---
title: Unlocking Performance with Next-Gen Topology-Aware CPU Scheduler in VMware Cloud Foundation
url: https://www.vmware.com/explore/video/6403821746112
description: With the advent of high core-count processors, chiplet-based designs, and non-uniform cache hierarchies, traditional scheduling assumptions are broken. Modern workloads—from AI/ML pipelines to large databases and latency-sensitive microservices—demand a new approach to CPU scheduling and Non-Uniform Memory Access (NUMA) locality in VMware Cloud Foundation® (VCF). This session reveals the next-generation ESXi scheduler, with deep awareness of CPU, memory, cache, network, and storage to drive deterministic, topology-aware VM placement. It minimizes data migration costs, enforces locality, and maximizes host memory efficiency across VMs and VMware® vSAN™. These innovations unlock consistent performance gains of up to 45% across enterprise workloads, including databases like Microsoft SQL Server, Oracle, in-memory databases like SAP HANA, Monster VMs, and Kubernetes-based (VKS) platforms.
---

# Unlocking Performance with Next-Gen Topology-Aware CPU Scheduler in VMware Cloud Foundation

[Unlocking Performance with Next-Gen Topology-Aware CPU Scheduler in VMware Cloud Foundation](https://players.brightcove.net/6164421911001/lUBT2rAMW_default/index.html?videoId=6403821746112)  
![Unlocking Performance with Next-Gen Topology-Aware CPU Scheduler in VMware Cloud Foundation](https://cf-images.us-east-1.prod.boltdns.net/v1/jit/6164421911001/2632093a-143e-4195-a4ff-0e6a89de827a/main/1280x720/24m36s971ms/match/image.jpg)

**124 Views**

## Abstract

With the advent of high core-count processors, chiplet-based designs, and non-uniform cache hierarchies, traditional scheduling assumptions are broken. Modern workloads—from AI/ML pipelines to large databases and latency-sensitive microservices—demand a new approach to CPU scheduling and Non-Uniform Memory Access (NUMA) locality in VMware Cloud Foundation® (VCF).

This session reveals the next-generation ESXi scheduler, with deep awareness of CPU, memory, cache, network, and storage to drive deterministic, topology-aware VM placement. It minimizes data migration costs, enforces locality, and maximizes host memory efficiency across VMs and VMware® vSAN™.

These innovations unlock consistent performance gains of up to 45% across enterprise workloads, including databases like Microsoft SQL Server, Oracle, in-memory databases like SAP HANA, Monster VMs, and Kubernetes-based (VKS) platforms.

## Session Info

- **Session Code:** CLOB1112LV
- **Type:** Breakout Session
- **Track:** Cloud Infrastructure
- **Products:** VMware vSphere
- **Level:** Technical 300
- **Event Delivery:** [Las Vegas](https://www.vmware.com/explore/us)
- **Year:** 2026

## Speakers

### ![Qasim Ali](https://static.rainfocus.com/vmware/explore2026lv/att/1712168731929001aZcs/attprofile/Qasim-profile_1782771475299001PSIo.jpg)  
**Qasim Ali**  
Distinguished Engineer  
Broadcom  
Qasim Ali is a Distinguished Engineer at Broadcom in the VCF division with 20+ years of experience in virtualization, cloud infrastructure, resource management and performance engineering. He holds a Distinguished Speaker award at VMworld and is known for his deep technical insights.

### ![Rupesh Shimkhada](https://static.rainfocus.com/vmware/explore2026lv/att/1713215611403001IywK/attprofile/Rupesh_1782864362332001tpvJ.jpeg)  
**Rupesh Shimkhada**  
Sr. Manager Technology Product Management  
Broadcom  
Rupesh Shimkhada is Product Management lead for vSphere platform and a Solutions Architect. He is responsible for all vSphere platform capabilities in ESX. He has been with VMware since 2009 leading various roles, his full profile can be found here https://www.linkedin.com/in/rupesh-shimkhada/

### ![Haoqiang Zheng](https://static.rainfocus.com/vmware/explore2026lv/att/1783636458478001xK32/attprofile/IMG_4853_1785785908782001vAPz.JPG)  
**Haoqiang Zheng**  
Distinguished Engineer  
Broadcom  
Haoqiang Zheng is a Distinguished Engineer at Broadcom. Since joining VMware in 2005, he has played a pivotal role in architecting and developing both the ESXi kernel CPU scheduler and the DRS. He holds a Ph.D. from Columbia University and a Bachelor's degree from Tsinghua University.

