Andy Patrizio

Author Archives: Andy Patrizio

Arm creates virtual IoT chips to accelerate development

It can take years for a CPU to go from design to silicon, so Arm is helping developers get a jump on things by putting virtual models of its chip designs in the cloud. The virtual models will allow developers to write and test applications before the actual silicon ships.Dubbed Arm Total Solutions for IoT, the project is a full-stack solution intended for Internet of Things applications and use cases. Arm says the early access for developers, OEMs and service providers, as well as the reduction in product design cycles, could accelerate deployments by up to two years.Arm doesn’t make chips the way Intel and AMD do. It makes designs and licenses them to more than 800 OEMs, which are responsible for everything from embedded devices to servers. Once Arm releases the basic chip design to its partners, the partners then add their own IP to differentiate from the competition, which takes time.To read this article in full, please click here

Intel’s 2nd-gen neuromorphic chip is 10x smarter than the first

Four years after Intel first introduced Loihi, the company’s first neuromorphic chip, the company has released its second generation processor, which Intel says will provide faster processing, greater resource density, and improved power efficiency.CPUs are often called the brains of the computer but aren’t, really, since they process only a handful of tasks at once in a serial manner, nothing like what the brain does automatically to keep you alive. Neuromorphic computing attempts to replicate the functions of the brain by performing numerous tasks simultaneously, with emphasis on perception and decision makingChip shortage will hit hardware buyers for months to years Neuromorphic chips mimic neurological functions through computational “neurons” that communicate with one another. The first generation of Loihi chips had around 128,000 of those digital neurons; the Loihi 2 has more than a million.To read this article in full, please click here

Intel’s second-generation neuromorphic chip is 10x smarter than the first

Four years after Intel first introduced Loihi, the company’s first neuromorphic chip, the company has released its second generation processor, which Intel says will provide faster processing, greater resource density, and improved power efficiency.CPUs are often called the brains of the computer but aren’t, really, since they process only a handful of tasks at once in a serial manner, nothing like what the brain does automatically to keep you alive. Neuromorphic computing attempts to replicate the functions of the brain by performing numerous tasks simultaneously, with emphasis on perception and decision makingChip shortage will hit hardware buyers for months to years Neuromorphic chips mimic neurological functions through computational “neurons” that communicate with one another. The first generation of Loihi chips had around 128,000 of those digital neurons; the Loihi 2 has more than a million.To read this article in full, please click here

A dive into Kyndryl, IBM’s managed-services spin-off

Thanks to  a US Securities and Exchange Commission filing new details have emerged about Kyndryl, the IBM spin-off of its managed-infrastructure services unit into a separately traded public company.Kyndryl does exactly what the managed-infrastructure services unit of IBM’s Global Technology Services segment does: manage enterprises IT infrastructure, whether it comes from IBM or another vendor. That’s a challenge for Kyndryl because it has to deal with the trend toward cloud services and against on-premises infrastructure.The split is expected to be complete by the end of 2021, and when it was announced last year, IBM CEO Arvind Krishna said IBM will focus on open hybrid-cloud and AI capabilities while Kyndryl would focus on modernizing customer infrastructure.To read this article in full, please click here

Marvell announces some very smart SmartNIC processors

Marvell has begun to sample the Octeon 10, a server microprocessor aimed at intelligent network management that has up to 24 Arm-compatible cores, making it as powerful as any server processor.Marvell refers to the Octeon processor line as data processing units (DPUs). They are designed to run high-throughput data in the cloud and on-premises. The DPU is more commonly called the SmartNIC because it can offload non-computational tasks from the CPU like network packet processing, data encryption and compression. That frees up CPU cores to run general-purpose applications.The Octeon 10 has a few firsts. It's the first processor made by TSMCs 5nm manufacturing process and the first processor to feature Arm’s Neoverse N2 core. The N2 core uses the new Armv9 architecture that the company claims can deliver 40% more single-threaded performance for a variety of workloads vs. the N1, but still retains the same level of power and area efficiency as N1.To read this article in full, please click here

