isc0o0

Robots, Ai, humanity, freelancing
فانتزى بريمرليج ، خيال واسع ، روبوتات وذكاء اصطناعي
لعبتى المفضلة برو سوكر ايفولوشن pes 2013

[isc0o0 @essam_madridi]

Milestone 1
A real-time voice chat assistant powered by LLMs🧑‍🦯

Backend: 98% MiMo-V2.6 Flash @XiaomiMiMo
UI: Kimi K3 @Kimi_Moonshot

Spent nearly 10 days vibecoding this.
Thought it would be easy. It wasn’t.
Discovered other engines, models, and some hard scenarios along the way.

[Marouane Lamharzi Alaoui @marouane53]

The biggest cheat code in agentic coding is having Claude Opus 5.5 open a PR, wait for Codex to automatically review it, then fix whatever Codex finds. Repeat this a few times before merging.

Opus might be the best coding agent, but GPT Sol 6.1 is the best code reviewer. It catches things Opus misses.

[isc0o0 @essam_madridi]

If you can't update your ram update your bios 😄
Thanks @GIGABYTEUSA you teached me to never leave @ASUS

[isc0o0 @essam_madridi]

@vvvvqqqqzzzx Before i went to gen14, I used gen4 build for over a decade heavy use and still fine. Go for the used market if budget is tight.

[isc0o0 @essam_madridi]

@dhh I was trying your try installer on windows but found it tries to download huge files which i don't like I prefer offline installer plus image should be seperate.

[isc0o0 @essam_madridi]

@Electronynet المستخدمين الشراء الان قبل ارتفاع جديد ام الانتظار لهبوط قريب! ماذا تتوقع؟

[isc0o0 @essam_madridi]

@mr114514A @deepseek_ai Yes I know i just hoped they join the path of Qwen in this area instead of relying on one lab.

[isc0o0 @essam_madridi]

@chairsister Interesting didn't think that was possible but could save us the price of another 64

[isc0o0 @essam_madridi]

@AIVibecode صحيح انتظارك لترى نتائج لمخيبة تلو الأخرى محبط ومقرون بغضب داخلى .. لكن أعتقد الأمر يختلف مع استخدام نماذج قوية

[isc0o0 @essam_madridi]

@KosmischerK @coldniko What!! is 9070 is that strong vs rtx?

[isc0o0 @essam_madridi]

@sudoingX There are already CXL memory expander cards (Type-3 devices) that plug into a PCIe slot and add hundreds of GB (or even terabytes) of extra memory. Servers and some high-end workstations can use them But bandwidth and latency are still far worse than real on-board GDDR/HBM.

[isc0o0 @essam_madridi]

@murasametech It's just two and There's no place for other GPUs 😂

[Niko Veit @coldniko]

@murasametech I mean more is always better BUT 64gb is efficient enough

[isc0o0 @essam_madridi]

@QwenDevs used to ship LLMs nonstop...
Now at the most exciting times with Strata making Qwen3.8-Flash-Next fly on consumer GPUs .. they are quiet on Qwen4!

October is almost over… time to drop Qwen4 guys 🧑‍🦯

[Niko Veit @coldniko]

@mapte_fireball I always base things on 64GB + 12GB Vram

[isc0o0 @essam_madridi]

@Dr_Singularity When we'll get consumers chips/Gpus that worth buying like old days?

[isc0o0 @essam_madridi]

@coldniko @murasametech Windows eats half of that! 🤔

[Niko Veit @coldniko]

@murasametech You dont need to upgrade your memory. 64GB is just fine to run Q4

[isc0o0 @essam_madridi]

@NVIDIARTXSpark @ASUS #ShareYourSpark I will build WordPress themes that's awesome yet so fast 🤓

[isc0o0 @essam_madridi]

@VectorCrossProd This what my mind headed i have 1 3060 already hope that could combining it with the 50 give me a nice speed boost plus that additional 4gb vram i was can considering if identical old cards out weight the speed of mixing 3060 with 5060ti

[isc0o0 @essam_madridi]

@VectorCrossProd I'm thinking of 5060 ti 16gb do u think it's worth over two 3060!