## Presentation PDF
[Download](https://static.rainfocus.com/vmware/explore2026lv/sess/1775076119958001y8Nh/presrevpdf/CLOB1112LV_1788483529457001qhTi.pdf)  

## Related Videos

### [Running VMware vSphere® Kubernetes Service: Enterprise Kubernetes Without Full VMware Cloud Foundation](https://www.vmware.com/explore/video/6403822115112)  

![Running VMware vSphere® Kubernetes Service: Enterprise Kubernetes Without Full VMware Cloud Foundation](https://cf-images.us-east-1.prod.boltdns.net/v1/jit/6164421911001/2f420d42-be9b-4001-8f97-3c23f8fcc4ec/main/1280x720/22m55s18ms/match/image.jpg)

**Duration:** 45:50  
**Code:** CLOB1217LV | 48 views  
VMware Cloud Foundation® (VCF) provides a comprehensive platform for running VMware vSphere® Kubernetes Service (VMware VKS), but the full stack can feel like a large commitment for organizations just starting their Kubernetes journey. What many teams do not realize is that you can run production-grade VKS clusters on vSphere 9.x with Foundation Load Balancer (FLB) and VDS networking, without the full VCF footprint.

This session demonstrates the "lightweight VKS" deployment model: enterprise Kubernetes capabilities using infrastructure you already have. We will deploy and manage VKS clusters in vSphere 9.x and explain when this approach makes sense.

You will leave with a practical deployment guide, architecture decision framework, and the confidence that you can start your Kubernetes journey today without waiting for a full VCF transformation.

### [VMware vSphere Kubernetes Service at the Edge: Production Design and Architecture Best Practices](https://www.vmware.com/explore/video/6403821919112)  

![VMware vSphere Kubernetes Service at the Edge: Production Design and Architecture Best Practices](https://cf-images.us-east-1.prod.boltdns.net/v1/jit/6164421911001/44a1798d-8ab0-4ad0-aa9b-29a5740951f6/main/1280x720/27m34s826ms/match/image.jpg)

**Duration:** 55:09  
**Code:** CLOB1146LV | 40 views  
Deploying Kubernetes at the edge introduces unique challenges: resource sizing; scaling to a large number of locations; limited bandwidth; high latency; and managing connectivity loss. This session explores best practices for designing and deploying VMware vSphere® Kubernetes Service (VKS) on VMware Cloud Foundation® (VCF) to support production edge workloads. Attendees will learn how to architect small-footprint clusters, design connectivity between central and edge locations, and enable resilient operations in low/no-connectivity scenarios. You will leave with a framework to deploy and operate VMware VKS at the edge with confidence, ensuring scalability, reliability, and operational simplicity.

### [Four Production-Ready Storage Capabilities Every VMware Cloud Foundation Admin Needs to Know](https://www.vmware.com/explore/video/6403821749112)  

![Four Production-Ready Storage Capabilities Every VMware Cloud Foundation Admin Needs to Know](https://cf-images.us-east-1.prod.boltdns.net/v1/jit/6164421911001/48ecba7e-50c4-4d89-a587-059bba243bfb/main/1280x720/23m10s763ms/match/image.jpg)

**Duration:** 46:21  
**Code:** CLOB1085LV | 46 views  
Are you ready to move beyond the roadmap and start architecting for today? Join VMware® vSAN™ experts Duncan Epping and Pete Koehler for a deep dive into the high-impact storage features currently shipping in VMware Cloud Foundation® (VCF) 9.1.

We will go under the hood of the vSAN Express Storage Architecture (ESA) to show how to achieve massive space savings without compromising performance using global deduplication and the new compression algorithm. We will explain how the adoption of QLC can improve your total cost of ownership and enables new use cases on your VCF-based private cloud. Lastly, we will demonstrate how to deploy VCF Native S3 Object Storage to support cloud-native applications and show the Ransomware Recovery workflow, illustrating how to leverage native recovery tools and ESA’s deep snapshot capabilities to recover with confidence.
