14 DAY TRIAL // With every single day that passes, we seem to hear more and more rumors regarding the possible delay of Nvidia's next-generation graphics core, known by the name of Kepler, and the latest reports to make their appearance suggest that the company's GPUs isn't performing as well as expected.
Until recently, most of the rumors concerning Kepler's delay stated that TSMC is having problems with its 28nm yields, which forced Nvidia to postpone the launch of the GPU to 2012.
DigiTimes' sources, however, don't entirely agree with this explanation and suggest that Nvidia is also having troubles with Kepler's performance as current silicon delivers lower than expected results. These can most probably be resolved by developing a new revision of the Kepler GPU.
Outside of the cited sources, DigiTimes also bases its report on the fact that other TSMC customers, that also rely on the foundry's 28nm node for building their chips, haven't announced any delays.
One such example is Qualcomm that still expects to release the MSM8960, MSM8270 and MSM8260A SoC solutions, based on the 28nm Krait architecture, in the fourth quarter of this year.
Kepler is the code name used by Nvidia to refer to its next-generation graphics core and, just like AMD's Southern Islands GPUs, this is a 28nm die shrink of the current Fermi architecture with a few minor tweaks and improvements meant to increase the performance of the chip.
According to previous Nvidia estimates, these changes, combined with the new manufacturing process, should deliver 3 to 4 times the performance per Watt of the Fermi in double-precision 64-bit floating point operations.
If these rumors about Kepler turn out to be true, it would seem like the GPU could share a similar faith with Fermi, since this Nvidia graphics core has also suffered from a series of yield and (power related) performance issues before being released.
