Dell XE9640 Deep Dive, Edge Servers, 50 DWPD SSDs, More...

Dell PowerEdge XE9640 Liquid-Cooled GPU Server Deep Dive

The Dell PowerEdge XE9640 is a 4x GPU-accelerated rack-mount server capable of delivering AI power in a power-efficient way, thanks to liquid cooling. The XE9640 was announced during?SC22?along with the XE8640 and one of our favorites, the 8-way?XE9680?GPU server. Today, the XE9640 is generally available, and we’re taking a deep dive into the underlying hardware. Check out the details for the?Dell PowerEdge XE9640.

Intel’s License Free GPUs for VMware Horizon Changes the VDI Game

GPUs have become a mainstay in modern virtual desktop infrastructure (VDI) implementations, and rightfully so. They greatly increase the power of virtual desktops and expand use cases. However, one of the biggest pain points for VDI implementation is the complexity of licensing. Get the scoop on?Intel's license-free GPUs.

Lenovo ThinkSystem Towers Are Cost-Effective SMB Solutions

Going Cloud has advantages, but it also means perpetual monthly costs. Lenovo’s ThinkSystem ST50 and ST250 are cost-effective tower servers that can be more economical for small businesses and remote offices. Read about the cost-effective?Lenovo ThinkSystem Towers.

Supermicro X13 Edge Servers Unveiled for AI & Telco

Supermicro has announced the expansion of its line of purpose-built servers for Edge AI and Telco workloads. This announcement is highlighted by the newest addition to their X13 Edge series, the company’s line of edge servers aimed at a broad range of use cases, from telecommunications to enterprise applications. Read about?Supermicro X13 Edge Servers.

Lenovo?Unveils Next-Gen Edge AI Solutions to Simplify Deployment, Accelerate Business Transformation

Lenovo is stepping up its efforts to bring AI to the edge with a new suite of services and solutions to enable businesses to deploy advanced computing capabilities remotely. The company has launched Lenovo TruScale for Edge and AI, a pay-as-you-go service model designed to accelerate AI readiness directly at data creation points. Read the details on?Lenovo's Next-Gen Edge AI Solutions.

Solidigm P5810 Write-Intensive Enterprise SSD Announced

Solidigm has unveiled the D7-P5810 at the recent Tech Storage Field Day, designed for high-endurance and extreme write-intensive workloads. Built on the company’s proven 144-layer Single-Level Cell (SLC) 3D NAND technology, the PCIe Gen4 drive thrives in handling write-heavy tasks such as caching, high-performance computing (HPC), and logging. Read more details on the?Solidigm D7-P5810.

Dell APEX Cloud Platform For Microsoft Azure Announced

Dell announced the availability of Dell APEX Cloud Platform for Microsoft Azure, designed for IT professionals pursuing advanced hybrid cloud solutions. This launch represents a significant milestone in Dell’s strategic collaboration with Microsoft, focusing on providing seamless cloud integration. Read more on?APEX Cloud Platform for Microsoft Azure.

Infinidat SSA Express Software And Expanded InfiniBox Platform Announced

Infinidat introduced two upgrades that significantly enhance its all-flash capabilities and storage capacity for enterprise-level storage needs. Infinidat unveiled SSA Express Software for its InfiniBox platform, which seamlessly integrates a high-speed, all-flash engine directly into the hybrid array.?? Get the details on?Infinidat's SSA Express Software and InfiniBox Platform.

Sabrent Rocket nano XTRM Review

The Sabrent Rocket nano XTRM (SB-XTMN) is an external SSD available in a 1TB capacity model, designed to offer both versatility and portability. Check out the review for the?Sabrent Rocket nano XTRM.

Samsung LPCAMM Promises a Leap Forward in Memory Module Technology

Samsung Electronics has announced the creation of the industry’s first Low Power Compression Attached Memory Module (LPCAMM). This groundbreaking technology has the potential to redefine the DRAM market for PCs, laptops, and possibly even data centers. Read more about the?Samsung Low Power Compression Attached Memory Module(LPCAMM).

OCI Compute BM.GPU.H100.8 Brings 8 NVIDIA H100 GPUs to the Cloud

Oracle Cloud Infrastructure (OCI) has launched a new bare metal compute shape that packages eight NVIDIA H100 GPUs, twin Intel 56-core Xeon CPUs, 2TB of DDR5, 16x NVMe SSDs, and 400Gb networking into an AI dream in the cloud. OCI Compute BM.GPU.H100.8 is generally available now. Get the details on?Oracle Cloud Infrastructure's new compute shape.

Bluetti PV350 Solar Panel Review

The Bluetti PV350 is a portable 350-watt folding solar panel designed to extend the runtime or enable the charging of portable power stations out in the field. Keeping devices operational without grid power can be challenging with remotely deployed hardware. With a 23.4% efficiency, rugged design, and water-resistant rating, the Bluetti PV350 is geared to provide solar power in nearly all remote environments. Get the details for the?Bluetti PV350 Solar Panel?in this review.

Podcast #123: Cloud-Based GPUs For AI

This podcast features a conversation between Brian and Jeffrey Gregor (he prefers Gregor). ?Gregor is the General Manager for OVHcloud? US, a leading cloud provider with a global presence. Prior to joining OVHcloud, Gregor was a Navy Officer, ?a highly decorated Navy Officer, with over 15 years serving the US.

OVHcloud was established in 1999, with some borrowed funds and hardware, ?as a web-hosting startup in northern France and has grown steadily, expanding across Europe. Following the expansion into Canada in 2015, ?OVHcloud opened offices in the US. Gregor joined OVHcloud when the company opened offices in northern Virginia. And the expansion continues.

We recently posted a deep-dive review for OVHcloud US, focusing on GPUs in the cloud and how to approach?cloud training for AI.?The review is worth the time as it provides background on OVH, how we performed our tests, and the results.

Full Pod on YouTube

Social Highlights

How about GPU-assisted VDI without breaking the bank? Intel’s Flex Series GPUs offer an affordable, high-performance solution, eliminating extra licensing costs and complexities associated with hardware-assisted GPU virtualization. Flex GPUs are available in both a 75W and a 150W single-width PCIe add-in card offering 12GB or 16GB of GDDR, respectively. Here’s the first demo of this capability from Intel Innovation. There’s still much more to understand but if Intel is trimming cost from GPU-enabled VDI, this could be a massive entry for the Flex GPUs.

Your weekend data center ASMR is here. 26 Solidigm P5316 30.72TB SSDs slipping into the toolless drive carriers for the Tyan Transport HX FT65T-B8050 single proc AMD server, which incidentally has a 96-core AMD Genoa-X inside, alongside Kingston DDR5. We did, of course, go RAID0 and will be sharing this storage with a workstation for the most epic ~740TB network share ever.

A bit ridiculous at first sight, we have a feeling the Dell UltraSharp 43 4K USB-C Hub Monitor - U4323QE won’t be leaving the lab any time soon.

We’ve been working with this box a lot lately, Supermicro is one of the few producing a dense E1.S server. Right now we’re using the server with 24x KIOXIA SSDs with a dual purpose. First, it’s providing high-speed storage to three workstations over 100GbE, giving them access to more data for training AI models. Once models are ready, we can inference with the T4 in the storage server itself, no need to move data.