majabbar

Realist, AI Augumented Human..... 🤔

[majabbar @MindLedger]

@TheAhmadOsman How to get a donation of a 4 slot NVLink as I am unable to find one myself. 😀

[Ahmad @TheAhmadOsman]

People keep asking me why do I
focus on fundamentals instead
of agents or shiny products

Shortcuts don’t compound
- Models are still improving
- Agents come and go
- Frameworks churn
- Products age fast
- We don’t know what the next-gen models unlock

Fundamentals stick
> Architectures & models
> Inference
> Memory
> Hardware
> Latency
> Failure modes

When you understand the stack end-2-end,
you can build anything on top
- Agents
- Products
- Companies
- Labs

When you don’t,
you’re gluing demos together
hoping the abstraction doesn’t crack

I’m not optimizing for the next launch
I’m optimizing for the next decade

[Ahmad @TheAhmadOsman]

Let me make Local AI easy for you

Give Codex Cli the article below & tell it:

- Infer the right Inference Engine from your hardware + article below
- Use uv+venv
- Pick the right kernels
- Tune flags, batching, KVCache, etc
- Optimize for your hardware & chosen model

See? Easy

[majabbar @MindLedger]

@TheAhmadOsman Aged well like s wine.

[majabbar @MindLedger]

@TheAhmadOsman Everything is setup at my end 2x3090 ti but I am struggling to find a 4 slot nvlink 🤪 that support amp.

[majabbar @MindLedger]

@TheAhmadOsman It seems you lean more towards Codex than Claude Code. Both have corporate guardrails, though.

[majabbar @MindLedger]

@TheAhmadOsman Guess what! I asked my agent yesterday to find the absolute best way to maximize Tokens Per Second (TPS) while running Claude Max supervision.

The move? It actually went out and scraped your x and repository to optimize itself.

[majabbar @MindLedger]

@TheAhmadOsman Let me ask my agent to take over my x and do as you have just said. 🤪

[majabbar @MindLedger]

@TheAhmadOsman She is a proud lady. Respect 🙏

[majabbar @MindLedger]

@TheAhmadOsman Have some karkadi. 🤪

[majabbar @MindLedger]

@TheAhmadOsman What about gen 4 gpus on gen 5 slots? I have installed 2x 3090ti on 2x Gen 5 slots with x8x8 bifurcation.

[majabbar @MindLedger]

@TheAhmadOsman Tomorrow is a big day for me inshallah. I'm going to fix my new psu, motherboard, and 2x 309ti with nvlink. It's all because of you. 🙏

[majabbar @MindLedger]

@TheAhmadOsman Opensource all the way to AGI 🤪

[majabbar @MindLedger]

@TheAhmadOsman @AnjneyMidha @CS153Systems @LisaSu @satyanadella @garrytan @sdianahu @matiii @andi_blatt @mabb0tt @karpathy @LiamFedus @amppublic It should be a must.

[majabbar @MindLedger]

@TheAhmadOsman THANKS

[majabbar @MindLedger]

@TheAhmadOsman Wow. Useful .. very useful ...

[majabbar @MindLedger]

@TheAhmadOsman Very informative.

[majabbar @MindLedger]

@TheAhmadOsman @sudoingX When is the discord? I have so much to ask during my 0 to Hero journey and may be may regularly share my milestones. I started with 1 3090ti, just got the second 3090ti and am now deciding on a new motherboard and stuff, etc.

[majabbar @MindLedger]

@TheAhmadOsman RTX Pro 6000 is the way to go. A hard wish, though.......

[majabbar @MindLedger]

@TheAhmadOsman @VitalikButerin Spot-on