With the beginning of 2023, we already have the new PCI-E 5.0 PLATFORMS, both from the hand of AMD with Genoa, such as Intel with Sapphire rapids.
All this new technology means doubling the bandwidth of the communication buses, as happened previously with the change from PCI-E 3.0 to PCI-E 4.0.
Exchanges
Greater number of Cores
memory increase
More speed
Better interconnectivity
New cooling technologies
Exchanges
There have also been other changes, which come hand in hand with the new buses and which we will see in the coming months:
Greater number of Cores
0
Initials
0
Ready
0
Initials
0
Ready
Increased number of Cores in processors: in the case of AMD, initially up to 96 Cores. In the immediate future up to 128 Cores, with the new Pergamo series.
In the case of Intel, initially up to 48 Cores, but Sapphire will go up to 60.
memory increase
DDR5 at 4800 MHz and 12 pipeline
Before: DDR4 at 3200 MHz and 8 pipeline
The transition from DDR4 memory at 3200 MHz to DDR5 memory at 4800 MHz and from 8 pipeline to 12 memory, depending on the processors. This is logical, because faster processors and more Cores, more need for speed access to memory.
More speed
0
IOPS
Although they have already been announced, we will soon see PCI-E 5.0 technology NVME drives, which in theory will allow up to 4 million IOPS.
These speeds can now be achieved with current and new controllers from LSI/Broadcom and Supreme Raid (based on the chip from the Nvidia T1000)
Better interconnectivity
During the first half of this year, Gigabyte HGX platforms based on the NVIDIA H100 SXM5 will be released. These platforms will have a PCI-E 5.0 bus and improvements in intercommunication through the internal NVLINK bus.
In addition, they will have connectivity via NVSWITCH with the new 400 Gb/s Infiniband cards and will be able to connect directly to this bus. All this is focused on Artificial Intelligence solutions.
The H100 significantly improves the performance of its predecessor, the A100. In HPC, the improvements are of the order of 2-3 times, while in AI they are 5-6 times.
It also has a new bandwidth for better performance in Artificial Intelligence tasks. This platform is ideal for those looking for high performance in AI and HPC solutions.
New cooling technologies
During this quarter, we will have very dense platforms in these new technologies of Giga Computing
That will mean processors with consumptions of up to 500 W and the new 100 W H5 SXM700
That's why we will also launch all new oil immersion technology
