We will do a better job next time
While there have been a lot of reports regarding memory allocation issues on the GTX 970 graphics card as well as recent class-action lawsuit for false advertising regarding that same memory issue, number of ROPs and L2 cache, there has not been any official response from Nvidia, at least until now when we have an official response from Nvidia's CEO, Jen-Hsun Huang.
Writing on the official Nvidia blog, Nvidia's CEO and founder, Jen-Hsun Huang has explained the new memory architecture in Maxwell where Nvidia has invented a new memory architecture with capability so that reduced-configurations of Maxwell can have a larger framebuffer, or in the case of the GTX 970, the graphics card is not limited to 3GB but can have an additional 1GB.
According to Jen-Hsun, the GTX 970 is a 4GB graphics card while the upper 512MB of its 4GB buffer is "segmented and has reduced bandwidth," which allowed Nvidia to add an additional 1GB and keep less frequently used data in the 512MB segment.
In that same blog post Jen-Hsun notes that while Maxwell memory architecture was a good thing but also confirms that the company failed to communicate this internally to the marketing team and reviewers. He also adds that while some were disappointed that the company did not better describe this segmented memory architecture, he says that their "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". Jen-Hsun Huang says that this 4GB of memory on GTX 970 is used and useful to achieve the performance and engineers will continue to enhance game performance.
At the end of the blog post, Nvidia's CEO wrote: "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."
While we certainly agree that the Geforce GTX 970 still represents good value, we tend not to agree with some things as while Maxwell's memory architecture and segmentation is "a new feature" as Jen-Hsun describes it, there are still plenty of problems when that last 0.5GB is filled, and announcing the wrong number of ROPs and L2 cache is definitely not a small deal.
Source: Blogs.Nvidia.com.
Writing on the official Nvidia blog, Nvidia's CEO and founder, Jen-Hsun Huang has explained the new memory architecture in Maxwell where Nvidia has invented a new memory architecture with capability so that reduced-configurations of Maxwell can have a larger framebuffer, or in the case of the GTX 970, the graphics card is not limited to 3GB but can have an additional 1GB.
According to Jen-Hsun, the GTX 970 is a 4GB graphics card while the upper 512MB of its 4GB buffer is "segmented and has reduced bandwidth," which allowed Nvidia to add an additional 1GB and keep less frequently used data in the 512MB segment.
In that same blog post Jen-Hsun notes that while Maxwell memory architecture was a good thing but also confirms that the company failed to communicate this internally to the marketing team and reviewers. He also adds that while some were disappointed that the company did not better describe this segmented memory architecture, he says that their "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". Jen-Hsun Huang says that this 4GB of memory on GTX 970 is used and useful to achieve the performance and engineers will continue to enhance game performance.
At the end of the blog post, Nvidia's CEO wrote: "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."
While we certainly agree that the Geforce GTX 970 still represents good value, we tend not to agree with some things as while Maxwell's memory architecture and segmentation is "a new feature" as Jen-Hsun describes it, there are still plenty of problems when that last 0.5GB is filled, and announcing the wrong number of ROPs and L2 cache is definitely not a small deal.
Source: Blogs.Nvidia.com.
News by Luca Rocchi and Marc Büchel - German Translation by Paul Görnhardt - Italian Translation by Francesco Daghini
Previous article - Next article
comments powered by Disqus
comments powered by Disqus