RTX 50 Series launch was «frustrating and stressful»
The day the RTX 50 Series went on sale was anything but satisfactory. And that goes for us as well as for you. Category Manager Luca Franziscus gives us a peek behind the scenes of the recent launch.
On 30 January, the RTX 5080 and RTX 5090 went on sale at 3 p.m. sharp. Or did they? If the numerous Community comments posted under our article on the RTX 5080 are anything to go by, the launch may not have gone completely as planned. One user says:
3 p.m. has come and gone. Where are the cards?
As a matter of fact, the cards were here. But only a few, and not all of them at 3 p.m. Luca Franziscus from our Category Management team explains why this was the case and shares some other insights on the launch.
What was the RTX 50 Series launch like for you?
It was frustrating and stressful. My main goal as a Category Manager is to make sure our customers are satisfied. Unfortunately, I knew even before the launch that there wouldn’t be a satisfactory solution for everyone involved, or that everyone interested in buying a card could get one.
Why?
We feel the same way as the users in the comment section: everything always comes to us at the last second. On top of the frustration this causes, it always creates a huge workload for us shortly before launch – even though we thought we’d made all the necessary preparations in advance this time.
Was that why certain graphics cards weren’t available on our website at 3 p.m.?
In principle, all the cards went online at 3 p.m. sharp. Our online shop’s product cache occasionally takes a few minutes to update, so some models might not have been listed at 3 p.m. The other thing worth mentioning is that we only had a very limited quantity of cards available. We’re talking a few hundred units, spread out across dozens of models. If you think about it, ten units of a specific model sell within a few seconds.
Why didn’t you put the graphics cards online earlier?
We’re simply not allowed to do that. We’re contractually obliged to stick to that launch time. Otherwise, confidential data could be made public in advance, as is unfortunately often the case when leaks occur on online shops.
Could you give us an example?
Apparently on one online shop in Germany, cards were listed in the background pre-launch. Although they were untraceable to customers, they were visible to bots, which then tapped into the data using crawlers and bought up the entire stock at 3 p.m. sharp. With this in mind, it’s in our customers’ interests if we really only list the cards at that time point.
By the way, we have a zero-tolerance policy on bots, and follow up on suspicious orders. We also cancel orders for more than one card per person. Plus, we take a hard line when it comes to our employees placing orders. They don’t get to benefit from pre-orders, nor do they receive any discounts.
When did you find out how many cards you’d be receiving at the warehouse?
We were told what our allocation would be two weeks before the launch. At the time, we were assured the cards would be with us in time for the release. We then entered them into our system one by one so that we were ready for launch day. Naturally, we did our best to stay updated on when the cards would arrive, but unfortunately the delivery date was continually pushed back.
How come?
Basically, Nvidia wanted to wait until the last possible minute for the cards to enter the supply chain. If a card’s left in a manufacturer’s or distributor’s warehouse for too long, there’s potential for it to be leaked. Then again, Chinese New Year also affected this launch. When we asked the distributor and manufacturer about it, it turned out that certain RTX 5090 models were left in China for longer. But deliveries are often delayed as it is.
The first cards really did arrive at our warehouse on launch day itself. Unfortunately, none of them were RTX 5090 models. Though we expected to receive our shipment in time for the deadline, we were told on the 30th that it’d take a few more days.
Why couldn’t you communicate that to customers early on?
Partly because we ourselves just didn’t know. Our warehouse then had to book in the cards so that we could keep track of them. The thing is, we don’t have the infrastructure for that. We can create blog pages, but they’re difficult to maintain and our automated system means we have little influence over how they’re displayed.
Speaking of automated systems, there was an incident on the weekend after the launch. What happened there?
An offer for 500 GPUs came in from one of our suppliers. Our system counted this as stock and put the cards online. Unfortunately, our supplier was unable to deliver the products and we had to cancel the orders. Needless to say, this was exasperating for the customers affected, and I’d like to apologise to them.
Let’s go back to the launch itself. Shortly after 3 p.m., there were no more cards left on our site. Why?
By then, we’d already sold all the cards we had in stock. As a result, the system took them offline to prevent overselling. That was an issue we had when the RTX 30 Series launched, which led to really long delivery times. We also take this approach because we base our products’ sales prices on their current purchase prices. Scarce availability is bound to affect those numbers. In fact, we can assume that prices will rise in the coming months. We’ve got to do ourselves a favour and make sure we don’t add to that.
There was a huge rush on the cards despite their performance increase being pretty mediocre. How do you explain that?
One reason behind the flood of orders is bound to be the fact that RTX 40 Series models are hard to come by. Nvidia discontinued the GPUs in the middle of last year. As a result, top-of-the-line models in particular were in short supply and prices have remained high. A lot of people held back from buying them and opted to wait for the RTX 50 series instead. At the same time, the new generation isn’t much more expensive than the previous one, despite the moderate performance increase. So, waiting seems to have paid off.
The availability of the new generation is also very limited. Why doesn’t Nvidia just make more GPUs?
It’s anyone’s guess. As far as Nvidia’s concerned, it might just make more profit from AI accelerators than consumer graphics cards. Given TSMC’s limited production capacities, it prefers having chips (the more profitable product) manufactured. That leaves less room for consumer products. Nvidia itself announced on launch day that it wouldn’t be able to meet demand.
Customers will now be able to see the cards on our site again, but they can’t be ordered from us directly. So, what’s going to happen now?
We should receive the last graphics cards from our original allocation in the next few days. We’ll be publishing a new blog page to announce which models we have in stock and when they’ll go on sale. We did something similar when there was a chip shortage during the pandemic. It’s something we simply weren’t able to do on launch day, as we didn’t have the necessary information.
What should people interested in buying the cards do now?
Take a look at our website occasionally and keep your eyes peeled for our announcement. We’ll let you know when the cards are due to go on sale well in advance so that you’ll be prepared for it.
When will the situation in general start to calm down?
That’s difficult to judge, but things are likely to continue this way for a few more months until the cards are widely available. There’ll probably be bottlenecks with the upcoming RTX 5070 Series when it’s first released too. At least the situation with AMD cards seems to be much better.
Header image: Luca Franziscus
103 people like this article
Kevin Hofer
Senior Editor
kevin.hofer@digitecgalaxus.chFrom big data to big brother, Cyborgs to Sci-Fi. All aspects of technology and society fascinate me.
