"With Graviton5, capacity is in ample supply with 192 Arm Neoverse V3 cores fabbed on TSMC's 3nm process tech. Those cores are backed by a larger 192MB L3 cache, which helps to cut down on cache misses and improves performance by reducing the amount of data that needs to be fetched from slower DRAM. The chip also features an improved memory subsystem that boosts speeds to 7200 MT/s, with support for 8800 MT/s DIMMs in the works."
""Linking processors introduce new latency paths, and when a core needs to access memory on the other CPU, the request has to move across that interconnect, and that adds latency, extra protocol overhead, and sometimes even queuing. In certain scenarios, it could take up to three times longer," AWS EC2 VP David Brown said in Thursday's keynote. According to Amazon, Graviton5's higher core count effectively cuts inter-core latency by roughly a third, benefiting workloads like online gaming, high-performance databases, electronic design automation, and analytics."
Graviton5 integrates 192 Arm Neoverse V3 cores manufactured on TSMC's 3nm process, creating a single-socket CPU with a larger 192MB L3 cache to reduce cache misses and lower DRAM accesses. The memory subsystem supports 7200 MT/s today with 8800 MT/s DIMMs planned. New M9g instances provide about 25% higher performance than Graviton4-based M8g instances by consolidating two 96-core CPUs into one 192-core socket. The single-socket design cuts inter-core latency by roughly a third, improving performance for online gaming, high-performance databases, electronic design automation, and analytics. Compute nodes still contain two sockets that share a Nitro smartNIC.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]