Refresh
We now have one other associate becoming a member of Su up on stage. This time it is Karan Batta, SVP at Oracle.
AI, naturally, is the most important speaking level right here. Oracle is seeing “incredible levels” of efficiency and effectivity on AMD GPUs, Batta says.
The firm is constant to scale up capability for MI300X sequence GPUs for patrons, Batta says. There's extra to come back in the 12 months forward as effectively.
Rapid fireplace buyer chats right here with Naveen Rao, VP of AI at Databricks joins Su on stage now to debate the corporate's partnership with AMD.
“We've been on this journey for a while,” Rao says. “We've achieved remarkable results.”
Leveraging the MI300X the corporate has delivered marked compute enhancements. It's an actual “powerhouse” of a GPU, Rao says.
So what's AMD's large promoting level right here then with regard to GPU capabilities?
The Instinct MI300X sequence is the chip maker's flagship product right here, and in phrases of efficiency it's miles exceeding what opponents can present.
The MI300X outperforms Nvidia's H100 sequence on a number of key benchmarks, Su explains. All informed, this represents 1.3x greater efficiency on Llama 3 and Mistral-based workloads.
Data middle AI demand is completely skyrocketing. If you are a daily reader of ITPro then you definately'll have come throughout a number of the varied bits of analysis we have lined on this topic.
Insatiable demand from enterprises globally has prompted an enormous surge. This, Su says, represents an enormous alternative for AMD to capitalize on this pattern with its Instinct GPU sequence.
Looking forward, the numbers concerned listed below are merely thoughts boggling.
So how's this benefitting Google Cloud? Well, leveraging EPYC machines, the tech large has minimize prices by 40% and delivered a 15% efficiency enchancment.
Big numbers right here and once you have a look at this from a monetary perspective, very important for the hyperscaler.
Image 1 of 2
We've bought our first associate on stage now from Google Cloud.
“The demand for AI compute power is insatiable.”
We've moved onto the information middle portfolio right here. A very thrilling space for AMD proper now – large features and large revenues boosts in latest months.
AMD EPYC now boasts a file market share – and it is rising, Su says. At the top of the second quarter, AMD exited with a 34% income share in this area.
“When we think about AI, it really is about using the right compute for the right application,” Su says.
There's actually no ‘one-size-fits-all' method to AI, and AMD is eager to allow enterprises to get these inventive juices flowing and adopting the expertise on their phrases, at their tempo, and primarily based on their distinctive particular person wants.
We're getting a quick rundown of AMD's present technique.
“It's about employing the entire ecosystem – the cloud, OEMs…our goal is to create an open ecosystem,” Su says.
Su says AMD is “really committed” to driving open innovation in excessive efficiency computing and AI infrastructure. This has been a recurring speaking level in the corporate's technique of late – ecosystem constructing and collaboration.
And right here we go! AMD chief government Lisa Su is on stage to kick issues off at Advancing AI.
“Let's talk a little bit about AI,” Su says. “Actually, we're going to talk a lot about it.”
Setting the tone right here proper off the mark.
We are seated now and good to go. Just a matter of time till we hear from Lisa Su in her opening remarks right here at AMD Advancing AI. An excellent buzz across the keynote corridor forward of issues commencing.
Like we stated, an actual distinction in the amount of individuals right here on the Moscone Center right now.
Image 1 of 2
But what's AMD doing in this regard? Its Ryzen sequence is a key focus for the corporate in its large AI PC push, so we'll seemingly hear extra on this through the opening keynote.
There has been some concern from trade stakeholders on the Ryzen sequence, nonetheless.
After AMD unveiled its Ryzen Pro enterprise-focused processors, the corporate promised important efficiency features, however stories in April urged they had been unlikely to fulfill the steep efficiency threshold required to be designated as an ‘AI PC' by Microsoft.
A key speaking level right here was that the AI processing energy of those chips appeared to fall in need of the 40 trillion operations per second (TOPS) commonplace set by Microsoft.
You can learn extra on this in our protection under.
• AMD's new sequence of AI-enabled processors seemingly fall in need of Microsoft’s efficiency necessities for next-gen AI PCs
With this in thoughts, we may see AMD come out weapons blazing with a brand new replace to the Ryzen sequence.
We're anticipating to listen to a couple of large push on AI PCs from AMD right now. The rise of the AI PC has been a giant speaking level during the last 12 months, with producers and chip makers alike ramping up improvement in this house.
Intel CEO Pat Gelsinger claimed final 12 months that the AI PC could be the “star of the show” in 2024, and he wasn't fallacious. This has been a giant space of focus for Intel, one among AMD's key opponents in this house.
It's nonetheless early – and we expect a torrent of individuals to converge on the Moscone Center this morning – however I've to say town middle in San Francisco has been a breath of recent air in comparison with the carnage that Dreamforce introduced with it simply a few weeks prior.
No sea of lanyards or the cacophony of automotive horns in chaotic visitors.
We've nonetheless bought a short while till the opening keynote session kicks off, however in the meantime why not try our preview forward of the convention for a style of what we are able to anticipate to listen to about?
• AMD has put in the groundwork for a serious AI push whereas the tech trade fawned over Nvidia
It's clear that AMD has been cooking up a storm to date in 2024 because it primes itself for a battle with Nvidia.
All eyes have been on the latter during the last two years because the emergence of generative AI, however AMD has been placing in the groundwork of late to meet up with its nice rival.
Today we'll hear extra about the way it plans to ramp issues up in the 12 months forward, however now we have seen a raft of thrilling acquisitions, together with two in the final couple months alone. These included a $4.9 billion deal to amass ZT Systems, which marked an enormous assertion of intent from the chip maker.
In July, AMD additionally introduced the $655 million acquisition of SIlo AI amidst its large AI push.
