
Thanks for keeping us in the loop Alex. Personally I use both Windows and OSX. I like them both.
If/when they do port to Linux and the desktop version has PCIe, this will hopefully allow people to choose their own networking level. With any luck, this could result in very large clusters for an almost reasonable price. Will it be a more practical solution than DGX? I have no idea.
The tomatoes fooled me. It took like 15 seconds before I remembered they weren't real.
I currently have an M1 Air with 8GB of RAM, and now I'm unsure if it's worth upgrading to the M5 Pro with its 3-nanometer processor, 24GB of LDR5X RAM, and 300GB of "top-tier" LDR5X (DRAM) memory, or if I should wait for the M6 Pro with its 2-nanometer processor, 24GB of LDR6X RAM, and 370GB of "top-tier" memory.
It seems to me it depends on your use case. It would be great for me at the right price. I need a lightweight cool running machine with good graphics to replace my nano gen one
Interesting, but we will miss the stacking capability when we need more vram, more or less will be like current Strix Halo devices
Let me know how many games run out if the box (-)
Memory bandwidth 300GB/s… Cuda will work nice for image/video inference, but local model inference speed is gonna be just ok unless you only use the Qwen MOE models. Nemotron 3 local models in practicality is just meh compared to Qwen3.6 models. Price guess, maybe $3k to $4.5k range considering the DGX Spark sells for $4k when it came out. Maybe chips crisis may up that too.
Can you install the windows on arm in the dgx spark, and review it ?
I'm sticking with my plan of an AI workstation build with a used Threadripper 3970x and either a set of four Intel Arc B70s, or perhaps even a Crescent Island 160GB vram card. I'm hoping Crescent Island has a variant with a built in blower for workstation build otherwise I'm going to turn to one of those custom modded solutions.