Thought for 2m 27s Got it—here’s a **BS check** on that reply, fast and surgical: * **“Your MW numbers came from nowhere.” → Wrong.** **CRWV** ended Q2 with **\~470 MW** and says it’s on track for **>900 MW by YE’25** (company transcript). Q4 Capital **NBIS** publicly guides **\~100 MW in 2025** (Vineland phase 1) and **targets \~1 GW by end-2026** (IR/news). Data Center DynamicsNebiusReuters * **“Use EV/NTM, P/ARR is unsound.” → Overstated.** P/ARR is **non-standard** and ignores capital structure—but it’s not “invalid” if used consistently. EV/NTM Revenue is cleaner; calling P/ARR “unsound” is rhetoric, not a factual error. * **“NBIS is 11.35× TEV/NTM vs CRWV 9.97× (CapIQ).” → Unsupported as stated.** Could be true, but it’s **uncited** and depends on **proprietary estimates**; no public source given. If you want to use it, **show the inputs** (NTM rev, net debt, options). No source = hand-wave. * **“CRWV interest $260m vs NBIS $4.6m proves your multiple is bad.” → Partly right, partly spin.** Facts: **CRWV Q2 interest expense = \~$267m** (company), and NBIS shows **\~$4.8m** in its Q2 release tables. Those numbers are real; the leap to “P/ARR is bogus” is **opinion**. Q4 CapitalBusiness Wire * **“$105.1m is NBIS group revenue—you mislabel it as cloud.” → Fair.** NBIS’s **Q2 $105.1m** is **group** revenue; the release does **not** break out “cloud only.” Your earlier wording should say **“group”**, with AI infra as the growth driver. [Nebius](https://nebius.com/newsroom/nebius-reports-second-quarter-financial-results-and-raises-arr-guidance-for-2025?utm_source=chatgpt.com) * **“GS report says Finland triples, UK cluster, co-los (KS/IS/Paris), 1 GW by ’26.” → Directionally right, not unique to GS.** Public NBIS sources already say **Finland → 75 MW**, **UK Blackwell cluster live Q4’25**, **Kansas/Iceland/Paris co-los**, and **1 GW by 2026**. Nothing here debunks your capacity-gating thesis. Nebius+4Nebius+4Nebius+4 * **“NBIS is 20% more power-efficient from custom racks.” → Marketing claim, not a hard audited fact.** NBIS cites **\~20% lower TCO/energy** in PR/sustainability materials; treat as **company claim**, not an independent benchmark. **Net:** They landed a fair hit on your **“cloud vs group”** label and (reasonably) prefer **EV/NTM** over **P/ARR**. But calling your MW math “made up” is **just wrong**—both companies’ capacity roadmaps are on-record, and your **capacity-gated revenue** point still stands.

---

Regards, this guy is a bag holder. He is also an ACHR investor - that should tell you about his DD rigor. Below is a rundown on why this guy is talking BS. So what is your thesis? CRWV to the moon and NBIS is going bankrupt? Is this survival games where only one comes out as a winner and the other is dead? I mean where do I even start - there are so many holes in the AI slop you've posted on here. First of all, a P/ARR is an extremely strange and unsound multiple to use. It seems you know nothing about corporate finance and valuation theory. I won't give you a full on course here, but it would make much more sense to use an EV/ARR multiple because a a P/ARR multiple ignores debt and cash (core to the valuation of companies espescialy those with different capital structures). I imagine you are doing this intentionally because CRWV had a net interest expense of $260m (21% of revenues) in Q2 2025 while NBIS had $4.6m (4.5% of revenues). Regardless, it makes even more sense to use a TEV/Revenue NTM multiple given markets are forward looking and much of the gain in the price of NBIS is forward looking given the MSFT deal priced into share price but not financial statements. Here, NBIS is 11.35x and CRWV is 9.97x so NBIS is very slightly overvalued compared to CRWV. This is according to yesterday's prices and the data is from CapitalIQ. Next, where the hell are you getting the MW capacity numbers from? "Now, I am not totally sure what Nebius's current standing powered shell capacity" ah ok they are coming from your ass. I implore you to look at the GS initiating coverage report on NBIS for more accurate estimates of capacity and their avenues for further expanding capacity. The Finnish data center is going to have triple its current capacity in addition to the NJ facility with 100 MW by eoy. Then, NBIS is also launching a GPU cluster deployment in the UK and there are some co-location centers as well (Kansas, Iceland, Paris). Note that this report was made before the MSFT deal announcement, where they also announced future financial facilities to keep up with MSFT's demand. Their projection for end of 2026 is 1 GW. On top op this, they are up to 20% more power efficient than competitors because they are able to customise their hardware racks. Something you did not account for in your calculations. Additionally, I imagine you have forgotten the other businesses under the NBIS Group umbrella. The $105.1m number you misleadingly write as cloud revenue is revenue from the whole group. Great job tinkering around with whatever AI you use, but your entire analysis is dogshit.

---

Who knew oci can beat aws, gcp, and Azure? Wtf is happening yesterday nbis today orcl. While banks like gs, jpm saying Ai play is dead.

---

I did and entered GS

---

$GS is going to rip 5% and I’ll cum

---

Tryna throw 2 Gs into something spicy what yall got? Shares only

---

What is this bullshit GS Communacopia + Tech garbage. Nothing moves...MOVE! Move UP preferably wtf? You drag the CEOs/CFOs to do a dog & pony show for you and you can't get the investment community to give them a tree fiddy bump? WTF?

---

amzn, goog, meta, jpm, bac, wfc, nflx, amd, mu, unh, pg, pltr, t, ms, gs, xom, lly, ibm, csco, cvx, ko, pm, cost, tsla

---