Amit Jaiswal

Whatever you do, don’t retire!

[Amit Jaiswal @amitkjai]

3C's - Towards making it worth.
CONFIDENCE on the Business!
COMMITMENT on the Investment Decision!
CONVICTION to stick to it!

[Amit Jaiswal @amitkjai]

@hajiyev_rashad $74-$79 looks to be the supply/demand fight zone.

[⚡️The Traveling Trader⚡️ @ZTheTrader]

Took me a while to understand this when I first started trading/investing so listen up.

If you constantly have thoughts such as:

- It hit an all time high, i'm afraid it tanks now
- I didn't catch the bottom so it's not worth finding a trade now
- It looks like a continuation but the minute I jump in it's going to fall
- I entered a position and it's red now...I must've gotten in too late or done something wrong
- If I don't enter now I'm going to miss the gains everyone else is making
- I'm feeling too much FOMO because I'm not in that trade. I better enter and increase my position size to make up for lost time
- My position is up today and if I don't take my small gains I may lose them tomorrow

None of this...and I truly mean none of this...has to do with trading. These are all human emotions baked into us through a mix of nature and society-induced retail shopper's mentality. These thoughts are trying to protect you from short term pain. They are not meant to be used to objectively analyze trades.

I know it's hard but try to drown it all out. These thoughts will come to you but it doesn't mean you have to act on them. Try to make decisions solely based on data, analysis, backtested setups, patience, and a realistic approach to risk/reward.

[Amit Jaiswal @amitkjai]

@MauroBianchi24 Did you chart Silver and Gold?

[Amit Jaiswal @amitkjai]

@MauroBianchi24 I haven't seen anyone else getting so close with Elliot waves.

[Amit Jaiswal @amitkjai]

@MauroBianchi24 Welcome back!

[Amit Jaiswal @amitkjai]

@MauroBianchi24 Question: When you trace fib series to get the Golden pocket levels, do you use logarithmic or regular scale on the chart?

[Patrick Karim @badcharts1]

We are STILL close to a VERY RARE & HISTORICAL event.

This would be the 2nd time in 125 years we get a "SUPER" Capital Rotation Event.

Stock markets get DESTROYED when adjusted for inflation.

While gold, silver & friends enjoy their BEST years of GAINS and OUTPERFORMANCE!

[Justin Banks @RealJGBanks]

HOW THE BEAR MARKET STARTS

Bull Trend:

8 EMA = momentum
13 EMA = control
21 EMA = structure

Price losing:

8 → momentum weakens
13 → trend control is lost
21 → bear market begins

Don’t overcomplicate it.

Structure breaks Weekly the trend changes.

[Amit Jaiswal @amitkjai]

@WeekendInvestng It may fall thru a little more below 70K, where I see a good support per the volume profile.

[Alok Jain ⚡ @WeekendInvestng]

Those who bought AsianPaints at 10 -20 rs don't really bother if stock is moving +/- 20% near 3000...

generational wealth has been made..

Same for Gold /Silver guys who entered at 5000 per 10gm/25000 per kg

[Rashad Hajiyev @hajiyev_rashad]

Going forward volatility in precious metals market is going to increase. Single day 10-15% moves likely to become a norm. Good luck trading it…

[John @market_sleuth]

Today was a simple garden variety liquidation break. The one thing that’s a concern is the volume. Every down move since Nov is happening on 20-35% more volume than up days. Today is no exception. 25% more volume than any up day this week. I’ve seen this pattern precede every single major correction. Doesn’t mean it will happen soon but the risk is growing.🚧

[Vatsala Singh @_vatsalasingh]

संस्कृत की पूरी वर्ण माला शरीर के अंदर ही पैदा होती है! ध्यान के समय अनहद नाद शरीर के अंदर सुनाई देता है...

[optionGeek ✨ @StockShark16]

Triple witching is tomorrow and it is propping up the market short term. Elevated put open interest (OI) across single-stock futures is creating a support cushion as dealers hedge exposures. This “pinning effect” can stabilize price action into expiry.

However, once these contracts roll off, that support disappears. With positioning reset and hedges unwound, we anticipate the market could face a serious sell-off next week.

[Amit Jaiswal @amitkjai]

@vulturetrades Play

[Jim @JP_Money_95630]

Let's Talk TECH!

My Background =
Valedictorian US ARMY Signaling Intelligence School Late 90's
Then Cisco CCNA, CCNP, CCIE.
(This Took me 10+ Years and $$$ of my Own Money)

I also hold RECENT Certifications from $NVDA = NCA, NCP.
A.K.A NVIDIA Certified Associate, NVIDIA Certified Professional.

My Global Technology Consulting Rate is $400 Per Hour.

I help decide the technology for numerous companies and Many Wall Street Firms who want Multi-Year Growth and Investments.

I've been doing this for the last few years after 25 Years in Corporate Technology Roles as the Leading Delivery Architect.

