September 17th was the official launch day...1 month later and barely any cards available with Nvidia even confirming that they will not be widely available until early 2021...Do you have proof of that?
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
September 17th was the official launch day...1 month later and barely any cards available with Nvidia even confirming that they will not be widely available until early 2021...Do you have proof of that?
September 17th was the official launch day...1 month later and barely any cards available with Nvidia even confirming that they will not be widely available until 2021...
And how is this any different for the GTX 1080 launch, which also used special GDDR5X memory?
Will you please shut up about how your little toesies have been stomped upon by the big Nvidia Giant, and just pretend for once that you actually understand how demanding it is to launch a piece of hardware this impressive worldwide?
https://www.extremetech.com/gaming/...are-in-short-supply-experiencing-price-spikes
https://wccftech.com/nvidia-geforce-gtx-1080-success/
Both of these stories make it clear how poor the availability was in the first few months of production. Anyone still want to argue that water is not somehow wet with me?
The RTX 3090 is twice the die size of the 1080, and is the first card to be mass-produced at such a large size on an advanced node. They could also be running out of GDDR6x chips, as those did not exist until August (and only have a single supplier).
The key here is, we will never know what caused the shortage, just like we never found exactly why the Pascal shortage happened. There are likely many different things working together to hold-back getting these cards to retail.
There’s also the Chinese/India border dispute which effects mining, work stoppages on many others for COVID which is messing with raw materials. Quarantine protocols in place for the transport of said materials and entire fleets of ships that are embargo’ed preventing them from docking. Global trade routes are really impacted by everything going on which can’t be good for any bodies distribution, most of those issues aren’t expected to be sorted until March of 2021 if at all.And how is this any different for the GTX 1080 launch, which also used special GDDR5X memory?
Will you please shut up about how your little toesies have been stomped upon by the big Nvidia Giant, and just pretend for once that you actually understand how demanding it is to launch a piece of hardware this impressive worldwide?
https://www.extremetech.com/gaming/...are-in-short-supply-experiencing-price-spikes
https://wccftech.com/nvidia-geforce-gtx-1080-success/
Both of these stories make it clear how poor the availability was in the first few months of production. Anyone still want to argue that water is not somehow wet with me?
The RTX 3090 is twice the die size of the 1080, and is the first card to be mass-produced at such a large size on an advanced node. They could also be running out of GDDR6x chips, as those did not exist until August (and only have a single supplier).
The key here is, we will never know what caused the shortage, just like we never found exactly why the Pascal shortage happened. There are likely many different things working together to hold-back getting these cards to retail.
Huh? I thought the lack of product was a nefarious plot by Nvidia and they have tons of chips just waiting to flood the market.
September 17th was the official launch day...1 month later and barely any cards available with Nvidia even confirming that they will not be widely available until early 2021...
I remember when the heatsink was the issue...GDDR6X not Samsung 8nm is the issue.
I remember when the heatsink was the issue...
That is still possible... Perhaps they are all having problems getting good supply of retail packaging. I mean cardboard companies are hit hard right now.
I'm going to go with Occam's razor on this one.... most likely its the actual GPUs, not components.
If AMD has a hard launch in a couple weeks I guess the its the ram supply argument will have to shift.
If you'd pay more attention to the article you're mocking, Nvidia is holding back GDDR6X, not the Samsung based GPU .
No what caused the shortage was nvidia rushing to market before the product was ready, to get a leg up on amd. The proof is in the pudding with the quality control (bad caps) and supply. But at least we can still brag about "master race". Got to give them props on keeping us in the lead! LONG LIVE PC!!!And how is this any different for the GTX 1080 launch, which also used special GDDR5X memory?
Will you please shut up about how your little toesies have been stomped upon by the big Nvidia Giant, and just pretend for once that you actually understand how demanding it is to launch a piece of hardware this impressive worldwide?
https://www.extremetech.com/gaming/...are-in-short-supply-experiencing-price-spikes
https://wccftech.com/nvidia-geforce-gtx-1080-success/
Both of these stories make it clear how poor the availability was in the first few months of production. Anyone still want to argue that water is not somehow wet with me?
The RTX 3090 is twice the die size of the 1080, and is the first card to be mass-produced at such a large size on an advanced node. They could also be running out of GDDR6x chips, as those did not exist until August (and only have a single supplier).
The key here is, we will never know what caused the shortage, just like we never found exactly why the Pascal shortage happened. There are likely many different things working together to hold-back getting these cards to retail.