The Dell/VMware relationship remains strong despite split

VMware is in the process of spinning out from Dell Technologies, but the working relationship remains as strong as ever with a bunch of announcements from VMworld.All told, the pair made four significant announcements at the show, the first being that VMware Cloud will be sold on systems acquired through Dell's Apex pay-as-you-go program. The new Apex offering gives customers the ability to move workloads across multiple cloud environments and scale resources quickly with predictable pricing and costs.The new offering combines Dell’s hyperconverged infrastructure VxRail with VMware Cloud, VMware Tanzu for building cloud-native applications, and VMware HCX for application migration. Businesses can deploy the offering in their data center, at an edge location or a colocation facility with partners like Equinix.To read this article in full, please click here

The Dell/VMware relationship remains strong despite split

VMware is in the process of spinning out from Dell Technologies, but the working relationship remains as strong as ever with a bunch of announcements from VMworld.All told, the pair made four significant announcements at the show, the first being that VMware Cloud will be sold on systems acquired through Dell's Apex pay-as-you-go program. The new Apex offering gives customers the ability to move workloads across multiple cloud environments and scale resources quickly with predictable pricing and costs.The new offering combines Dell’s hyperconverged infrastructure VxRail with VMware Cloud, VMware Tanzu for building cloud-native applications, and VMware HCX for application migration. Businesses can deploy the offering in their data center, at an edge location or a colocation facility with partners like Equinix.To read this article in full, please click here

Very quietly, Oracle ships new Exadata servers

You have to hand it to Larry Ellison, he is persistent. Or maybe he just doesn’t know when to give up. Either way, Oracle has shipped the latest in its Exadata server appliances, making some pronounced boosts in performance.Exadata was the old Sun Microsystems hardware Oracle inherited when it bought Sun in 2010. It has since discontinued Sun’s SPARC processor but soldiered on with servers running x86-based processors, all of them Intel despite AMD’s surging acceptance in the enterprise.When Oracle bought Sun in 2010, it was made clear they had no interest in low-end, mass market servers. In that regard, the Oracle Exadata X9M platforms deliver. The new Exadata X9M offerings, designed entirely around Oracle’s database software, include Oracle Exadata Database Machine X9M and Exadata [email protected] X9M, which Oracle says is the only platform that runs Oracle Autonomous Database in customer data centers.To read this article in full, please click here

Very quietly, Oracle ships new Exadata servers

You have to hand it to Larry Ellison, he is persistent. Or maybe he just doesn’t know when to give up. Either way, Oracle has shipped the latest in its Exadata server appliances, making some pronounced boosts in performance.Exadata was the old Sun Microsystems hardware Oracle inherited when it bought Sun in 2010. It has since discontinued Sun’s SPARC processor but soldiered on with servers running x86-based processors, all of them Intel despite AMD’s surging acceptance in the enterprise.When Oracle bought Sun in 2010, it was made clear they had no interest in low-end, mass market servers. In that regard, the Oracle Exadata X9M platforms deliver. The new Exadata X9M offerings, designed entirely around Oracle’s database software, include Oracle Exadata Database Machine X9M and Exadata [email protected] X9M, which Oracle says is the only platform that runs Oracle Autonomous Database in customer data centers.To read this article in full, please click here

Kioxia seeks to make the SSD more programmable

NAND flash maker Kioxia has expanded its Software-Enabled Flash technology to bring a greater degree of programmability to NAND storage. The move will benefit hyperscalers the most but will have benefits for enterprises and SMBs as well.Kioxia (formerly Toshiba) first introduced SEF last year. It’s an open-source API that operates as a new kind of hardware flash controller to offload some functions to a controller, thus freeing up the CPU, while allowing large data-center environments to manage at scale.Because the API is open source, competitors in the flash space can adopt the API and customize it for their hardware. Hyperscalers think about SSDs in terms of deploying and serving workloads at scale. Kioxia notes that cloud providers often have different types of drives they deploy for different use cases, like block storage versus file storage or ZNS.To read this article in full, please click here

