Atlas AI, Part II
Chips Out Loud
Guests Abe Tarapani, CEO, and Marshall Burke, Co-Founder of Atlas AI rejoin the podcast for a continuing conversation about the symbiosis between the companies as an investment of Micron Ventures. Jim Greene explores the impact of the CHIPS Act on infrastructure in the US and how Atlas AI plays a role for companies like Micron in that effort. ”Think about just the visionary nature of these investments, and how substantial they have the opportunity to be,” Abe imagines, “There is an opportunity for companies … to proactively think about how to direct the investment in a way that drives...
info_outline
Atlas AI, Part I
Chips Out Loud
The Chips Out Loud podcast greets new guests Abe Tarapani, CEO, and Marshall Burke, Co-Founder of Atlas AI. Jim Greene discusses the mission alignment between the companies, with Atlas AI as an investment of Micron Ventures. ”I was wandering around in Sub-Saharan Africa, working with a number of organizations, both big and small, trying to understand how they could improve livelihood outcomes among lower income populations in Sub-Saharan Africa,” Marshall begins, “it was really rich data, but it was fundamentally just not very scalable and it got very frustrating for me in that we could...
info_outline
Trends in High Performance Computing
Chips Out Loud
As the supercomputing conference SC22 takes place in Dallas, Texas, Steve Pawlowski, Corporate Vice President Advanced Memory Systems, joins host Jim Greene on the newest edition of the Chips Out Loud podcast to discuss topics and trends in High Performance Computing. ”If you really try to break out the cost of the different compute structures versus the cost of moving data in order to be able to do a single floating point operation, you spend about 100 times more energy doing the data movement than you do actually doing compute, “ Steve confides, “And so the idea was, okay, instead of...
info_outline
Micron DDR5 for the Data Center
Chips Out Loud
Malcolm Humphrey joins host Jim Greene on the latest installment of the Chips Out Loud podcast. Malcolm serves as the Vice President and General Manager of the Compute DRAM at Micron. Today the excitement surrounds the impact of the new generation DDR5 memory for data center customers. “DDR5 refers to Double Data Rate Generation 5 Dynamic Random-Access Memory, which is a mouthful,” Malcolm confides, “ I just call it D5 for short. But really what it is, it's the newest generation of high-speed memory standard, optimized for compute applications, and in particular, the data...
info_outline
1-beta DRAM leadership
Chips Out Loud
On the heels of the news that Micron has begun shipping QS-sampling LPDDR5X components developed on the new 1-beta process node to its smartphone customers, host Jim Greene welcomes Thy Tran, Vice President DRAM Process Integration, to the Chips Out Loud Podcast to discuss the emergent technology. “One beta technology is our next generation of DRAM … which enables further shrink of our memory array cell size, as well as the rest of the circuitry in the die,” Thy reveals. “Besides over 35% memory bit density gain, our customers can expect approximately 15% in power savings and...
info_outline
Driving What's Next in Mobile
Chips Out Loud
Ross Dermott, Vice President, Mobile Product Line Management, Mobile Business Unit at Micron Technology joins Chips Out Loud host Jim Greene to talk about the new mNAND solutions for top-of-line smartphones, its intelligent vertical integration, and the roadmap for future-proofing the mobile experience. “The NAND-based products that we're building for mobile are a category called Managed NAND,” Dermott explains. “Managed NAND combines NAND technology with a controller and firmware. And that controller and firmware are really what help optimize the performance of that NAND...
info_outline
Micron’s Journey to 232L Tech
Chips Out Loud
In this episode, Russ Meyer, Corporate Vice President Non-Volatile Memory, drops by to talk about the announcement of 232-layer memory in Micron’s new 3D NAND offering. “I think everybody's aware of the torrential pace of NAND scaling over the last decades and this 232-layer technology is the next progression to enable that scaling,” Meyer states. “That scaling, as we all know, has been a foundation for the revolution that we're all living through, the way we interact with the data..” Host Jim Greene queries Russ on a bevy of topics: Why do layers matter and what makes stacking...
info_outline
Addressing Automotive Requirements
Chips Out Loud
Guests Robert Bielby and Marco Montagnana sit with Chips Out Loud host Jim Greene to talk about learnings from Micron Technology’s foray into the automotive sector, with its standards-based technologies and stringent safety requirements. Robert serves as Senior Director, Automotive System Architecture. Marco operates as Senior Manager, FUSA & Cybersecurity Office. “I think we're at a point where we have sufficient compute performance to really be able to address self-driving vehicles,” Robert Bielby explains. “The problem is a little bit more difficult than people had expected or...
info_outline
CXL in the Data Center
Chips Out Loud
This week’s episode of Chips Out Loud features guests Ryan Baxter and Eric Caward discussing the topic of Compute Express Link. Ryan works at Micron as Senior Manager, Data Center Segment while Eric resides as Senior Manager, Emerging Technology. CXL is the premiere open standard for high-speed CPU connection to device and memory in high performance data centers. “It's really the first interface of its kind in servers,” Baxter explains. “It's really an interface built for efficient memory accesses. The protocol itself is byte addressable. It's low latency. It's high bandwidth....
info_outline
Authenta and Security at the Edge
Chips Out Loud
On today’s episode of Chips Out Loud, guest Chris Jacobs shares his thoughts on Micron® Authenta™ technology, a unique level of cyber protection for IoT devices. Chris works as Vice President, Marketing Embedded Unit at Micron Technology. “Authenta is a set of features added to a flash storage device,” Jacobs reveals. “These features add security layers by protecting data, maintaining its integrity and providing recovery mechanisms for damaged data. From a bigger picture point of view, Authenta is a combination of flash-based technology and Authenta cloud platform that serves as a...
info_outline“If you think about the broad industry support, it's really going to drive inertia that builds innovation and will accelerate in the data center, which will be very exciting,” Eric Caward concurs. “This is not just for folks that want to innovate on the interface. This would be enterprise OEMs. This would be cloud service providers and everyone in between.”
Host Jim Greene further explores the advancement of Data Center architecture with his guests, and the customer use cases that might most benefit from the introduction of CXL. Is CXL a memory story, an accelerator story, or both? Baxter and Caward opine on additional advancements the future holds.
To find out more about CXL, revisit Ryan Baxter’s blog post from VMworld 2021 in which he called the technology, “a re-imagined memory and storage hierarchy, and an open, agnostic interconnect,” or see this 2022 blog entry from Micron’s Bryan Madden which also discusses CXL technology and features a fireside chat video with previous Chips Out Loud guest Jeremy Werner discussing its effect on the evolution of the data center [26:34].