By the technical details of Nvidia Graphics Technology Conference in September for the first time described in the next-generation graphics cards for Fermi-based information flow continues. By Nvidia released new information on the official web site according to the Fermi-based solutions with 512 and (or) 448x parallel processing unit is included. 530mm ^ 2 with a membrane structure which has quite large for the Nvidia GPU clock speed and power consumption GF100 value and that some problems related to working on new revisions of updates for some time that was often claimed.

According to a new documentation to light the Nvidia GeForce 300 series in the GF100-based parallel processing units instead of 512 parallel processing units that can be 448x. Regarding desktop solutions available information is not yet clear, but 7 / 24 Tesla to work with C20 series based on Fermi developed a parallel processor model 448x is officially verified. Support 384-bit memory bus and the 1.4GHz operating frequency with parallel processor model for the Nvidia Tesla stated that the 225 watt thermal design power. But this information is related to the Tesla series has once again reminded benefits.

As mentioned above, since Tesla series of professional graphics solutions under full load it is designed to work for days than desktop solutions to particular time rates have some important differences can be. On the other hand GeForce GTX 360 and GeForce GTX 380 models running at speeds which would exist for any official explanation about the moment does not exist.

Name: nvidia_fermi_tesla_dh_fx57.jpg Views: 88 Size: 88.7 KB