TSMC already produces the A100 chip for the enterprise sector so they wouldn't have to retool anything but it would significantly drive up the cost of ampere if they did. samsung producing it is likely why they were able to cut the price by 40%I find this rumor hard to believe. Retooling it to work with TSMC wouldn’t be cheap unless NVIDIA feels Navi 21 is too big of a threat so they’re creating 7nm super cards on tsmc. Either way I don’t see that happening till at least June if not longer.
I remember when the heatsink was the issue...
And handcrafted by artisanal heatsink smiths who produce them in perfect small batches, the way heatsinks were originally forged, with care, one at a time.It is. It's expensive.
I don’t believe this. May be for super series late next year. September time frame. Rumor is also that tsmc needs more 7nm order. I call bullshit on that. Between AMD GPUs consoles, CPUs I don’t think they are hurting much.
I guess we will see may be they will be forced to do it because of competition. But either way I don’t expect it to be until late next year.
Either way they are already hurting with BOM on the card I don’t see them going to tsmc without increasing costs. May be super series that will be another 100-200 bucks next year to cover that cost.
And handcrafted by artisanal heatsink smiths who produce them in perfect small batches, the way heatsinks were originally forged, with care, one at a time.
There's nothing special about the design. It looks different but it functions just the same. It's not a cost outlier, and it's not a production bottleneck. That's spin.Mocking it isn't going to change its cost.
Either way they are already hurting with BOM on the card
Why? every rumor says that AMD is sticking with the same old GDDR6 plus cache. It's the brand-new single-supplier GDDR6x that is likely to be a major cause for RTX 3080 (10 chips) and RTX 3090 (24 chips) not appearing in retail.
AMD can even ship 16GB GDDR6 on that same 256-bit RX 5700 XT memory bus, with higher-density chips. They have several different supplier by now, so it's a bit easier to keep-up with demand by now.
GDDR6x is still lower-density-only (which is why Nvidia smartly went with 10GB for the 3080.) I'm sure having even more chips on both video cards than the GTX 1080 did will result in a shortage.
If you assume 33/66% split for 3090/3080, they are shipping ALMOST DOUBLE THE AMOUNT OF GDDR6X -per-card they did for each video card, compared to the GDDR5XX GTX 1080 launch. It would take a supply miracle to pull that one off. on the day of launch in the Year of our Covid.
Unfortunately, it was too far along to have been avoided after ccovid hit.
Have we seen any proof of this anywhere?
You do realize your posting in a thread with the word allegedly in the title.... If you wanted facts this isn't the thread for you . At least he was nice enough to literally put it in the title so you could avoid it if you desired, but you obviously felt the need to comment anyways.Its these leakers who make up shit and tweet it and/or make clickbait YT videos and never offer a source or shred of proof. Then that gets regurgitated by their followers like it’s fact.
While you make some good points, you are pretty much falling into the same trap that they do. What you think Intel and nVidia are angels? Who here believes that? No one, which is why you have all of these rumors in the first damn place. Yes there are shortages everyone has experienced them. The reason why people are so up in arms is because most people can recall previous launches. We're getting ready to close out month 2 since the Nvidia launch with people still largely not able to get the card they want at the original suggested MSRP.Why wouldn't I comment? I've noticed a concerted effort lately to smear NVIDIA for part shortages as if they're the only company on the planet that's experiencing it and then you get a range of ridiculous conspiracy theories that go with it--like them withholding hundreds of thousands of cards to create artificial demand to barely being able to meet their BOM and making no profit on FE cards. But when you probe even a little bit, none of these tinfoil hat theories have any substance to them, they're basically propagated by a few people (e.g. MLID) and treated as fact.
This isn't just limited to NVIDIA only, we also have guys like Not An Apple Fan (who MLID likes to give credit to) hyping up AMD fans with nonesense like, "yeah lets get on the choo choo train, oh btw don't forget to donate to me" and telling everyone he's got inside knowledge that AMD has a faster card they're hiding because they want to "jebait" NVIDIA.
What's factual is the Ampere cards have severe shortages and nobody outside Samsung and NVIDIA know why. Jensen has said it's due to unprecedented demand and he's probably not far off, people have gone nuts in trying to get 3080s, especially since they got Covid money to spend and a lot of people who skipped Turing have the upgrade itch. Then another fact is the pandemic and the disruption in the supply line, brand new memory type and a new foundry they've never done business with. All those would add up to shortages, but that doesn't mean it's some grand conspiracy NVIDIA has cooked up to really screw it's customers--that's just ridiculous.
