# VMware / Broadcom 日跟踪 · 2026-09-29
> 日期：2026-09-29（Asia/Shanghai，周二）。原料：archive/2026-09-29/vmware/。上一日成功日跟踪 2026-09-28（周一；周末 09-26/27 无早报）。抓取全库：Saved 265 / skipped 29 / errors 40；blocked：blocksandfiles.com / docs.redhat.com / e.huawei.com / support.huawei.com / www.nextplatform.com / www.reddit.com / www.theregister.com。

## 摘要
- **新稿 · Optimizing AI Deployments with VMware Cloud Foundation (Part 2)**（厂商博客 09-28）：Maximizing GPU Utilization for Best TCO；passthrough（DirectPath I/O）vs NVIDIA vGPU（time-sliced / MIG）；系列 Part **2/4**（Part3 CPU/Memory、Part4 Time-Slicing）；厂商实验称合并 VM/共享 GPU 可抬高吞吐与 SM 利用率（如 Config3 Offline ~94% SM）——**Broadcom Performance 白纸/博客实验，数字厂商自称，非独立实测**。
- **新稿 · The Cost Case Is Clear: Infrastructure Economics with VVF 9.1**（厂商博客 09-28）：VVF 9.1 TCO 叙事——Memory Tiering 自称 offload **20–25%**、server TCO **up to 42%**、DB vs 9.0 **up to 15%** better；vSAN dedupe/compression storage TCO **up to 39%**；Quick Patching 30→**<5** min（**80%**）；diagnostic **up to 60%**；Forrester TEI downtime **up to 32%**（VCF9 复合组织）。**全部 Broadcom internal testing / Forrester commissioned——标厂商自称，勿当独立实测**；本地无 TEI PDF。
- **续档（09-28 首见，今日不升格）**：esxcli memtier 命名空间；HCX 9.1 Part 3 Network Extension/迁移 howto。
- **续档（09-25 首见，今日不升格）**：双 MQ Leaders 厂商 PR + CC 六用例 #1 自称；VKS Add-Ons webinar。
- **续档**：AI Part 1 / HOL 9.1 / Resilience / Memory Tiering NVMe / VCF 9.1.1。
- VR 09-23 MQ 媒体转述仍有效；**本地仍无 Gartner PDF**。

## 原文摘录

### [Optimizing AI Deployments with VMware Cloud Foundation (Part 2)](https://blogs.vmware.com/cloud-foundation/2026/09/28/optimizing-ai-deployments-with-vmware-cloud-foundation-part-2/) · 新稿
- 归档：`vmware/optimizing-ai-deployments-with-vmware-cloud-foundation-part-2-8d0d611d.txt`
- URL：https://blogs.vmware.com/cloud-foundation/2026/09/28/optimizing-ai-deployments-with-vmware-cloud-foundation-part-2/
> Part 2 (this blog post): Maximizing GPU Utilization for Best Total Cost of Ownership… Part 3… Part 4: Utilizing Time-Slicing…（**系列 Part2/4；厂商博客**）
> In VCF, DirectPath I/O (passthrough) and NVIDIA vGPU technologies are used to allocate GPU resources to virtual machines… vGPU includes… time-sliced vGPU and MIG vGPU.（**厂商产品主张**）
> … four servers… ~34-38% GPU SM utilization… Config 2… 2328→9314 tokens/second… 49%… Config 3… 18621… 79%… Offline… 23365… 94%.（**Broadcom Performance 实验数字——厂商自称，未独立复测**）

