Perplexity makes AI inferencing on old chips less perplexing
Research also points to success beyond Nvidia on AWS architecture
News - AI Hardware
-
-
Microsoft’s Blackwell Ultra VMs push AI performance past the million token milestone
MLPerf Inference tests see the new Azure ND GB300 v6 VMs achieve token performance that ‘fundamentally alters the calculus of AI efficiency’
-
Intel plans to cut up to 20% of its manufacturing workforce: report
Majority of layoffs expected to take place in July
-
Google unveils seventh-generation TPU, Ironwood
Comes in both 256 and 9,216 chip configurations
-
Tesla’s 50,000 GPU Cortex supercomputer went live in Q4 2024
The Nvidia H100 cluster is alive and kicking in Texas
-
FPGA company Altera announces independence from Intel
Almost a year since spinout was first announced
-
Nvidia invested $1B in AI companies in 2024
Participated in more investment rounds than Amazon and Microsoft
-
Edgecore Networks debuts 800G switch to power AI/ML workloads
The new switch, dubbed the DCS560, packs an impressive 51.2 Tb/s of total switching capacity with 64 ports supporting 800G interfaces
-
SK hynix posts first loss in 10 years as memory market struggles
Joining Micron and Samsung in trying to survive the memory market crash
-
AMD takes on Intel with cache stacked Milan-X EPYCs
AMD's EPYC 3 Milan-X stacks 768 megabytes of level-three cache atop the die for a claimed 66% boost in “technical computing workloads.”
-
Intel, Linux Foundation launch open source silicon groups
Both new groups formed just days before the granddaddy of open source hardware groups, Open Compute Project (OCP), holds its annual summit. All three focus on “fundamentally different areas”