I'll put money on Zen 3 selling out instantly and people not being able to find 5800/5900s in stock anywhere for weeks. The same may or may not happen to Navi 21 because traditionally AMD has far less demand than NVIDIA and they might be able to whether shortages better but even they'll sell out pretty fast and not have new inventory for awhile. If you look at past graphics card releases, they always instantly sold out and were unavailable for long periods of time but again, Ampere is unique because of the pandemic, new memory type and new foundry so it compounded things and just resulted in a terrible launch. But some nefarious NVIDIA conspiracy? Nah.
As far as this topic goes, we all know NVIDIA typically refreshes their cards, anyone could make that kind of educated guess. Whether they go to TSMC for a 7nm refresh is up in the air, they could do that or maybe they start getting better bins with Samsung and can release higher clocked + memory Super cards on 8nm. But even this rumor tries to paint Ampere as some grand failure and that NVIDIA is shitting in it's pants to rush to make 7nm Super cards.
P.S. One thing I can say with certainty is MLID is a huge AMD shill. The guy was making excuses like crazy for the $50 price hike and then told his viewers to buy Zen 2 CPUs if they can't afford Zen 3's $50 price hike, yeah ok, like that's a really good alternative despite the competing Intel CPU being cheaper. His reason for Intel being such shit? Because they don't have PCI-e 4.0, like who gives a shit about that? Yet he never holds back at making up conspiracies about NVIDIA and how it's out to fuck everyone. Leakers like him have no credibility in my eyes, I used to somewhat take him seriously until his latest videos and now I tuned him and all the other fake leakers out.
That's true but when the main thing overlapping with Nvidia and Samsung is the GPU, it makes sense to assume the actual silicon yields are low. We've heard between 20,000 and 30,000 cards at launch, for the 3080 and 3090 combined, with 300,000 in the next six weeks, combined, across all SKUs. If not yield, then what else could hold up two waves of product?What's factual is the Ampere cards have severe shortages and nobody outside Samsung and NVIDIA know why.
Not a grand failure, just a trainwreck of a start. Getting on a denser node will help out tremendously. The dies are just too big to make without flaws. They're also being pushed outside of their power envelope. I also wonder if they're going to market the 2x RAM cards as the Supers. Nvidia's said it a bunch, they want to streamline their brand portfolio.But even this rumor tries to paint Ampere as some grand failure and that NVIDIA is shitting in it's pants to rush to make 7nm Super cards.
They also have a disgustingly large order for the A6000's from Dell and HP, those use the better version of the chips used by the 3080 and 3090 and that needs to be filled first, they have also received massive orders for their DGX A100 systems. NVidia is processing a massive amount of very large high-profit orders first as most of those have delivery deadlines. OEM orders will always come first consumer market gets the short end of the stick.Why wouldn't I comment? I've noticed a concerted effort lately to smear NVIDIA for part shortages as if they're the only company on the planet that's experiencing it and then you get a range of ridiculous conspiracy theories that go with it--like them withholding hundreds of thousands of cards to create artificial demand to barely being able to meet their BOM and making no profit on FE cards. But when you probe even a little bit, none of these tinfoil hat theories have any substance to them, they're basically propagated by a few people (e.g. MLID) and treated as fact.
This isn't just limited to NVIDIA only, we also have guys like Not An Apple Fan (who MLID likes to give credit to) hyping up AMD fans with nonesense like, "yeah lets get on the choo choo train, oh btw don't forget to donate to me" and telling everyone he's got inside knowledge that AMD has a faster card they're hiding because they want to "jebait" NVIDIA.
What's factual is the Ampere cards have severe shortages and nobody outside Samsung and NVIDIA know why. Jensen has said it's due to unprecedented demand and he's probably not far off, people have gone nuts in trying to get 3080s, especially since they got Covid money to spend and a lot of people who skipped Turing have the upgrade itch. Then another fact is the pandemic and the disruption in the supply line, brand new memory type and a new foundry they've never done business with. All those would add up to shortages, but that doesn't mean it's some grand conspiracy NVIDIA has cooked up to really screw it's customers--that's just ridiculous.
I'll put money on Zen 3 selling out instantly and people not being able to find 5800/5900s in stock anywhere for weeks. The same may or may not happen to Navi 21 because traditionally AMD has far less demand than NVIDIA and they might be able to whether shortages better but even they'll sell out pretty fast and not have new inventory for awhile. If you look at past graphics card releases, they always instantly sold out and were unavailable for long periods of time but again, Ampere is unique because of the pandemic, new memory type and new foundry so it compounded things and just resulted in a terrible launch. But some nefarious NVIDIA conspiracy? Nah.