Kioxia seeks to make the SSD more programmable

NAND flash maker Kioxia has expanded its Software-Enabled Flash technology to bring a greater degree of programmability to NAND storage. The move will benefit hyperscalers the most but will have benefits for enterprises and SMBs as well.Kioxia (formerly Toshiba) first introduced SEF last year. It’s an open-source API that operates as a new kind of hardware flash controller to offload some functions to a controller, thus freeing up the CPU, while allowing large data-center environments to manage at scale.Because the API is open source, competitors in the flash space can adopt the API and customize it for their hardware. Hyperscalers think about SSDs in terms of deploying and serving workloads at scale. Kioxia notes that cloud providers often have different types of drives they deploy for different use cases, like block storage versus file storage or ZNS.To read this article in full, please click here

HPE expands GreenLake services into new markets

Hewlett Packard Enterprise has announced three new cloud-related offerings to more effectively protect data and make it more available to analytics.The first is called HPE GreenLake for Data Protection that relies on the company’s on-premises Greenlake data-center hardware sold on a pay-per-use model rather than purchasing everything upfront.The service includes HPE Backup and Recovery Service for VMware and GreenLake for Disaster Recovery.How to choose the best NVMe storage array The backup and recovery service allows enterprises to back up on-premises virtual machines to the public cloud. This is purely a service with no hardware purchase requirements. Customers can recover instantly on-prem, and it is particularly aimed at protecting against ransomware attacks.To read this article in full, please click here

HPE expands GreenLake services into new markets

Hewlett Packard Enterprise has announced three new cloud-related offerings to more effectively protect data and make it more available to analytics.The first is called HPE GreenLake for Data Protection that relies on the company’s on-premises Greenlake data-center hardware sold on a pay-per-use model rather than purchasing everything upfront.The service includes HPE Backup and Recovery Service for VMware and GreenLake for Disaster Recovery.How to choose the best NVMe storage array The backup and recovery service allows enterprises to back up on-premises virtual machines to the public cloud. This is purely a service with no hardware purchase requirements. Customers can recover instantly on-prem, and it is particularly aimed at protecting against ransomware attacks.To read this article in full, please click here

Gigabyte and CoolIT partner for liquid cooled servers

Gigabyte Technology isn’t the first name that comes to mind in data-center hardware. It’s better known as a consumer player, but it is a significant server player none the less, making server motherboards on par with other top names like Supermicro.Now the company has teamed with CoolIT Systems to provide two high-density servers equipped with liquid-cooling technology.The servers, H262-ZL0 and H262-ZL2, are equipped with direct liquid cooling for CPUs designed to support the high-performing but super-hot 280 watt AMD EPYC 7003 (Milan) processors.The servers, based on the company's H262-Z6x family of air-cooled servers, are hyperconverged and very dense, targeting HPC, HCI, in-memory-computing, and scientific-research markets. They both pack four nodes with two sockets each and eight DIMM slots per node in a 2U form factor. To read this article in full, please click here

Gigabyte and CoolIT partner for liquid cooled servers

Gigabyte Technology isn’t the first name that comes to mind in data-center hardware. It’s better known as a consumer player, but it is a significant server player none the less, making server motherboards on par with other top names like Supermicro.Now the company has teamed with CoolIT Systems to provide two high-density servers equipped with liquid-cooling technology.The servers, H262-ZL0 and H262-ZL2, are equipped with direct liquid cooling for CPUs designed to support the high-performing but super-hot 280 watt AMD EPYC 7003 (Milan) processors.The servers, based on the company's H262-Z6x family of air-cooled servers, are hyperconverged and very dense, targeting HPC, HCI, in-memory-computing, and scientific-research markets. They both pack four nodes with two sockets each and eight DIMM slots per node in a 2U form factor. To read this article in full, please click here

