Hmm, you're right, it's not entirely Nvidia's fault.
Maybe Nvidia should've enforced their own specs more strongly, as the quality of the FEs is good, but as I understand it, the cost of the FE design was a slap to partner profits - that's why none of them used it.
Well, not the first time the green guys have screwed over their partners...
The problem goes back to Nvidia's decision to start producing their own cards, which in many ways mirrors the 3Dfx decision in times when they were the top dog with its Voodoo cards.
Now Nvidia has learned for mistakes that eventually led to 3Dfx's downfall, but the position it has put its board partners in is very much the same, leaving companies like EVGA, ASUS, Gigabyte and MSI in the position where at launch of any product line, NVIDIA is their key supplier and direct competitor at the same time. And even after FE is done leaves them with ridiculously small margins to work with and precious little room to manevour. It is worth nothing that when NVIDIA produced their first cards, there was a huge backlash from board partners, because the 3Dfx situation was still quite fresh in their their minds, and more importantly in their books. All of the sudden NVIDIA started saying how limited numbers these cards were, and they had no intention of becomming a full blown card producer. For a while after that, there were still gruntled voices in the industry on how "limited" those runs were, and thats about the time when the "Founders Edition" moniker was born.
Now fast forward to the present day, we have an ecosystem, when NVIDIA skimms the large share of the cream from the top of each launch. Additionally (at least before covid) with its FE sets the pricing level board partners must fall in line behind ( at least for the volume SKUs) and entirely controls the pricing of key components in board partners product. Now add the relatively short product cycle of each generation, factor in how long and how much money it takes to produce boards based on those those new GPUs and you have an enviroment where lion's share of risk involved is with board partners who work with as little as 3% margin on the volume SKUs. And finally take a moment to consider how many new NVIDIA board partners we've seen to enter the market in the last 15 years compared to how many we've seen to fade away. The bil. dollar question is, how much room is there for quality when you work in environment that tight ( and no, those cards are not leaving the factory anywhere near the prices you as customer paid for them, not even before covid).
And there's the question how to differentiate to appeal to your customer, you bump up the clocks, touch up power delivery ( only if you really really have to) and make your own cooler.
What i believe NVIDIA needs to do is to finally decide if it wants to become a full fledged consumer board manufactuer or pure chip designer and supplier.
Facebook
Twitter
Instagram