I've created Billions in Revenue for many companies, much of which are on Auto-Pilot Now that they are fully deployed with revenue creating systems that all started with the right infrastructure decisions many years ago.

** Fun Fact: If you Call 911 and they know where you are at, guess what, that is what we designed in the US Army way back in the 90's called Triangulation. You Take the Distance from 3 Towers using the amount of Power used to reach the handset or mobile device then calculate distance and in turn 3 End Points Land within 10M of the User, Target or person of Interest.
Modern Day E911 on your Mobile Device. =p

So yeah, been around a minute. I used to Split Beams of light in the Livermore Lab to put messages on different colors. We call that DWDM - Bouncing Light off a Prism.

I have 20+ Years Designing Data Centers, Edge Computing and hold an AI Product design background from MIT.

Now let's Talk AI GPU + CPU + High Bandwidth Memory as well as AI GPU Global Saturation.

As investors, All we care about is this 2000/2001 or 1995/1996/1997/1998/1999?

So Let Us Begin.....

Question:
Are We Fully Saturated, Moderately or Not Even Close to Full Global AI GPU/CPU Saturation?

2022: Ignition (ChatGPT moment)
2025: ~5–7M GPUs installed Globally
2028–2029: Moderate saturation
(~30M, plateau begins)
2032–2033: Altman-scale saturation
(~100M, plateau)

The shaded bands highlight the following:

1. Acceleration - Adoption Phase
2. Moderate Saturation (~30M GPU's
3. Altman Expansion phases ( ~100M GPU's)

Let’s frame this AI GPU saturation cycle as a timeline. To answer “when did it begin, when might it slow, when might it end,” we can align it with data, shipments, and adoption curves.

1. When Did It Begin?

2017–2020 (Incubation):

• NVIDIA Volta (V100) & Ampere (A100) enter hyperscaler clusters.
• AI = mostly research + early enterprise pilots.
• Installed base in hundreds of thousands globally.

2022–2023 (Ignition):

• ChatGPT (Nov 2022) → sudden mainstream proof of GenAI.

• 2023 saw the first “GPU crunch”; supply shortage:

H100s selling above MSRP.
• Installed base: <1M H100-eq by end 2022 → crossed ~2M by end 2023.

2024–2025 (Acceleration):

• Mass deployment: Meta, Microsoft, Amazon, Google each order hundreds of thousands of GPUs.
• Installed base: 5–7M H100-eq by mid-2025.
• This is the true start of the “AI Revolution build-out phase.”

📈 2. Moderate Saturation (~30M H100-eq)

Trajectory:
• With ~5.3M NVIDIA accelerators shipping in 2025 and ~6M+ projected in 2026 (Mizuho), plus AMD/TPUs, global adds could be ~8–10M H100-eq per year after 2025.

• Installed base reaches ~20M by 2027 and ~30M around 2028–2029.

Slowdown factors:

• Enterprises + consumer assistants fully deployed.
• Capex intensity starts to taper as replacement/upgrades overtake greenfield installs.

End of cycle:

• Plateau in late 2020s, with refreshes
(Blackwell → Rubin → after) rather than exponential growth.

• Similar to how cloud data centers hit maturity ~10–12 years after their launch.

🚀 3. Altman-Scale Saturation
(~100M H100-eq)

Trajectory:

• If OpenAI’s 100M GPU vision is the true horizon, at current/future ramps this means ~15M units per year must be shipped & installed for the next 5–6 years.
• That implies hitting 100M around 2032–2033.

Slowdown factors:
• Power grid limits — IEA says data center demand doubles by 2030, AI being the main driver.
Grid build-outs may lag chip shipments.

• Capital intensity — At $25–40K per GPU-equivalent, 100M units = $2.5–4T hardware capex.

Even spread over 8 years, this rivals cloud build-outs.

• Tech shifts — Breakthroughs in efficiency (e.g., low-bit inference ASICs, optical compute) could reduce GPU-equivalent needs.

End of cycle:

• Early 2030s plateau, unless demand scales beyond consumer+enterprise into ambient AI everywhere.

🧭 4. Summary Timeline

Start: ~2022 (ChatGPT moment)

Moderate Saturation (~30M GPUs):
~2028–2029

Altman Saturation (~100M GPUs):
~2032–2033 Slowdown/End:

• Moderate case: late 2020s → refresh cycle dominance
• Altman case: early 2030s → plateau unless “AI in everything” accelerates further

👉 TL;DR:
We’re ~3 years into a 7–10 year build-out cycle.
Moderate world ends before 2030.

Altman-scale world extends into early 2030s, but requires massive power + capex alignment.

[Amit Jaiswal @amitkjai]

@LuxAlgo Which indicator is this?

[Amit Jaiswal @amitkjai]

@VijayThk The monthly chart also provides a good view: