NVIDIA’s Jen-Hsun Huang Issues Apology For GTX 970 Inaccuracies

Nearly an entire month after the GeForce GTX 970 memory debacle kicked-off, NVIDIA CEO Jen-Hsun Huang has issued an open letter to help close the chapter on these unfortunate events, and also to offer an apology.

After reading, I’m left a little uncertain as to whether or not the letter is going to have the impact that NVIDIA wants it to, because no admission of an architectural mistake is offered. If you’ve read my previous posts, that shouldn’t come as much of a surprise, as NVIDIA has remained adamant that the slowed final 500MB of the framebuffer was a design choice, and not an accident. In his letter, Jen-Hsun backs that up.

Here’s his full letter:

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

Even with the crippled final 500MB in the framebuffer, Jen-Hsun believes that the GTX 970 is still a fantastic graphics card, and that’s something I’d also have to agree with. With the events of the past month, it’s unfortunately easy to overlook all that the GTX 970 brings to the table. I was personally left so impressed back at launch, that I had this to say about it in my review of ASUS’ Strix model:

“Last week, I was left very impressed with the GTX 980, but I was totally oblivious of just how impressive the GTX 970 would prove to be. I can safely say that I am even more impressed with the GTX 970 than the 980, because the value is through the roof.”

Has my opinion changed? Only slightly. As I mentioned before, if you’re rocking a 1440p display and are running only a single GTX 970, you’re in all likeliness never to see an issue. If you’re running multiple GTX 970s and a resolution of 4K, I’d definitely be a little concerned. As AMD’s 8GB Radeon R9 290X has proven, it’s not that hard to push beyond 4GB when sending over 8 million pixels’ worth of high-end detail to the screen.

Actually – one thing this entire debacle has really highlighted to me is that people don’t just want the 4GB they’re promised, they want more if it’s offered to them. With AMD offering an 8GB card (and is apparently able to compete against the GTX 980 nicely in multi-GPU configurations at high-resolution), I think the time is now that NVIDIA should offer its own high-density GTX 970 and 980.

That aside, the biggest lesson that needs to be learned from this relates to internal communication at NVIDIA. Whether this is a big issue, or a small one, consumers have proven that they aren’t willing to be misled.