Thomas Lewis

First there was planning and then there was chaos.

[Thomas Lewis @ThomasLewis717]

@SamErde Exactly! Thank you PS master. Trying to catch up :D

[Leica Rumors @LeicaRumors]

Light Lens Lab 35mm f/2 “Apochromatic” M-mount lens production and design update:
https://t.co/vgajJObgd7

[NSFVoyager2 @NSFVoyager2]

I am currently ~19h 52m 51s of light travel time from Earth (2026:251:000000:2L)

[Dark Web Informer @DarkWebInformer]

"whats the best wallet for xmr?"

Feather in my opinion.

Dread Thread: https://dreadytofatroptsdj6io7l3xptbet6onoyno2yv7jicoxknyazubrad[.]onion/post/1eae3d3e885062f0b695

[Eleonora Sayaka Chialva @EleonoraSayaka]

Here are a couple of photos from my latest repair🔍

[Interesting AF @interesting_aIl]

FireFox is the last major browser that continues to support uBlock Origin

This comes after Microsoft Edge and Google Chrome removed the original version

[vxdb @vxdb]

LiquidNetwork hackers refunded 3,400 of the 4,000 stolen bitcoin. Keeping 600 BTC for themselves as a bounty.

[International Cyber Digest @IntCyberDigest]

‼️ Switzerland is moving away from Microsoft 365 and is making open-source alternatives available to 3,000 federal staff.

The Federal Chancellery's programme covers:

* Document editing: Collabora Online
* Email: Open-Xchange
* Calendar, contacts, tasks: Open-Xchange
* File storage: Nextcloud
* Project management: OpenProject
* Knowledge management (wiki): XWiki
* Audio and video conferencing: Nordeck (Jitsi)
* Chat; product: Element
* Identity and access management: Univention

They are investing millions in the development of a sovereign Swiss open-source platform for office automation. A 172-person pilot rated it strong on documents and collaboration, weak on large video calls, and viable as a fallback if Microsoft 365 goes down.

[Volatile Markets @volatilemarkts]

Twelve machines. 2.4 terabytes of unified memory. One coordinated inference problem.

Seven DGX Sparks: 896 GB, CUDA, brutal at compute.
Five Mac Studios: 1.5 TB, Metal, brutal at bandwidth and unbeatable per watt.
Plus three Mac minis, a 5080, some V100s.

At 4-bit, that Studio memory holds a 1 to 2 trillion parameter model with room left over for a 900,000-token context. The weights fit. That was never the problem.

The problem is prefill. Before a model with that context says one word, it has to read everything you gave it. On Apple Silicon that is roughly 400 tokens a second. A 900K-token load is over half an hour of silence. Decode is fine, 25 to 30 tokens a second all day, quiet, 300 watts. It’s the first word that costs you.

The Sparks prefill four to five times faster and can’t hold the model. The Studios hold the model and can’t prefill. Everyone with mixed silicon owns both halves of the answer and no way to join them.

So we joined them. NVIDIA prefills, Apple decodes, one request. Two engines that share no cache format, no framework, no vendor. Instead of transferring a cache neither can read, the prefill box computes the decoder’s finished cache using the decoder’s own weights and writes it into the decoder’s prefix store. About 10 KB per token crosses the wire, over plain 10 gigabit Ethernet through the two switches in the first picture. No RDMA, no Thunderbolt.

Measured today on DeepSeek-V4-Flash, 284B, 241,000-token cold load:
Mac Studio alone, 12 minutes to the first word.
Two Sparks feeding it, 3 minutes.
Same prompt again, 19 seconds.
Decode identical. Answers identical.

That ratio is what makes the goal real: prefill 900K on the Sparks for a trillion-parameter model living on five Studios. Tonight we took the prefill window from 262K to 524K. Not finished, and every number gets posted either way.

Why this is a paradigm shift for us: our agent is persistent and has 54MB of memory files..and then we drop a transcript or a codebase on top of it mid-conversation. The wait was the product’s real cost. It isn’t anymore.

All credit to everyone who contributed to these concepts before us. We distill knowledge from all the greats and give credit to all. Standing on the shoulders of GitHub wizards unapologetically without fear of failure or judgement. Local ai must win!

https://t.co/X94wAZsND8

#localai #heterogeneousinference #dgxspark #applesilicon

[Thomas Lewis @ThomasLewis717]

@GrapheneOS Please make it opt out at least, please! The RCS introduce a wide attack surface and not a viable option for some users.

[Sam Erde @SamErde]

Raid my stash of PowerShell scripts from the past 10+ years of work in Entra, Active Directory, Microsoft 365, Exchange Server, Server, and more. Hopefully these save you from having to write a few things yourself!👨‍💻
https://t.co/FAAw0RG0qw

[Mr. Macintosh @ClassicII_MrMac]

The current state of VMware:

[Thomas Lewis @ThomasLewis717]

@S1lky_1337 Never publish the details till 0day get patched. Just profittttttt :D
P.S. the mosaic is reversible.

[Silky @S1lky_1337]

Found a built-in Windows driver that gives you arbitrary kernel memory read/write from local admin. No BYOVD required. My PoC elevates the current process to SYSTEM and neuters any target process token in-place (privileges, integrity, groups) via DKOM.

[Lintaroh @tanakayottt]

MC68030の回路書いてる
32bitアドレスのデコードがだるいですねえ

[misaki @tdatwja]

Justice Department and FBI Seize Platforms Operated and Used by China State-Sponsored Hackers to Target U.S. Critical Infrastructure

・南京鑫玖维网络科技有限公司

https://t.co/UPIGvAJLEK

PDF: https://t.co/NEfehIWcIQ

[Kağan IŞILDAK @kaganisildak]

🧿🧿🧿🧿🧿🧿🧿🧿🧿
🧿 ios 26.4 on QEMU 🧿
🧿🧿🧿🧿🧿🧿🧿🧿🧿
fully emulated