Refresh
Take a look at AMD Instinct MI300X GPU…
The AMD Pensando Software-in-Silicon Developer Kit (SSDK) gives customers the ability to develop or migrate services to deploy on the AMD Pensando P4 programmable DPU in coordination with the existing set of features already implemented on the AMD Pensando platform.
The breakout sessions have begun as we hear from some of AMD’s partners giving insight on AMD’s next generation DPU codenamed “Giglio,” which aims to bring enhanced performance and power efficiency to customers, compared to current generation products.
It’s expected to be available by the end of 2023.
Up next… lunch time.
Thanks for tuning in this morning. We’ll be back shortly to dig into all the news updates that were announced so far.
The MI300X is based on the next-gen AMD CDNA™ 3 accelerator architecture and supports up to 192 GB of HBM3 memory to provide the compute and memory efficiency needed for large language model inference and generative AI workloads.
We got to see it live in action for the first time as CEO Dr. Lisa Su shows what great fun can be had with MI300X as AI writes a poem about San Francisco.
And it did… Quickly!
Take a look at AMD Instinct MI300X, offering 2.4 times more memory and 1.6x more HBM bandwidth.
The announcements keep on coming!
AMD also unveiled that the AMD Instinct MI300A, the world’s first APU Accelerator for HPC and AI workloads, is now sampling to customers.
The rate of innovation caused by AI is unprecedented. AMD’s president says it’s ready to achieve the next breakthrough in AI for its customers.
AMD officially announces its partnership with Hugging Face, an American company that develops tools for building applications using machine learning.
Victor Peng, President of AMD unveils ROCm 5 – a comprehensive suite of data center optimizations.
“We are very early in the life cycle of AI, but the largest opportunity is in the data center,” says AMD CEO, Dr. Lisa Su.
It’s time to hear how AMD is ramping things up using AI…
Introducing the Smart Switch. Another way AMD has evolved the data center.
With the launch of AMD Genoa-X, the fourth-gen EPYC processors, Microsoft Azure is supporting the chips with new Hbv4 and HX cloud instances for technical computing.
A brief mention of AMD Siena, which is coming to market during the second half of this year.
With the launch of AMD Genoa-X, the fourth-gen EPYC processors, Microsoft Azure is supporting the chips with new Hbv4 and HX cloud instances for technical computing.
Up next…
Dan McNamara, SVP and GM, server business unit unveils AMD’s 4th Gen AMD
EPYC processors with AMD 3D V-Cache technology, the world’s highest performance x86 server CPU for technical computing.
Meta is aggressively scaling their data centers with the helping hand of AMD. Meta prepares to deploy Bergamo to help its workloads.
First look at the 4th Gen AMD EPYC “Bergamo”. With up to 128 “Zen 4c” Cores. Bergamo is shipping in volume to hyperscale customers.
AWS vice president Dave Brown joins Dr. Lisa Su on stage to announce Amazon EC2 M7a Instances, which delivers up to 50% more performance than M6a instances. It’s available in preview today!
AMD chief executive officer, Dr. Lisa T. Su kicks things off with the opening keynotes where she reveals Epyc delivers 1.8x more performance and best in class energy efficiency.
What to expect?
Since this is AMD’s flagship data center event, one can expect to hear about
- New EPYC CPUs including new 4th generation parts. We expect to get more details about Bergamo, which AMD unveiled back in November 2021. This is a cloud-native CPU with 128 cores and aimed at the growing threat from Arm-powered processors, either from hyperscalers (e.g. Amazon Graviton) or from the likes of Ampere Computing.
- The Instinct MI300. AMD’s data center APU that’s expected to combine GPU, CPU and high bandwidth memory, lots of it. Check out the latest news about the most exciting piece of silicon that AMD announced for, well, over a decade.
- Pensando DPU. A critical piece in AMD’s datacenter story remains is the SmartNIC, allowing hyper scalers to dramatically improving performance by offloading cloud/virtualization overheads to a dedicated piece of silicon.
Five minutes till show time and all seats have been filled. The mood is lively, following the breakfast selection and unlimited coffee supply. Everyone is eagerly waiting.
We’ve checked in and are all waiting in anticipation for the grand hall doors to open and the conference to kick off. Just 20 minutes left till kick off…