KitGuru KitGuru.net – Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards
Jen-Hsun Huang, chief executive officer of Nvidia Corp. on Tuesday wrote an open letter to the company’s customers covering the recent GeForce GTX 970 controversy. The head of the GPU developer admitted that the company failed to communicate how memory sub-system works on the GeForce GTX 970, but said that the graphics card should still be considered as a “4GB” adapter. He promised not to make the same mistake again.
“Some of you are disappointed that we didn’t clearly describe the segmented memory of GeForce GTX 970 when we launched it,” wrote Mr. Huang. “We invented a new memory architecture in Maxwell. This new capability was created so that reduced-configurations of Maxwell can have a larger framebuffer. […] GTX 970 is a 4GB card. However, the upper 512MB of the additional 1GB is segmented and has reduced bandwidth.”
Although Mr. Huang admitted that the GeForce GTX 970’s memory sub-system works rather oddly, he did not reveal why the company incorrectly stated the amount of raster operations pipelines (64 instead of 56) and level-two cache size (2048KB instead of 1792KB).
“Unfortunately, we failed to communicate this internally to our marketing team, and externally to reviewers at launch,” Mr. Huang acknowledged.
This is the not the first time Nvidia admits is mistake and blames poor communication within the company for the incorrect advertising. However, Nvidia offers no compensation to customers, who bought the GeForce GTX 970 based on specifications of the product.
Last week an end-user sued Nvidia for incorrect advertising and demanded a refund from the company.
Discuss on our Facebook page, HERE.
KitGuru Says: While it is good to see that Nvidia’s chief executive officer takes the GeForce GTX 970 issue very seriously, he did not reveal anything new and did not propose anything to owners of the graphics cards. While it is obvious that the GeForce GTX 970 is a great performer for the money, it is clear that many people are disappointed with the fact that they did not get what they paid for (and some even report about issues like micro-stuttering, which is a clear driver problem) and providing them some kind of a compensation would be a good step. It is unknown why Nvidia does not want to give anything back to its customers in this case.
The full text of Jen-Hsun Huang’s letter reads as follows:
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
So “There is no bug it’s a feature” 🙁 new feature from nV is Lie (or maby not new but very old, cuz they do the same with nV 660Ti no 2GB GDDR5 there, just 1.2 + DDR3 lol
http://youtu.be/spZJrsssPA0
ok dude its funny but I know spanish and he is not saying that lol. I think they got an excuse on the memory since its indeed an architecture feature but they cannot get away with advertasing the wrong ROPs and cache. They got caught and should issue at least some free games or a small refund plus a price reduction of at least 10%.
A better job would be stopping on gimping down the GPU and use more hardware to improve Thread Level Parallelism instead of dumbing down the GPU so much that it requires clever compiler optimizations, taxing the CPU to make the GPU run good, specially with nVidia’s bad tracking of stopping optimizing for GPU’s after a new generation is launched (GTX 780 anyone?). Maxwell is the least CPU like design of this whole generation of DX11 cards.
That will be nice 🙂 But ain’t gonna happen. Some call it not nvidia but nGreedia lol
Im waiting for R390X now, and im very happy with mine R280X XFX DD Black OC GHz Ed.
Im plaing games mostly: 1792:1344 or 1920:1440 with 85Hz on CRT Sony Black Trinitron ~23″ visible size. And its really great Hardware + 3 Games for free on AMD as Standard in AMD Game Evolving -> http://sites.amd.com/us/promo/never-settle/Pages/never-settle.aspx
I have nV 285GTX 1.2GB 512Bit -> when i’m changing GPU its in my PC. now only Radeon GPU’s is in my mind.
Rest is no comment…
What about “the amount of raster operations pipelines (64 instead of 56) and level-two cache size (2048KB instead of 1792KB)??!!!! Why didn’t he explained this?!
What about “the amount of raster operations pipelines (64 instead of 56) and level-two cache size (2048KB instead of 1792KB).”??!!!! Why didn’t he explained this??!!!
it was a gimped 980 if it wasn’t it wouldn’t be a 970….
And the solution is…
NONE…
Tks for the open letter, next time ill sure have it in mind when I buy and AMD card!
But neither an advertised 64 ROP/2MB card that resulted was a blatant lie and misleading.
Well Mr Huang is certainly full of shit,just look how he tries to deflect with the “maxwell is a revolutionary architecture be grateful you got 4GB not 3GB !” Sure the Nvidia apologists will defend him and his BS and guys like Ryan Shrout at PcPer will continue to make excuses for them so they continue to recieve review samples of future cards so they have content for their “articles”.
The 1 torpedo that sinks the USS Bullshit is the fact that they not only gave out misleading and inaccurate specs but sat on it for months and would have likely stayed silent about it had users not discovered the truth.It appears to be the “fashionable trend” to lie and mislead your consumers and then blame it on a “communication error” when they get caught red handed.
Ooops Our Baaaad, Marketing didnt get the memo and you got screwed, So sowwy. PFFT
I am so glad I ditched nvidia for high end cards years ago. They are a bit flakey on customer service. The fact that he ducked several other issues and made excuses shows they are not at all sympathetic. The whole, lets build a $200 GPU chip, then cripple it down to $190, $180, $160 versions it just pure greed. IF they built the $200 gpu and only that chip and reference design they could make a lot more money. I mean its a bit like building a Bugatti and If you pay full price you get full performance pay 10% less we will just disable 4 cylinders. The problem is though is us users are so desperate to have the most they can afford this very minute they buy into this stupid culture. Thank god when you buy a discounted pet…..
When he said “We’ll do a better job next time” all I could think of was this http://33.media.tumblr.com/9c8b68a088542fc52fe7cb8d000c7373/tumblr_mt14la3Neq1r1vf9eo1_400.gif
It’s true – I mean, the tech guys *must* have noticed the errors in the press statements and the advertising, and they will have notified the marketing guys. NV chose to keep their mouth shut on this one, so whether it was a mistake in the beginning by the time we found out they had plenty of time to rectify it so it was intentional misleading.
Ppl are just blind and dont get it… If you bought a i7 convinced that it would have 8 cores and you ended up with 4, how would you feel?
If it only had 3.5gb fine! But should be advertised like that… If I knew I probably would have bought a 290x… Just for that bit of future proOf.