Andy Patrizio is a freelance technology writer based in Orange County, California. He's written for a variety of publications, ranging from Tom's Guide to Wired to Dr. Dobbs Journal, and has been on staff at IT publications like InternetNews, PC Week and InformationWeek.
Once known for its gaming cards, the GPU giant with a commanding hold on the AI chip market dominated 2024.
The company says its Smart Retimer SoC can maintain signal strength through thin wiring and reduced power.
Amazon is seeing success with its Arm-based Graviton CPUs, which shows the viability of hyperscalers making their own custom silicon.
Although it isn’t sharing the designs, Microsoft is influencing other data center operators with its zero water consumption cooling system.
The coming year will see continued interest in familiar data center topics, including the ROI of AI, power constraints, liquid cooling, and Intel leadership.
Verizon will provide on-premises edge connectivity for AI apps and other workloads.
Equinix stopped commercial sales of its Metal service to focus on colocation, interconnection and hyperscale offerings.
New reference design addresses sustainability issues surrounding AI.
Legacy copper services are no longer meeting customer needs for speed, reliability and always on-connectivity, AT&T says.
Already struggling, Intel now faces even more challenges.
MRDIMM promises faster memory with no hardware or software changes.
New PowerEdge servers are aimed at accelerating enterprise AI adoption.
Is one massive chip more powerful than hundreds of smaller chips? Cerabras would argue yes.
What do these five urban regions in the US have in common? Readily available power.
The six-chip configuration is meant for space-constrained locations.
Reports say racks must be redesigned to accommodate the heat generated.
Micron is also launching a high-capacity SSD. Both vendors promise greater storage density and lower power consumption.
During the campaign, President-elect Donald Trump hinted at trying to repeal the CHIPS Act, but reality contradicts him.
So far, the Supermicro saga is tied to accounting, not products. How deep it goes remains to be seen.
The reference designs are meant to ease and accelerate deployment of AI-oriented data centers.
An interesting question is raised by an anonymous source on X, but there is nothing to substantiate it.
Liquid cooling continues to gain legitimacy. HPE's architecture doesn't rely on fans, while Supermicro's rack solution is specifically designed for Nvidia‘s Blackwell platform.
New storage system features multiple Nvidia GPUs for high-speed throughput.
The two competitors have become allies in a bid to stave off Arm.
Jointly designed architecture is intended for GPU-loaded AI factories, which will generate tremendous amounts of heat.
HPE unveils one server, and Dell launches several compute and storage products.
New Epyc chip to offer scale-up performance or scale-out capacity for a wide range of data center workloads.
Xeon 6 delivers twice the performance of its predecessor, Intel claims.
Altera CEO Sandra Rivera shares ‘big, audacious, ambitious goal’ to dominate FPGA market.
An employee may have inadvertently tipped a major reboot of Intel’s x86 architecture.
SmartNICs are designed around CPU-to-CPU communication, making them less optimal for GPUs, the company argues.
Microsoft and BlackRock aim to launch a $30 billion fund for AI infrastructure with future plans for up to $100 billion in total investment potential.
The news isn’t all good on the foundry front, but it is a step in the right direction for Intel.
The antitrust investigation appears to have no public complainants or perceived victims.
The next generation of processors from Ampere will have up to 512 cores and is designed to fit into existing air-cooled data centers.
Intel’s photonics group demonstrated a new high-speed optical interconnect chiplet co-packaged with an Intel CPU and running live data at the show.
AMD says the $4.9 billion deal to acquire the server maker is 'our next major step' in AI.
Softness in Intel's server business allowed AMD to make progress, but the gap remains very large.
Concerns about the new direction VMware is taking are driving some enterprise customers to consider alternative platforms from vendors such as Scale Computing, Nutanix and Oxide Computer.
Among cloud infrastructure providers, Oracle is finally breaking out, although Amazon, Microsoft and Google still maintain strong leads. Meanwhile, on-premises data center capacity remains stable.
The damage is contained to high-end 13th- and 14th-generation desktop CPUs processors at gamers.
Shipments of Nvidia’s next-generation Blackwell architecture could be delayed for three months or more, according to published reports.
NeuReality claims its AI appliance can significantly cut costs and energy use in AI data centers.
Diversity of AI workloads means there’s a market for many types of chips.
British chip designer Graphcore has been acquired by Japan's SoftBank Group, which retains significant control over Arm Holdings, its earlier semiconductor investment.
The Dell PowerEdge XE9680 ships with Instinct MI300X accelerators, AMD's high-end GPU accelerators that are designed to compete with Nvidia Hopper processors.
As high-density data centers continue to add AI workloads, there’s growing interest in liquid cooling, thanks to its ability to transfer heat more efficiently than air cooling.
AWS says Graviton4 is its most powerful and energy efficient processor, suited for a broad range of workloads running on Amazon EC2.
Updated infrastructure-as-code management capabilities and expanded SLAs are among the new features from Pure Storage.
‘I'm not sure yet whether I'm going to regret this or not,' said Nvidia CEO Jensen Huang as he revealed 2026 plans for the company’s Rubin GPU platform.
Sponsored Links