NVIDIA AD102 et AMD Navi 31 dans une course pour atteindre 100 TFLOP FP32 First

Une course technologique se prépare entre NVIDIA et AMD pour savoir quel GPU de la marque atteindra le 100 TFLOP/s pic de débit FP32 marque en premier. L'architecture graphique TeraScale d'AMD et le “RV770” silicium, were the first to hit the 1 TFLOP/s mark, chemin du retour 2008. It would take 14 years for this figure to reach 100 TFLOP/s for flagship GPUs. NVIDIA’s next generation big GPU based on the “Créer la malédiction – PlayStation.Blog,” the AD102, is the green team’s contender for the 100 TFLOP/s mark, selon kopite7kimi. Pour y parvenir, all 144 La prochaine carte graphique phare de NVIDIA (La prochaine carte graphique phare de NVIDIA) ou 18,432 couleurs CUDA, of the AD102 will have to be enabled.

From the red team, the biggest GPU based on the next-generation RDNA3 graphics architecture, “Navi 31,” could offer peak FP32 throughput of 92 TFLOP/s according to greymon55, which gives AMD the freedom to create special SKUs running at high engine clocks, just to reach the 100 TFLOP/s mark. Le Navi 31 silicon is expected to triple the compute unit count over its predecessor, résultant en 15,360 processeurs de flux. Both the AD102 and Navi 31 are expected to be built on he same TSMC N5 (5 nm EUV) nœud, and product launches for both are expected by year-end.