IBM and Atos partner to help financial businesses migrate to the cloud

IBM has partnered with Atos, the closest thing it has to an equal in Europe, to help boost the digital transformation and cloud migration initiatives for banks and insurance companies in a project called Atos Cloud Centre of Excellence.Finance is one of the most regulated industries and, therefore, one of the most reluctant to move to the cloud. The center’s goal is to increase security and regulatory compliance for financial services companies around the world that wish to move their workloads to the cloud.Atos and IBM said the center will provide technology and financial services expertise for clients, backed by dedicated Atos professionals who are trained on IBM Cloud for Financial Services, IBM Cloud Paks and Red Hat OpenShift.To read this article in full, please click here

Intel: Under attack, fighting back on many fronts

At first glance, Intel doesn’t look like a company under siege. In its last fiscal year, it recorded $77.8 billion in sales and $20 billion in profit. Its market capitalization is $220 billion as of mid-September 2021.And yet it is. When you’re the leader, all your competition is gunning for you. Intel is wrestling with a loss of leadership in manufacturing and process nodes, it’s losing share to a very resurgent AMD, an unrelenting Nvidia is challenging Intel for AI dominance, the Atom processor failed spectacularly against Arm in the mobile market, and it’s on its third CEO in three years. More about Intel: A closer look at two newly announced Intel chips Intel shifts to a multiarchitecture model Intel revises its chip terminology and branding CEO Gelsinger shakes up Intel But Intel revels in the competition. “Our success in so many markets makes us targets for lots of companies,” said Lisa Spelman, corporate vice president and general manager of the Xeon and memory group. “So it’s not a surprise that we have competitors that want a piece of that.” To read this article in full, please click here

Intel: Under attack, fighting back on many fronts

At first glance, Intel doesn’t look like a company under siege. In its last fiscal year, it recorded $77.8 billion in sales and $20 billion in profit. Its market capitalization is $220 billion as of mid-September 2021.And yet it is. When you’re the leader, all your competition is gunning for you. Intel is wrestling with a loss of leadership in manufacturing and process nodes, it’s losing share to a very resurgent AMD, an unrelenting Nvidia is challenging Intel for AI dominance, the Atom processor failed spectacularly against Arm in the mobile market, and it’s on its third CEO in three years. More about Intel: A closer look at two newly announced Intel chips Intel shifts to a multiarchitecture model Intel revises its chip terminology and branding CEO Gelsinger shakes up Intel But Intel revels in the competition. “Our success in so many markets makes us targets for lots of companies,” said Lisa Spelman, corporate vice president and general manager of the Xeon and memory group. “So it’s not a surprise that we have competitors that want a piece of that.” To read this article in full, please click here

Lenovo extends TruScale as-a-service model to its entire portfolio

Lenovo is expanding its TruScale pay-per-use model to cover all its data-center products—servers, storage—and client-side devices—laptops, tablets.This transition to a fully integrated, end-to-end, as-a-service model is part of the company’s “One Lenovo” strategy of providing its entire portfolio of clients and servers as a fully managed, on-premises cloud environment through TruScale leasing.One Lenovo simply means laptops and desktops will be sold along with data-center products together all under one sales program. Lenovo will launch a new channel program in 2022 to encompass the One Lenovo strategy.The everything-as-a-service announcement came at the company’s virtual Lenovo Tech World 2021 eventTo read this article in full, please click here

Lenovo extends TruScale as-a-service model to its entire portfolio

Lenovo is expanding its TruScale pay-per-use model to cover all its data-center products—servers, storage—and client-side devices—laptops, tablets.This transition to a fully integrated, end-to-end, as-a-service model is part of the company’s “One Lenovo” strategy of providing its entire portfolio of clients and servers as a fully managed, on-premises cloud environment through TruScale leasing.One Lenovo simply means laptops and desktops will be sold along with data-center products together all under one sales program. Lenovo will launch a new channel program in 2022 to encompass the One Lenovo strategy.The everything-as-a-service announcement came at the company’s virtual Lenovo Tech World 2021 eventTo read this article in full, please click here

1 2 3 65