### [The Cost Case Is Clear: Infrastructure Economics with VMware vSphere Foundation 9.1](https://blogs.vmware.com/cloud-foundation/2026/09/28/the-cost-case-is-clear-infrastructure-economics-with-vmware-vsphere-foundation-9-1/) · 新稿
- 归档：`vmware/the-cost-case-is-clear-infrastructure-economics-with-vmware-vsphere-fo-0975a8ca.txt`
- URL：https://blogs.vmware.com/cloud-foundation/2026/09/28/the-cost-case-is-clear-infrastructure-economics-with-vmware-vsphere-foundation-9-1/
> Figure 1: vSphere Foundation 9.1 Key TCO Proof Points (source: Broadcom internal testing, May 2026)（**来源标注：Broadcom internal**）
> … enhanced NVMe Memory Tiering… offloads 20 to 25% of memory accesses… reduces server TCO by up to 42%… database… up to 15% better performance compared to version 9.0…（**厂商自称 [2] Broadcom internal testing**）
> … vSAN global Deduplication and enhanced compression, delivering up to 39% lower storage TCO…（**厂商自称**）
> Quick Patching for vCenter reduces maintenance windows from 30 minutes to under 5 minutes, an 80% reduction… Native diagnostic analysis… up to 60%…（**厂商自称**）
> … storage related unplanned downtime declines by up to 32%… Forrester TEI study of VMware Cloud Foundation 9… composite organization…（**Forrester commissioned on behalf of Broadcom——勿当独立实测**）
> Disclaimer: All performance and cost improvement claims based on Broadcom internal engineering estimates…（**文末免责**）

### [One Namespace to Rule Them All: The New esxcli memtier Namespace](https://blogs.vmware.com/cloud-foundation/2026/09/25/one-namespace-to-rule-them-all-the-new-esxcli-memtier-namespace/) · 续档（09-28 NEW）
- 归档：`vmware/one-namespace-to-rule-them-all-the-new-esxcli-memtier-namespace-b2b49da0.txt`
> （esxcli memtier 收敛续档；**今日不升格**）

### [The Complete Guide to Workload Mobility… Part 3](https://blogs.vmware.com/cloud-foundation/2026/09/25/vcf-9-operations-hcx-network-extension-migration-part-3/) · 续档（09-28 NEW）
- 归档：`vmware/the-complete-guide-to-workload-mobility-with-vmware-cloud-foundation-o-fbcaaf14.txt`
> （HCX 9.1 Part 3 续档；**今日不升格**）

### [Broadcom (VMware) Recognized as a Leader in Two 2026 Gartner MQ Reports](https://blogs.vmware.com/cloud-foundation/2026/09/24/broadcom-vmware-recognized-as-a-leader-in-two-2026-gartner-magic-quadrant-reports/) · 续档
- 归档：`vmware/broadcom-vmware-recognized-as-a-leader-in-two-2026-gartner-magic-quadr-8b97b53f.txt`
> （双 MQ PR + CC #1 自称续档；本地无 PDF）

### [Webinar Recap: Operationalizing VMware vSphere Kubernetes Service with Add-Ons](https://blogs.vmware.com/cloud-foundation/2026/09/24/webinar-recap-operationalizing-vmware-vsphere-kubernetes-service-with-add-ons/) · 续档
- 归档：`vmware/webinar-recap-operationalizing-vmware-vsphere-kubernetes-service-with--5b06fcf0.txt`
> （VKS Add-Ons webinar 续档）

### [Optimizing AI Deployments with VMware Cloud Foundation (Part 1)](https://blogs.vmware.com/cloud-foundation/2026/09/21/optimizing-ai-deployments-with-vmware-cloud-foundation-part-1/) · 续档
- 归档：`vmware/optimizing-ai-deployments-with-vmware-cloud-foundation-part-1-3ff2cfa6.txt`
> （AI Part 1 续档；系列接 Part 2）

## 对基础档案的影响
- **最近动态追加**：AI Deployments Part 2（GPU utilization / passthrough vs vGPU；系列 2/4）；VVF 9.1 TCO 经济学博客（Memory Tiering / vSAN / Quick Patching / Forrester TEI——**一律厂商自称/委托研究**）。
- 09-28 memtier / HCX Part3 改为续档指针，**勿再标 NEW**；双 MQ PR / VKS webinar 续档；**勿把 TCO/TEI 数字写入已核实实测位**。