As far as this topic goes, we all know NVIDIA typically refreshes their cards, anyone could make that kind of educated guess. Whether they go to TSMC for a 7nm refresh is up in the air, they could do that or maybe they start getting better bins with Samsung and can release higher clocked + memory Super cards on 8nm. But even this rumor tries to paint Ampere as some grand failure and that NVIDIA is shitting in it's pants to rush to make 7nm Super cards.
P.S. One thing I can say with certainty is MLID is a huge AMD shill. The guy was making excuses like crazy for the $50 price hike and then told his viewers to buy Zen 2 CPUs if they can't afford Zen 3's $50 price hike, yeah ok, like that's a really good alternative despite the competing Intel CPU being cheaper. His reason for Intel being such shit? Because they don't have PCI-e 4.0, like who gives a shit about that? Yet he never holds back at making up conspiracies about NVIDIA and how it's out to fuck everyone. Leakers like him have no credibility in my eyes, I used to somewhat take him seriously until his latest videos and now I tuned him and all the other fake leakers out.
This is why I believe gaming is really becoming a after though for Nvidia. Their focus is AI and large data centers. Also probably rather push Geforce now. I believe in the next ten years gaming as we know it will disappeared.They also have a disgustingly large order for the A6000's from Dell and HP, those use the better version of the chips used by the 3080 and 3090 and that needs to be filled first, they have also received massive orders for their DGX A100 systems. NVidia is processing a massive amount of very large high-profit orders first as most of those have delivery deadlines. OEM orders will always come first consumer market gets the short end of the stick.
That sounds like they should have planed better and launched these cards later so they can satisfy the oem as well as the large consumer demandThey also have a disgustingly large order for the A6000's from Dell and HP, those use the better version of the chips used by the 3080 and 3090 and that needs to be filled first, they have also received massive orders for their DGX A100 systems. NVidia is processing a massive amount of very large high-profit orders first as most of those have delivery deadlines. OEM orders will always come first consumer market gets the short end of the stick.
And just sit on inventory for 6-8 months, all while having 0 products to sell? The 2000 series stopped moving the second rumor started that a new product was coming soon, they stopped production of said 2000 back in June so that was 3-4 months of 3000 series production before the launch not sure waiting until Oct or Dec would have made any difference other than to annoy people even more than they already are.That sounds like they should have planed better and launched these cards later so they can satisfy the oem as well as the large consumer demand
That sounds like they should have planed better and launched these cards later so they can satisfy the oem as well as the large consumer demand
They need to launch in Q3 as well because most corporations and educational facilities place their largest orders in their Q4 to burn up their budgets and stretch any very large purchases over into their next Q1.this shows you (and no i am NOT picking on you but you brought it up) dont know how big businesses operate nowadays with respect to their share holders.
the release date of sept 17 was at the end of 3Q.
this was done entirely to get it on the books. period. no other discussion needed.
doesnt matter it they ship 1 part or a quadzillion...its on the books...all the execs meet their bonuses
stock goes up...yay team!
there are countless companies across every industry that does this and has been doing this for 30ish years now
more so in the last 15.
it is THE reason that paper launches occur.
ever notice why products come out at the start or end of a quarter and not in the middle? it is all about the accounting and bonuses
this shows you (and no i am NOT picking on you but you brought it up) dont know how big businesses operate nowadays with respect to their share holders.
the release date of sept 17 was at the end of 3Q.
this was done entirely to get it on the books. period. no other discussion needed.
doesnt matter it they ship 1 part or a quadzillion...its on the books...all the execs meet their bonuses
stock goes up...yay team!
there are countless companies across every industry that does this and has been doing this for 30ish years now
more so in the last 15.
it is THE reason that paper launches occur.
ever notice why products come out at the start or end of a quarter and not in the middle? it is all about the accounting and bonuses
that sounds like they should exclusively launch the quadros ahead of time and sit on consumer cards till they can actually launch them.They need to launch in Q3 as well because most corporations and educational facilities place their largest orders in their Q4 to burn up their budgets and stretch any very large purchases over into their next Q1.
Quadro line is dead, NVidia confirmed that in their release last week. So make all their GPU's OEM exclusive till April or so, I wouldn't complain as that is where I generally get my stuff from but it would make a lot of others a lot angrier, having to wait is a lot more palatable than being flat out told No.that sounds like they should exclusively launch the quadros ahead of time and sit on consumer cards till they can actually launch them.