Having already said that it plans to make sure the 28nm product roadmap arrives as smoothly as possible, NVIDIA appears to have chosen the more or less exact times of arrival for the next generation of GPUs.That NVIDIA was going to release a new line of graphics products, sooner or later, was obvious, but only now have the details of the so-called N13 family been gathered up.
Some time ago, NVIDIA came out and said that there will not be a repeat of the 40nm yield issues when the 28nm chips come out. It also said that it took more time to study the process.
Now, the list of GPUs, complete with VRAM, thermal design powers and times of arrival has made its way to the web.
Codenamed N13 Series, the chips are ten in number and are divided into three categories, according to the target consumer segment.
The N13E are the Enthusiast-grade models and will arrive next year, in January (N13E-GS1) and April or May (N13E-GTX), though the samples will ship earlier (early November and early 2012, respectively). They have 2 GB of GDDR5 VRAM backing them and 256-bit memory interfaces.
The mainstream and low-end GPUs will arrive earlier, during this year's December in fact (2011), with the mobile parts set to appear fastest. This means that their first samples will be sent out during early October.
Their memory will be of 1 GB or 2 GB DDR3/GDDR5 and the bandwidth will be of 64 bits or 128 bits, depending on model.
All in all, the codenames are similar, in pattern, to the N12 that spawned the 500 series of adapters, though changes may still happen by the time anything is made official.
There is also the very real possibility that the Santa Clara, California-based GPU maker will have some 40nm-based products to fill in the gaps left by this list.
What remains is to see how much longer NVIDIA takes to bring out the even more advanced Kepler chips on the mobile front (the above are Fermi).