Arm on Arm Panel Discussion 2019
summary
Panel documents that real Arm servers and cloud offerings now exist: vendors ship systems with Ampere and ThunderX2 silicon, a Lenovo-Ampere box launches with few integration issues, and cloud providers offer Arm instances from 1–16 cores and bare metal options. Workstations for native development are in use but low-volume; firmware and supply chain remain the hard work. Surprising detail: about half of an initial Ampere inventory sells to a single customer needing 32-bit builds, so legacy 32-bit tooling still matters. works on arm
AI-written summary · about
chapters (24)
- 0:00Introduction
- 0:38Introductions
- 1:25What is Arm on Arm
- 2:25EDA on Arm
- 3:27Dogfooding in
- 6:49ADA
- 7:21Marvel
- 8:54Caveum
- 9:31A long laundry list
- 10:35Launching a server
- 11:49Access to hardware
- 14:01Where do you want to see traction
- 14:42What are you doing internally
- 16:46Branding and marketing
- 18:21Marketing
- 22:36Its a long road
- 23:57What it takes to be successful
- 26:36Is there a danger of leaving AR32 to the death ports
- 29:06What would be needed to build natively for ARM
- 30:41Software development in the cloud
- 32:53Kernel development in the cloud
- 37:31What should I try today
- 39:24Motivation for adoption
- 44:36Feedback
related talks
47mIntroducing the Next Generation of Windows on Arm
44mLearn how to build the best Arm apps for Windows
10mNative Parity for Development & Testing of SOAFEE Software-Defined Vehicle
25mPutting developers first - Arm adds new developer solutions
06mMigrating to the Arm architecture from a developer’s perspective
40mThe POWER of Supercomputers at Your Fingertips
34mDeveloping Applications using Acceleration Hardware on Power
48mSponsored Session: Your Next Workstation Is In The Cloud