Nvidia RTX cards suggest an interesting sea change in workstation strategy

Nvidia’s cards focus on rending and a more specific niche for workstation users.

abstract virtual space

Workstations represent what I think is the strongest personal technology implementation in the PC space, albeit at the high end. What I’m talking about is the almost laser-like focus workstation vendors have on workstation users. This is largely something we lost in the general corporate PC space that I think was a good part of the cause in the general PC space, the vendors focused excessively on IT (which focused them on price, cookie cutter boring designs, and excessively long model runs) rather than user needs. In the workstation space no one seemed to lose track of the people that used the products but, as focused as these vendors were workstations started to look pretty similar even though the people that used them have vastly different jobs.

With this week’s launch of the Nvidia RTX (disclosure: Nvidia is a client of the author) line of rendering-focused cards it hit me that this is changing big time because these cards aren’t generically targeting every workstation user. They are focused like a laser just on those that do rendering, and this really isn’t the first Nvidia has done this.

I think this even tighter focus on users is not only a competitive advantage for Nvidia, it is a competitive advantage for workstation users who qualify for the related parts, and it is the way we should start to consider building the next wave of personal technology.


I really should have seen this when Nvidia brought their DGX workstation to market. The problem they were solving was that there really was no workstation designed to support people that were working in the artificial intelligence field. Yes, you could use generic GPUs, which formed much of the basis for current AI development, but like a Swiss Army Knife it was never going to be as good as something that was purpose built for the segment.

Nvidia created a wicked expensive workstation that had massive performance improvements but focused on this AI segment and used it heavily internally. This appears to have advanced their lead in the AI segment significantly and, for those on the cutting edge of AI development, short of having their own Watson implementation (which also uses Nvidia technology) the DGX is the most powerful AI development tool in market.

RTX, on the other hand, is massively focused on rendering, and by focusing they were able to bring a concept developed by J. Turner Whitted in 1979 to the desktop. This concept called “Multi-Bounce Recursive Ray Tracing” was, at the time, impossibly compute-intensive, requiring what was estimated to be a super computer for every pixel rendered. But this was what was required to properly deal with the biggest problem, according to Nvidia’s CEO Jen-Hsun “Jensen” Huang, light. Apparently light is really nasty to deal with. This is because of the way it bounces around a scene changing who textures and surfaces appear, created reflections and shadows, and generally making the creation of artificial objects that look real very, very difficult.

But by creating a platform that could aggressively deal with light they were able to focus on the platform and the RTX line of video cards is uniquely focused as a result. One other interesting thing about these cards is the way the get to their impressive rendering performance level (about 6 times faster than prior technologies according to Nvidia). They do this by first rendering in low resolution then using a unique AI technology branded NGX using deep learning to up-convert to high resolution. An incredibly creative approach to this problem.

However, the point is that both DGX and RTX product lines are focused not on general workstation users but on those that work on AI or rendering respectively. Though not quite the same thing, because the use model is very different, I’ve also noticed that both AMD (disclosure: AMD is a client of the author) and Nvidia have been working with their licensees to create cryptocurrency focused graphics cards as well.

I believe this focus on the kind of work the user does will not only significantly benefit related productivity, it will advance significantly advancements in the fields where these workstations are used. There are some indications of similar tends in the general PC space. For instance, the Always Connected PC effort driven by Qualcomm and Microsoft (disclosure: both Qualcomm and Microsoft are clients of the author) appear focused on those that are ultra-mobile as is the new Surface Go, which broke 12” minimum size barrier.

Wrapping up

In the book “Technically Wrong” by Sara Wachter-Boettcher the author goes into great detail on how technology companies lost focus on those that buy and use their products. Workstation offering have always been better focused on users who generally ignored IT to demand what they needed and generally got it. However, with the recent DGX and even more recent RTX announcements we can see that even more focus is needed and Nvidia, at least, is stepping up. I think this is a powerful trend and one that seems to be drifting to the PC space all of which should drive more interest and satisfaction in the segment. This should also significantly advance technology in the targeted segments. Overall, this should be a really good thing for the health of the PC – Workstation segment and a lesson we shouldn’t again forget. Workers are generally better when the tools conform to their needs and not the other way around, and it fascinates me how often we seem to forget this.

Copyright © 2018 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon