NVIDIA CEO Jen-Hsun Huang Talks GeForce GTX 970 Memory Specifications

GeForce GTX 980 and GTX 970

NVIDIA took a beating when gamers found out that the GeForce GTX 970 video card has a different memory architecture than the GeForce GTX 980. NVIDIA was fairly quite on the situation, but did explain why the performance drop was happening and pointed out that it was an oversight between some folks internally as to why that information wasn’t fully disclosed. NVIDIA CEO Jen-Hsun Huang has posted up a statement in response to the criticisms of the community. Jen-Hsun stopped short of issuing an all out apology, but did say that NVIDIA won’t let this happen again.

gtx970 specs

 

The comments come shortly after NVIDIA was given a class action lawsuit for allegedly falsely advertising the NVIDIA GeForce GTX 970 graphics card. The suit claims that NVIDIA sold the based on the misleading representation that the GTX 970 operates with a full 4GB of VRAM at GDDR5 (not a less performant3.5 GB with a less performant and decoupled .5 GB spillover), 64 ROPs (as opposed to 56 ROPs), and an L2 cache capacity of 2048KB (as opposed to 1792 KB). 

It will be interesting to see what happens with the lawsuit. You can read Mr. Huangs comments below.

Jen-Hsun Huang Kicks Off GTC 2014

Hey everyone,

Some of you are disappointed that we didn’t clearly describe the segmented memory of GeForce GTX 970 when we launched it. I can see why, so let me address it.

We invented a new memory architecture in Maxwell. This new capability was created so that reduced-configurations of Maxwell can have a larger framebuffer – i.e., so that GTX 970 is not limited to 3GB, and can have an additional 1GB.

GTX 970 is a 4GB card. However, the upper 512MB of the additional 1GB is segmented and has reduced bandwidth. This is a good design because we were able to add an additional 1GB for GTX 970 and our software engineers can keep less frequently used data in the 512MB segment.

Unfortunately, we failed to communicate this internally to our marketing team, and externally to reviewers at launch.

Since then, Jonah Alben, our senior vice president of hardware engineering, provided a technical description of the design, which was captured well by several editors. Here’s one example from The Tech Report.

Instead of being excited that we invented a way to increase memory of the GTX 970 from 3GB to 4GB, some were disappointed that we didn’t better describe the segmented nature of the architecture for that last 1GB of memory.

This is understandable. But, let me be clear: Our only intention was to create the best GPU for you. We wanted GTX 970 to have 4GB of memory, as games are using more memory than ever.

The 4GB of memory on GTX 970 is used and useful to achieve the performance you are enjoying. And as ever, our engineers will continue to enhance game performance that you can regularly download using GeForce Experience.

This new feature of Maxwell should have been clearly detailed from the beginning.

We won’t let this happen again. We’ll do a better job next time.

Jen-Hsun