Register and share your invite link to earn from video plays and referrals.

Alex Ziskind
@digitalix
Software engineer making videos about dev gear, local AI, and weird tech experiments
1.5K Following    27.5K Followers
Faster chips, new bottleneck 🤯 Two clustered M5 Ultra Macs ran SLOWER than one: 35 tok/s vs 53. Which is strange, considering my experiments with M3 Ultra were the opposite. Keep the GPUs awake and they jump to 58 tok/s.
Show more
M5 Ultra vs dual sparks on deepseek-v4-flash. Disag still on the table
i don’t understand why people are so into the benchmade. can someone explain? I have 3 Kizers and they all are great. even open and close easier than the benchmade
last one at my Micro Center. Buying before another price hike.
can’t wait to watch this. Everything about it is just so unique and fresh.
Here is official trailer, You Can See Everything
never renewed anything faster than this
1.2M views in 48 hours is crazy
as fast as computers are these days, I spent a LOT of time waiting for computers to finish what they’re doing
New MacOS 27 setting for clustering Macs for AI
0
36
1.1K
43
Forward to community
wow, this is fantastic
We are now at 3740 tok/s prefill, 149 tok/s decode batched on M5 Ultra for Qwen 3.8 Flash Next This is nearly double what it was yesterday
Another Trillionaire freshly minted
Congratulations to AMD, the worlds newest Trillion dollar company. 🎉 It's been such a long journey but no other company is as deserving of such a valuation.
Seems NVFP4 is slightly better in prefill, but not in decode
M5 Ultra trying out NVFP4 vs MLX 4-bit. Only found Qwen3.8 27B for testing so far
M5 Ultra trying out NVFP4 vs MLX 4-bit. Only found Qwen3.8 27B for testing so far
Ok here come the Duo apps! this one looks pretty slick. And it’s in TypeScript!
Guitar Pocketbook, my first iPhone Duo app. - AVAudioEngine graph in TypeScript - Folding drives the mix (reverb and tone) - Hinge glows with audio frequency - WebGPU (metal) handles all visual fx Built with @NativeScript
Show more
nice to see some color options
#ProArt# #GR1X# brings NVIDIA RTX Spark™ to a compact #AgenticAI# PC, built for AI agents, creative workflows, and always-on performance. Together with the #ROGNUC# 16, it showcased powerful computing in remarkably compact designs. Your AI Agent, always On. #IBC2026#
Show more
ola hej. this dell 52” is insane. how am i gonna transport it to my new office though. 🤔
NativeScript is already folding this way and that way day zero. nice!
iPhone Duo, develop it with NativeScript now.
Finally, Xcode 27.1b is out with fresh new foldable sims