NVIDIA talks about his new marketing strategy for Tesla
A few days ago Nvidia announced its new Tesla cGPUs K10 and K20 Tesla , the first one based on 2 cores GK104 and the second into the future core GK110 . Tesla k10 virtue of being based on 2 cores gk104 new architecture based on Kepler, could look like the Tesla most powerful chip manufactured by Nvidia, but it is not.
Nvidia itself clarifies that have changed their marketing strategy for its Tesla product line, where the new Tesla K10 is focused on servers for the defense industry, government, oil and gas, which does not require great computing power of double precision. The M2090 and M2075 Tesla cGPUs based on the Fermi GF110 core continue to represent the top end in the Tesla product line for high performance computing (HPC) and are aimed at Super-computers, science and manufacturing, and its replacement will be the future Tesla K20 that will be later this year.
Nvidia claims that this differentiation in its product line Tesla responds to current times, where the HPC industry is gradually mature, and consequently require specific solutions for their various needs of parallel computation.
A curious fact confirms Nvidia GPU Boost technology present in its product line Geforce, is not present in the product line architecture-based Tesla Kepler.
Link: NVIDIA’s GPU Computing Ecosystem – The Tesla Aspect (Hardware Zone)Tags: computing, gk104, gk110, GPGPU, HPC, industry, k10, K20, market, Nvidia, parallel, strategy, Tesla