Accelerating climate change, including extreme temperature events, has intensified the need to understand evolutionary responses of populations to heat stress. Hybridization - the exchange of genetic material between divergent species - can facilitate adaptation, because it instantly introduces new genetic variation throughout the genome. However, little is known about the long-term consequences of hybridization. For instance, how is the genetic material from opposite species sorted and arranged in hybrid genomes over time, and how does this affects their fitness? Several strains of hybrids between S.paradoxus and S.cerevisiae, two species with different thermal niches, have been evolved for 200 generations under conditions representing thermal extremes (cold 16°C, hot 30°C), and fluctuating temperatures. After measuring their fitness, these strains have been sequenced to identify the genomic basis of thermal adaptation, which can be achieved thanks to the computing ressources of UPPMAX.