The Next Platform
@TheNextPlatform
Followers
9K
Following
1K
Media
53
Statuses
3K
For all that matters at scale. HPC, hyperscale, large enterprise and now out to the edge.
Joined October 2015
We drill down into the compute and network blades for the forthcoming GX5000 supercomputer from HPE’s Cray unit. $HPE $AMD $NVDA https://t.co/ZkF7Nss8xw
nextplatform.com
Space has always been a premium in the datacenter, but the heat is on – quite literally – to drive up the density of GPU and XPU compute not just because
13
0
1
AMD can break $10 billion in sales for its datacenter products in Q4. $AMD $NVDA $INTC https://t.co/ok5ZSLpcOc
nextplatform.com
If all goes according to the high end of plan, then AMD should kiss $10 billion in revenues in the fourth quarter of this year, and if it was low-balling
0
0
6
Arista’s R4 switches have deep buffers, but high radix may be important in the GenAI era. $ANET $AVGO https://t.co/iFg3Guqotr
nextplatform.com
Chip designs are busting out beyond the reticle limits of lithography machines, making chiplets and high-bandwidth, in-package die-to-die interconnects
0
0
2
We rip apart and enhance IDC’s latest AI infrastructure spending forecast. https://t.co/zSQQ5m9xfr
nextplatform.com
Here is a question for you. What is more difficult: predicting the weather thirty days out or what global AI infrastructure spending will be out to the
0
1
3
Trainium3 with be a tipping point for AI processing for AWS. $AMZN $NVDA $MSFT $GOOG https://t.co/joQSA8NyQY
nextplatform.com
One only need look at the incredible revenues and profits of the datacenter business at Nvidia to know that the world’s biggest compute customers – the
0
0
10
We do a deep think about how Qualcomm might actually take on Nvidia for AI inference. $QCOM $NVDA https://t.co/bTSyedhZcb
nextplatform.com
Qualcomm had datacenter envy back when Intel ruled the bit barns of the world, and now it has it even worse now that Nvidia has shown how AI processing
0
1
14
Some thoughts about Intel’s process and datacenter products roadmaps. Happy Friday, everyone. $INTC $NVDA $AMD https://t.co/OtxWLcG6te
nextplatform.com
Technical debt is a real thing, as any IT manager, programmer, system administrator or SRE, or end user will tell you. If you save money in the short run
1
4
46
IBM is not rich enough to build huge AI clouds, but it is smart enough to build AI infrastructure for its 100,000+ enterprise customers. $IBM $NVDA $AMD https://t.co/QtAYADVnL6
nextplatform.com
Over the next few years, Big Blue may not build the biggest AI business in the world by any stretch of the imagination, but we do have confidence that it
0
0
3
The level of AI software performance improvement that used to take Nvidia two years to do only took a few weeks. $NVDA #OpenAI
https://t.co/52R3NUgaqK
nextplatform.com
One of the neat things about modern AI is that a whole new generation of people in a field outside of economics (but certainly driving the modern economy)
0
1
4
The extreme need to find electricity and conserve it for AI systems is going to accelerate co-packaged optics adoption. $AVGO $NVDA $CSCO $ANET https://t.co/ZSc9sGOiLT
nextplatform.com
If Broadcom says that co-packaged optics is ready for prime time and can compete with other ways of linking switch ASICs to fiber optic cables, then it is
1
1
4
TSMC has stuck to its etching, and it is profiting from its chip making prowess. $TSM $NVDA $AMD $GOOG $MSFT $AMZN https://t.co/vMDN8JiB63
nextplatform.com
The world’s largest and one might argue most important chip foundry is telling Wall Street that its AI-related sales are running ahead of schedule. But
1
2
3
Oracle is going to be the first to deploy AMD’s “Altair” MI450 series GPUs and its “Helios” rackscale servers next year. $AMD $ORCL $NVDA https://t.co/Zn4AEHSgh1
nextplatform.com
It is Oracle OpenWorld CloudWorld AI World this week, so we expect a lot of AI infrastructure announcements from Big Red, with AI being the biggest new
0
2
10
And then there were three: Cisco joins Nvidia and Broadcom with special ASICs for linking AI datacenters together. $CSCO $NVDA $AVGO https://t.co/udePdlUsCm
nextplatform.com
Wide area networks and datacenter interconnects, or DCIs, as we have known them for the past decade or so are nowhere beefy enough or fast enough to take
0
0
2
We pick apart the Dell AI server forecast and give you the numbers out to 2030 that the company hinted at but didn’t nail down. $DELL $NVDA $AMD https://t.co/aOlNVIXGqt
nextplatform.com
It is one thing for the market researchers of the world to make prognostications about hardware, software, and services spending relating to the GenAI
1
0
1
OpenAI and AMD have a new twist on round-tripping money between hardware vendors and model builders. $AMD $NVDA https://t.co/Mheb7fF5xN
nextplatform.com
In early 2024, we all had been wondering how OpenAI was going to pay for its $100 billion Stargate datacenter infrastructure project it was rumored to be
1
1
2
Meta Platforms has pulled an Annapurna Maneuver and bought RISC-V chip developer Rivos. $META $NVDA $AMD $INTC $APPL https://t.co/1O3DitPqcV
nextplatform.com
Of all of the hyperscalers and cloud builders, Meta Platforms has always been the one that we expected to design and have manufactured its own CPU and XPU
1
0
1
Cerebras has $1.1 billion to play with. We make some guesses and suggestions on how to spend it. https://t.co/sLQCGlAuAK
nextplatform.com
What does Cerebras Systems, the first successful waferscale computing commercializer and a contender in the race to provide compute for the world’s
0
0
3
Phase 1 of the Stargate I datacenter in Abilene is fired up, and OpenAI and Meta Platforms are spreading the money around to CoreWeave, too. $CRWV $NVDA https://t.co/T6CcxeTcL0
nextplatform.com
What’s the difference between Meta Platforms and OpenAI? The big one – and perhaps the most important one in the long run – is that when Meta Platforms
0
1
2
Equinix is known for its datacenters, but it was founded on the network effect and it is using AI to help companies move of data and build their own AI with it. $EQIX https://t.co/sv7Y2Az2OS
nextplatform.com
At some point, when the cloud bills get very high and the AI usage starts going up and up, the enterprise is going to want to stop paying a cloud and get
0
0
2
AI spending forecasts the world over are being revised, and up, up, and away. . . . $NVDA $AMD $MSFT $GOOG $AMZN $IT https://t.co/sQYSgNYxdH
nextplatform.com
When we try to predict the weather, we use ensembles of the initial conditions on the ground, in the oceans, and throughout the air to create a kind of
1
0
2