Softbank Sells Entire $3.6 Billion Nvidia Investment

I got out of Nvidia is April when the stock was at $240 by selling all my Nvidia cards and mining rigs, I am no hedge fund manager but I saw this coming
 
So with the stock up 3% they lost a potential gain of 3.6 billion x 0.03 or 108 million dollars.

On the other side AMD is up 18% since January 10 giving them a potential gain of 648 million had they reinvested it in AMD.

If I am reading this right since AMD has a market cap of 23.47B then this hypothetical 3.6B investment would be more than 10% of the company.
 
I don't care if it's NVIDIA, or who ever, but when you overprice your products, it's going to be hard to sell them. Makes perfect sense to me! And I believe most will agree, NVIDIA's latest GPUs are simply overpriced. It would do them good to slash those prices soon. But hey, what does some guy who goes by the name of DookieDraws, know? 😛
 


Poor competition doesn't help but the reality is the consumer market is not where the margins are. Thats HPC. They make more per Quadro or Tesla sold than they do Geforce by far.
 

Well yeah, if you look at the single most expensive model of the most expensive consumer GPU and gaming monitor on the market, it's going to cost a lot...
 
I sold my stock the day Turing launched and if I did so a few days later I'd lose money. The public reasons that Nvidia provided for crashing are nice to muddy the waters and I understand that if they flat-out admitted that it's because Turing failed, they'd see even fewer Turing sales and that would be all over the tech news. CPU sales are higher than ever, yet GPUs are failing when PC gaming and GPU compute are more popular than ever, what could it be though, I wonder!

The reality is that the majority of people don't need new GPUs - they upgrade when there's a good opportunity. Sort of like there is a good opportunity in the CPU market with Ryzen and the core count increases by Intel. Nvidia does not provide such products. They thought that in the times of limited competition they can get away with whatever they please, they released Turing, and found out that people not wanting to buy any GPU if the best options are so weak is a factor as well. That translates to their business customers as well - if their Pascal-based cards are almost as good and they'd have to invest a lot for smaller than ever performance-per-watt gains, what's the point?

Nvidia isn't selling because they're not offering any substantial performance progress. That means they don't have anything that people want to buy at this point, even if they had the vast majority of the market to themselves. If a 3 or 4 year old GPU is pretty much as great as the new offerings, why bother? We were all off-put when Turing launched, and there's no reason to believe it doesn't translate to the casual observers as well.
 


Some people would believe that when a major investor sells their share, it's a good time to buy for a bounce-back. I think some of those people also can't believe that Nvidia isn't "done going down" - they see this as a major company currently selling at below their value. That's probably why the stock is up. Again, I sold my stock when Turing launched, it dropped by double digit percentages since then, and I don't think there'll be a good time to invest until the successor to Turing comes out (and it's good) and Intel and AMD end up not being competitive with it, which is a lot of whens and ifs at the moment, and probably a while until Turing's successor releases in the first place.
 
Nvidia will learn their lesson and move on and likely fair very well in the future. The screwed up the 20XX cards by thinking people would pay those kinds of prices for features they won't need or use for years. A 1080 will work for years and costs 1/3 the price. They should have just made more of them watched the money flow in.
 


That's $1964 + $1511 USD = $3475. Sickening overpriced. I feel for you.

Here - $3600 for the screen, $2500 for the graphics card, AUD
In USD - $2557 + $1775 = $4332
In CAD - $3385 + $2352= $5737.

The graphics cards are typically priced > 40% over the highest price the GTX cards reached at the peak of crypto. The Australian retailers (for the most part), didn't drop prices after the crypto bubble burst.

Australian retailers typically don't lower prices for any reason - not by much anyway. For most people RTX cards will never be affordable. I predict in 2025 I'll still be running my GTX 1080 Ti (assuming it lasts that long), or be forced to drop to the bottom end of the market.
 
Nvidia is a marketing driven corp. The rift between the talky beancounters and their engineers is clearly visible.

A 1080 is just fine. It does my 3d cad, 4k and 8k video acceleration and I play games at 4k in high to ultra settings.

RT to me is an immersive, real and sharp thing. Thats what my raytracer does, of course much slower because IQ counts. But it looks real. A 2080 is incapable of doing anything remotely like that in real time. It looks liked shined up phong shading to me. The point of RT is to make it look real. Nvidia failed to deliver and the product is essentally dead in the water...at any price. So my lightly OC 1080 is near 1080ti speed and a 2080 is not far ahead.

I sit it out till a 4080 or similar specced AMD is available and I favour AMD over Nvidia. So the bias is set. Good luck and good ridance Nvidia. I prefer to buy from engineering corps instead of fluffy marketeers.
 

No, because market saturation. They need to sell upgrades to existing customers. If it weren't so expensive, a RTX 2080 Ti would be a decent upgrade from a GTX 1080.

Also, while 14 nm are certainly cheaper than they used to be, I imagine the price has leveled out. There's only so cheap they can make them. And they surely want to protect their margins, meaning they won't want to drop prices too much.
 
TJ Hooker 14 hours ago

Anonymous said:
its 350 million not 3.6 billion you missed a zero.

What? No, Softbank did indeed sell 3.6 billion USD worth of NVDA.

than it is not 39.8 billion yen.. hence he missed a zero.
 

I never said anything about AMD's stock, but I'm among those who thought AMD was in real trouble, with the launch of Nvidia's RTX cards.

What changed? Well, DXR performance is weak and support still hasn't really materialized. I think DLSS has yet to prove the game-changer they claimed. And the cards' prices haven't really dropped, even though production has had time to ramp and the GTX products they replaced are now pretty much sold out.

I will fully admit that Nvidia's new lead seems to have been illusory, and AMD now could have a small window where they could mount a credible threat. I still see Nvidia with an overall lead, but it looks like they overplayed their hand and it's cost them dearly. That said, in the long run, the new tech in their RTX cards could still turn out to look like a smart investment.

But, for the record, I'll admit to believing the hype, and I'll own my misjudgement. IMO, what makes this saga fun to watch is you never quite know for sure how it's going to turn out.

P.S. another thing I wouldn't have guessed is Nvidia's support for FreeSync coming at this time. I figured it'd take a period of comparable weakness for them, but I thought that wouldn't be until AMD had a stronger contender in the market. Did they get spooked by rumors of Radeon VII, similar to how rumors of Vega 64 seem to have caused them to launch the GTX 1080 Ti with more aggressive pricing than people expected? Or were they just desperate to improve the appeal of their products, in hopes of eking out a little more sales volume to help them through the current slump?