Update: Nvidia has provided an official response to the issue. It is lengthy, and is as follows:
The GeForce GTX 970 is equipped with 4GB of dedicated graphics memory. However the 970 has a different configuration of SMs than the 980, and fewer crossbar resources to the memory system. To optimally manage memory traffic in this configuration, we segment graphics memory into a 3.5GB section and a 0.5GB section. The GPU has higher priority access to the 3.5GB section. When a game needs less than 3.5GB of video memory per draw command then it will only access the first partition, and 3rd party applications that measure memory usage will report 3.5GB of memory in use on GTX 970, but may report more for GTX 980 if there is more memory used by other commands. When a game requires more than 3.5GB of memory then we use both segments.
We understand there have been some questions about how the GTX 970 will perform when it accesses the 0.5GB memory segment. The best way to test that is to look at game performance. Compare a GTX 980 to a 970 on a game that uses less than 3.5GB. Then turn up the settings so the game needs more than 3.5GB and compare 980 and 970 performance again.
Here’s an example of some performance data:
On GTX 980, Shadows of Mordor drops about 24% on GTX 980 and 25% on GTX 970, a 1% difference. On Battlefield 4, the drop is 47% on GTX 980 and 50% on GTX 970, a 3% difference. On CoD: AW, the drop is 41% on GTX 980 and 44% on GTX 970, a 3% difference. As you can see, there is very little change in the performance of the GTX 970 relative to GTX 980 on these games when it is using the 0.5GB segment.
Original Story: In the current PC gaming climate, the GTX 970 is known as the king of the jungle. At roughly $350, it outperforms everything else in its price range thanks to its Maxwell architecture. Its performance on benchmarks has earned it a valuable reputation that has led to widespread critical acclaim, and skyrocketing sales.
But there may be a blemish on the GTX 970 that could challenge its otherwise stellar reputation.
After running GPU memory benchmarks, a group of users at OverClock.net have come to the conclusion that the last 700MB or so of the GTX 970’s 4GB of VRAM has significantly reduced throughput. As a matter of fact, it could be more than a third less potent as the other 3.3GB of memory.
Surprisingly enough, the GTX 980 doesn’t fall victim to this issue. So the question is: is Nvidia marketing the GTX 970 as having 4GB of VRAM when it doesn’t?
Below are the benchmark results for both cards:
Prior to today there have been many reports from GTX 970 users stating that the card suffers from stuttering in certain environments. These environments just so happened to be high VRAM utilization oriented.
The issue has been significant enough that discussion has resulted in a 22+ page thread on OverClock.net with over 200 posts. There’s a similar thread on the official Nvidia forums with 10 pages and growing.
ManuelG, a moderator on the official Nvidia forum, has responded. He commented:
We are still looking into this and will have an update as soon as possible.
Reports indicate that this is a hardware issue. If that’s the case, then there will be no fix. So, GTX 970 owners are crossing their fingers hoping for the best.
VRAM has increased in demand significantly over the past couple years, especially with the introduction of affordable 4K displays. Also, the abundance of unified memory on the Xbox One and PS4 has led to unoptimized PC ports that require far more VRAM than they would if optimized.
Also See: Microsoft Unveils HoloLens, the “Most Advanced Holographic Computer the World Has Ever Seen”
As an owner of the GTX 970, I’m greatly concerned. I’ve been more than happy with the performance of the $350 card during the past month of ownership, but part of the reason I invested in the card was because I felt confident in its relatively future proof specs, part of which is its 4GB of VRAM. If it has less, the card may not last me as long as I hope.
We’ll update this article when a new statement is made by Nvidia.
Thank you OverClock user mrkk for the benchmark photo!
