The Next Platform

With the addition of stacked L3 cache AMD can boost the performance of an Epyc processor by at least 50%, which is *three* architectural generationshttps://www.nextplatform.com/2021/11/08/vertical-l3-cache-raises-the-amd-server-performance-bar/ …

Petabyte-scale graph databases are just around the corner.https://www.nextplatform.com/2021/10/28/the-accelerated-path-to-petabyte-scale-graph-databases/ …

It might not garner the same mainstream headlines as the Top 500 list of the world’s largest systems, but a new benchmark may fill in the gaps between real-world versus theoretical peak performance.https://www.nextplatform.com/2021/10/28/real-world-hpc-gets-the-benchmark-it-deserves/ …

Zettascale by 2027, Intel? Ok. 2X compute density yearly, it would still take somewhere around 116k nodes to do the zettaflops trick, and if it could keep the node power constant it would still be *772 megawatts.* Is this...magic?https://www.nextplatform.com/2021/10/27/intel-aims-for-zettaflops-by-2027-pushes-aurora-above-2-exaflops/ …

Our model suggests AMD’s datacenter CPU and GPU business rose by 38% sequentially from the second quarter of this year and more than doubled year on year, reaching $1.08 billion.https://www.nextplatform.com/2021/10/27/amd-datacenter-sales-break-through-1-billion-in-q3/ …

The Sunway and Phyitum architectures pushed LINPACK performance in March past exascale limits in both peak and sustained figures we've learned/confirmedhttps://www.nextplatform.com/2021/10/26/china-has-already-reached-exascale-on-two-separate-systems/ …

Intel facing incursions by AMD, Ampere, and the desire among the hyperscalers and cloud builders to control their own fate, such as Alibaba’s Yitlan 710 homegrown, custom Arm server chip but it's all a long game...https://www.nextplatform.com/2021/10/25/intel-back-to-playing-the-long-game-not-the-wrong-game/ …

Intel hsa reorganized its HPC business and says it is no longer interested in being prime on big supercomputing deals.https://www.nextplatform.com/2021/10/20/prime-contracting-no-longer-one-of-intels-hpc-aspirations/ …

If this Alibaba server chip is etched in 5 nanometer processes and if it is happening any time soon, it pretty much has to be coming from TSMC.https://www.nextplatform.com/2021/10/18/alibaba-rumored-to-enter-the-arm-server-chip-race/ …

Since the JAX API for all numerical elements is based on the familiar NumPy functions, which have already been in use in HPC, JAX can be relatively easy to onboard with, allowing faster uptake of ML functions.https://www.nextplatform.com/2021/10/18/why-jax-could-be-the-next-platform-for-hpc-ai/ …

If Google wants enterprises to use its Spanner database, it can't force them to speak Google Standard SQL. So they've woven PostgreSQL into the guts of Spanner...https://www.nextplatform.com/2021/10/13/google-adds-postgresql-interface-to-spanner-database/ …

Dell was able to steal away the Eni account from HPE with the HPC5 system, which had 1,820 nodes based on a pair of 24-core “Cascade Lake” Xeon SP-6252 Gold processors and 4 Nvidia V100 GPUshttps://www.nextplatform.com/2021/10/12/eni-chooses-utility-pricing-for-new-hpc4-supercomputer/ …

With Project Capitola, in dev for 3 years, the ESXi hypervisor is turned into a memory pooling and aggregation stackhttps://www.nextplatform.com/2021/10/11/vmware-stretches-esxi-to-be-a-disaggregated-memory-hypervisor/ …

If quantum bets are correct, a partnership between both could mean joint leadership in the emerging space—one that will require established companies to back untested technologies.https://www.nextplatform.com/2021/10/11/raytheon-ibm-partner-for-quantum-in-defense-aerospace/ …