Graphics specialist NVIDIA (NVDA 3.14%) has seen tremendous growth in its gaming-oriented graphics processor business over the last couple of years. That growth is believed to be driven by two major factors.

The first is that demand for the company's graphics processors is simply up -- more people are buying these gaming-oriented graphics processors today than they had in the past. Second, gamers appear to be interested in buying up the company's product stack, which is boosting average selling prices.

NVIDIA's GeForce GTX 1080 Ti graphics card.

Image source: NVIDIA.

Unit growth combined with average selling price is leading to significant revenue growth.

Although it's unlikely that NVIDIA can continue to drive a richer mix of products indefinitely, there's a clear path to average selling price growth here following the launch of the company's GeForce GTX 1080 Ti graphics card.

Allow me to explain.

The old product stack versus the new

Prior to the GeForce GTX 1080 Ti announcement, the following was NVIDIA's desktop graphics processor product stack (MSRP figures used):

GPU

MSRP

GeForce GTX 1080

$599

GeForce GTX 1070

$379

GeForce GTX 1060 6GB

$249

GeForce GTX 1060 3GB

$199

GeForce GTX 1050 Ti

$139

GeForce GTX 1050

$109

Data source: NVIDIA announcements.

Here's the company's latest product stack following the 1080 Ti announcement:

GPU

MSRP

GeForce GTX 1080 Ti

$699

GeForce GTX 1080 w/ 11Gb/s memory

>$499 (?)

GeForce GTX 1080

$499

GeForce GTX 1070

$379

GeForce GTX 1060 6GB w/ 9Gb/s memory

>$249(?)

GeForce GTX 1060 6GB

$249

GeForce GTX 1060 3GB

$199

GeForce GTX 1050 Ti

$139

GeForce GTX 1050

$109

Data source: NVIDIA announcements, author.

There are several things worth observing here. First, NVIDIA now has a pricier product at the very top of its product stack. Since NVIDIA is likely to sell such products to the same sorts of customers who would have otherwise purchased a GeForce GTX 1080 (before the 1080 Ti launch and subsequent price cuts on the 1080), the introduction of the 1080 Ti into the product stack should -- all else equal -- help boost average selling prices.

However, while it's easy to focus on the new 1080 Ti, NVIDIA has also taken actions elsewhere in the product stack that could help boost average selling prices.

For one thing, NVIDIA cut the prices of the 1080 from $599 to $499. It didn't have a product at this price point before -- it was either the $599 MSRP GeForce GTX 1080 or the $379 MSRP GeForce GTX 1070 (as an aside, the various add-in-board vendors offered boards spanning price points above those MSRPs).

Now, in addition to having products that extend the top end of the range, the 1080 now provides gamers with an interesting option at the $499 price point.

Though there may inevitably be those that would have purchased a GTX 1080 at $599 who are now going to buy the 1080 at $499 (potential mix down for NVIDIA), there may also be those who weren't quite able to stretch their budgets to $599 for the 1080 before and would have just settled for the 1070 at $379 who can now justify spending $120 more for the 1080 -- driving a mix shift up.

NVIDIA is also doing something interesting to try to capture the price point in between the $699 GeForce GTX 1080 Ti and the $499 GeForce GTX 1080 – allowing add-in-board vendors to build GeForce GTX 1080 boards using faster 11Gb/s GDDR5X memory (the standard GeForce GTX 1080 only has 10Gb/s GDDR5X memory).

And finally, NVIDIA seems to be allowing its board vendors to offer a variant of the GeForce GTX 1060 6GB model with 9Gb/s GDDR5 memory (up from 8Gb/s on the current 6GB versions of the 1060). This should allow NVIDIA and its board partners to drive upsell among customers willing to spend more than what a GTX 1060 costs today but can't quite stretch their budgets to the GTX 1070 price point.

The potential business impact

NVIDIA is both increasing the top end of the price range for its GeForce GTX gaming-oriented graphics processors while also taking actions to fill in large gaps between consecutive cards in the product stack.

Time will tell how this ultimately works out for the company, but I think the moves that NVIDIA made here are sensible and are likely to achieve their desired effects: boosting blended average selling prices while at the same time serving to increase unit demand from where it would have been without these actions.