NVIDIA STOCK DROPS -$37/share (-$18%) : $53mm loss in Q3 due to old Pascal inventory - YET I CAN'T find a 1080ti ANYWHERE for MSRP?????

[Misanthrope]

Sweet, sweet backlash for catering to miners.

[Canada EH]

11 hours ago, hazeyez said:

So - They posted a $53 million loss in Q3 because of old inventory

Well there are many other reasons as well, the story line you hear is what the company wants you to hear. Big corporations are Spin Doctors with a Masters Degree and a PHD and a Doctoral and a 

[YoloSwag]

11 hours ago, Canada EH said:

Well there are many other reasons as well, the story line you hear is what the company wants you to hear. Big corporations are Spin Doctors with a Masters Degree and a PHD and a Doctoral and a 

Definitely. The surplus they used as reason has to be exaggerated that they posted a loss mainly because of inventory costs. They're trying to steer investors into thinking once they get rid of the old stock they'll start making money again.

 

I say they didn't get enough pre-orders before launch on their RTX cards and they had to adjust their expectations. Even so now that it's not actually selling well because they tried to pull an "à la Apple" on folks. It was like their version of VEGA, much expectation yet so expensive.

[bbbbbbb99]

The reason you cant see a 1080ti is because it is perpetually behind you and in order to see it you need ray tracing.

[1988fido]

On 11/16/2018 at 9:13 PM, GoodBytes said:

Nvidia (mostly) makes and sells GPUs, not graphics cards (ok they do make graphics cards, but Titans mostly, and started recently to make Founder Editions models of select cards for consumers, which cost more and not as good). Graphics card manufactures are not buying them, probably because they don't want to be stuck with ones that don't sell.

 

So,

• Nvidia is stuck with too many GPUs, especially that they miss judge miners future demand.
• Manufacture are not making them because now the 10 series is old, or only doing low production number, making the production price increase, increasing the price at retail
• And, in addition, due to the low supply and high demand, retailers charges more.

so we can hope that new budget oriented graphic cards based on pascal will be released  on 20 series
something like 2050Ti will be same chip of 1070 ?

coz they have lots of extra chips laying around and not selling or vendors like Asus/gigabyte not buying to make more 10 series cards, so only way to push them is by rebranding it as new budget cards ?

[GoodBytes]

3 hours ago, 1988fido said:

so we can hope that new budget oriented graphic cards based on pascal will be released  on 20 series
something like 2050Ti will be same chip of 1070 ?

Well I'll argue that the RTX 20 series are just Pascal chip with some tweaks here and there to make it faster, with Ray Tracing cores slap on it, as the actual next generation chip was supposed to be Volta.

(Yes, Pascal is missing from the graph, and magically appeared once Volta unofficially got delayed (no date was ever set for it until later, when the roadmap was revise with Pascal added in between))

 

We do know that the lower end card won't have the RTX branding, as they have too few core for Ray Tracing to be actually useful (but watch Nvidia still attempt at this, with its usual model that makes no sense, next to the GeForce 2030 DDR3 model which will still be a low-end Fermi chip, or whatever old junk). So will there is Pascal? Maybe.. but i expect them to turn off some cores and reduce the performance of them and rebrand the GPUs, and sell them as GeForce GT 2050 or something. Or just through them away and take the hit.

 

[DuckDodgers]

50 minutes ago, GoodBytes said:

Well I'll argue that the RTX 20 series are just Pascal chip with some tweaks here and there to make it faster, with Ray Tracing cores slap on it, as the actual next generation chip was supposed to be Volta.

Not really. Turing inherits Volta architecture and adds the RTX hardware. The only missing thing is the HBM2.

Regarding the roadmap, Volta was simply moved as an exclusive HPC product to promote Nvidia's Tensor DNN acceleration and to serve as a platform to integrate the expensive HBM2 memory -- none of those were suitable for the consumer market. Then Pascal came to fill the gap between Maxwell and Turing for the consumers, as a test vehicle for the 16nm production, with minimum investment in the architecture to speed up the roll-out. It was essentially Maxwell 3.0 with several tweaks for faster VR, updated video codec, but mostly utilizing the smaller fabrication node to bump up the clock rate, as the main performance benefit.

[hazeyez]

NVDA ended Friday's trading session @ $163.41/share - $38.98/share less than the day started.

 

Took a BIG hit there - anyone holding this stock?