At NVIDIA GTC 2025, we saw the Pegatron NVIDIA GB300 NVL72 rack along with a number of lower-density PCIe and SXM GPU systems
Kioxia, AIO Core, and Kyocera are now demoing a faster PCIe Gen5 over optical SSD designed to move SSDs out of AI and server racks
The IBM z17 mainframe brings new AI capabilities with the Telum II processor and its built-in AI accelerator and the PCIe-based Spyre AI cards
We saw the ASUS AI Pod, a NVIDIA GB300 NVL72 rack at NVIDIA GTC along with new HGX B200 and HGX B300 systems at NVIDIA GTC 2025
We tried running the 94GB NVIDIA H100 NVL PCIe card as a single GPU without the NVLink bridge, and it worked, as expected
We have a quick look at the QNAP QNA-T310G1S, a Thunderbolt 3 to 10G SFP+ network adapter that we have been using for some time
At NVIDIA GTC 2025 we saw the new NVIDIA GB300 NVLink Switch tray open and on display, including its custom liquid-cooling solution
We show why you might prefer one DIMM per channel configuration using an AMD EPYC 4004 series server and 192GB of DDR5 ECC UDIMMs
We show you how you can run the 1.27TB Deepseek-R1 671B model at FP16, and even look at running it alongside VMs in your clusters
The new MLPerf Inference v5.0 results are out with new submissions for configurations from NVIDIA, Intel Xeon, and AMD Instinct MI325X