NVIDIA Continues to Blame GPU Shortages on Demand Rather than Supply

In a world where board partners are getting less than 20 units per country (per month), NVIDIA continues to insist that the shortages are primarily the result of strong demand. While technically that’s not incorrect, when retailers are getting less than 100 graphics cards in the entire Asia-Pacific region, it makes it hard not to blame the supply. Talking at the Raymond James Institutional Investors Broker Conference Call, the NVIDIA CFO said the following when asked whether demand would have been better if there was more supply:

Sure. So when we really think about our business, I think you really have to start the statement, not on the word supply, but really about the word, demand. Our demand is strong. Our demand has been strong across our businesses as we have been launching both the Ampere architecture to gaming and data center. We are now building, for example, for our overall gaming business, we are building GPUs for gaming both at Samsung and TSMC.

Colette Kress, NVIDIA CFO

Basically what this means is that regardless of how lean the inventory may be, NVIDIA can still make vast amounts of profits due to the present state of the graphics card market:

Q. Right. So it sounds like you don’t feel that the constraints in supply — I mean, you still may be constrained, but it doesn’t necessarily constrain your ability to grow from these levels. There’s additional supply coming on. Demand, is there. We can still get some growth.

A. Yes, absolutely, Chris. So this is, of course, a situation that we would love to have more optionality. We would love to have more overall supply. But remember, our issue is, is demand is strong. Demand is strong, and we do believe we will grow quite well as this demand will continue through this year in front of us.

And then, of course, there was the inevitable question on cryptomining and how it has affected the market. Once again, Kress brought up the introduction of the CMP boards for mining and that anti-mining measures introduced for the RTX 3060. We already know that the CMP boards are essentially a measure to prevent another crypto-hangover and make sure the revenue doesn’t suffer in the future, so that’s a moot point:

As for the anti-miner measures implemented on the GeForce RTX 3060, word on the street is that the restriction has already been overcome. Even those who can’t get over it can earn plenty by mining secondary currencies such as Octopus or Cortex.

Keep in mind that we are de-hashing, reducing the overall mining performance on our latest 60 series GPU and announced a CMP board, specifically for miners. We announced that CMP board as we felt if we saw demand from our professional miners, it would be a good opportunity for us to restart that CMP program for them. Now thirdly, Etherium mining increases that have occurred is shared with a lot of different overall drivers. Remember, we have an established installed base Ethereum this time around versus what we had in 2018. You’re also seeing new ASICs that are part of a lot of the hashing that is occurring, and there are multiple GPU providers at this time.

Colette Kress, NVIDIA CFO

The last bit was regarding the Average Selling Price (ASP) of the RTX 30 series GPUs. It looks like the early introduction of the RTX 3070 along with poor 3060 Ti supply (despite being the better offering out of the two), NVIDIA’s ASP for this cycle has increased significantly:

So we have volume SKUS, such as 3060, but given our interest in many of the gamers to be even above our volume SKUs and moving into that 3070, 3080, both helped us in terms of our average ASPs, our total revenue as a whole. And we will likely continue to see those be a driver for us.

Colette Kress, NVIDIA CFO

Source: SA


Computer hardware enthusiast, PC gamer, and almost an engineer. Former co-founder of Techquila (2017-2019), a fairly successful tech outlet. Been working on Hardware Times since 2019, an outlet dedicated to computer hardware and its applications.
Back to top button