The Next Platform

The Next Platform Deep dives on all that matters infrastructure at scale. Processors, accelerators, memory, storage, n

In an announcement for forthcoming 44PF NREL supercomputer, HPE let fly that Nvidia would be the GPU (no "Grace" Arm CPU...
12/01/2021

In an announcement for forthcoming 44PF NREL supercomputer, HPE let fly that Nvidia would be the GPU (no "Grace" Arm CPU despite 2023 timeline) with something we've never heard of: the A100NEXT Tensor Core GPU...

At this point in supercomputing, it’s becoming an anomaly to see an upcoming double-digit petaflops system not using AMD for CPU and GPU but the National

“There is a practical limit to the computational resources you can apply to this, even at Amazon. We could easily consum...
09/09/2021

“There is a practical limit to the computational resources you can apply to this, even at Amazon. We could easily consume 100X the compute but Amazon couldn’t afford it,” Chapman, who built the infrastructure to support Prime says.

These are still early days for quantum computing, far too soon to talk about domain-specific quantum systems. But if there are areas hungrier than ever

There are many interesting ways to use CXL to expand memory and storage in future systems. These are a few...
09/07/2021

There are many interesting ways to use CXL to expand memory and storage in future systems. These are a few...

The system world would have been a simpler place if InfiniBand had fulfilled its original promise as a universal fabric interconnect for linking all

Dave Brown remembers when Amazon executives would pop into their small office in Cape Town, populated by the entire EC2 ...
09/07/2021

Dave Brown remembers when Amazon executives would pop into their small office in Cape Town, populated by the entire EC2 team of fourteen, and tell them they would be working for a billion-dollar business one day.

Fifteen years later, as Brown looks back, he says he feels silly for finding that laughable. Today, at well over $50 billion, Amazon itself would not be profitable without AWS and plans, including a more physical presence to large customers, an ambitious chipmaking spree, and continued focus on anticipating workloads are to keep that trend kicking...

Dave Brown remembers when Amazon executives would pop into their small office in Cape Town, populated by the entire EC2 team of fourteen, and tell them

It is a pity that Intel didn’t do something more useful with all that excess Data Center Group profit for the past decad...
04/24/2021

It is a pity that Intel didn’t do something more useful with all that excess Data Center Group profit for the past decade – more useful than blowing tens of billions of dollars on share buybacks, at least. Perhaps it would have already built two foundries and already established its Intel Foundry Services business. Who can say?

All good parties come to an end, and the one that Intel has enjoyed for an unbelievable dozen years, starting with the rollout of the “Nehalem” Xeon E5500

Arista ate half of Cisco’s lunch in switching and now it wants to steal its router dessert...
02/22/2021

Arista ate half of Cisco’s lunch in switching and now it wants to steal its router dessert...

It’s a great thing when an upstart supplier of hardware or software lands one of the hyperscalers or large public cloud builders on Earth as a customer.

Big banks look to hyperscale web companies for IT inspiration but the irony is, these infrastructure role models have mo...
02/22/2021

Big banks look to hyperscale web companies for IT inspiration but the irony is, these infrastructure role models have morphed into their own most formidable competitors and are flexibility-native versus building unending workarounds for legacy systems.

Big banks look to hyperscale web companies for IT inspiration but the irony is, these infrastructure role models have morphed into their own most

Even Google had budget and personnel challenges on the road to its TPU. Its "bucket list" of priorities helped the team ...
02/16/2021

Even Google had budget and personnel challenges on the road to its TPU. Its "bucket list" of priorities helped the team navigate through iterations with cost, engineering, shifting workloads, and the need to reuse innovations scoring high...

The inception of Google’s effort to build its own AI chips is quite well known by now but in the interests of review, we’ll note that as early 2013 the

"These models are so adaptable and flexible and their capabilities have been so correlated with scale we may actually se...
02/11/2021

"These models are so adaptable and flexible and their capabilities have been so correlated with scale we may actually see them providing several billions of dollars worth of value from a single model, so in the next five years, spending a billion in compute to train those could make sense,”

There is a new challenge workload on the horizon, one where few can afford to compete. But for those who can, it will spark a rethink in what is possible

China's future Sunway exascale machine is pulling a bunch of different levers to try to get there first. And without usi...
02/10/2021

China's future Sunway exascale machine is pulling a bunch of different levers to try to get there first. And without using GPU or DSP acceleration and by taking an extensible architecture and stretching it

The trade war between the United States and China is not just a top-down political and economic one, but also a technical one. And one could argue that

Someone has to make the chips, and it might as well be Intel one third of the time...
01/20/2021

Someone has to make the chips, and it might as well be Intel one third of the time...

A little more than a week ago, Intel announced that Pat Gelsinger, its former chief technology officer and former manager of the predecessor of its Data

With comparable perf to a V100 GPU but better energy consumption + memory bandwidth potential, Graphcore can turn heads ...
01/20/2021

With comparable perf to a V100 GPU but better energy consumption + memory bandwidth potential, Graphcore can turn heads in lower-precision HPC. That is, if the software stack can be built to meet some tough portability, programmability requirements.

With performance comparable to the Nvidia V100 GPU, a common accelerator in HPC but better energy consumption numbers and memory bandwidth potential,

Address

San Francisco, CA
94108

Opening Hours

Monday 9am - 5pm
Tuesday 9am - 5pm
Wednesday 9am - 5pm
Thursday 9am - 5pm
Friday 9am - 5pm
Saturday 9am - 5pm
Sunday 9am - 5pm

Website

Alerts

Be the first to know and let us send you an email when The Next Platform posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Share

Category