Hardware
GPUs, workstations, AI accelerators and chip industry news that matter to 3D and AI work.
AI imageSupermicro Petascale Server Launches with PCIe 6 and AMD EPYC 9006
Supermicro announced the Petascale all-flash server with AMD EPYC 9006 processors and PCIe 6 support. Shipping has begun for the 2U system, which supports up to 3.93 PB of capacity with 32 E3.S NVMe SSDs.

NVIDIA RTX Spark Compact Desktop PCs Go on Sale
NVIDIA announced that compact RTX Spark desktop PCs from ASUS, Dell, and MSI went on sale on 18 November 2026. The systems target local AI, creative applications, and 3D workflows.
AI imageVertiv Expands Its 800 VDC Portfolio Across Racks, AI Pods and Data Halls
On November 17, 2026, Vertiv announced an expansion of its 800 VDC portfolio to cover rack, AI compute pod and data hall levels. The company says the rack architecture can preserve existing AC infrastructure.
AI imageNVIDIA Vera Rubin NVL72: ASUS ramps up shipments in Q4 2026
ASUS announced on November 17 that it is ramping up volume shipments of NVIDIA Vera Rubin NVL72 systems in Q4 2026. At SC26, the company is showcasing AI Factory production and facility simulation solutions.
AI imageNCSA receives $334 million grant for Garnet supercomputer
Schmidt Sciences has awarded NCSA a $334 million grant for the Garnet research supercomputer, which will be built in Illinois. The system, initially equipped with Blackwell GPUs, is expected to come online in mid-2027.
AI imageTOP500 November 2026: Top 10 unchanged, record-low number of new systems
The order of the top 10 supercomputers remained unchanged in TOP500’s November 2026 list, announced on 16 November. Just 27 new systems made the list—the lowest number in TOP500 history.

NVIDIA Photonics CPO powers Lambda’s 10,000-GPU cluster
NVIDIA and cloud provider Lambda are running an AI cluster with more than 10,000 GPUs using Photonics co-packaged optics switches. NVIDIA reports gains in power efficiency and reliability in its tests.

NVIDIA DGX Spark HPC Developer Kit Brings a Four-Node Cluster to the Desktop
NVIDIA unveiled the HPC Developer Kit at SC26, bringing together four DGX Spark systems. According to the company, the cluster offers 512 GB of memory and 20 TB of NVMe storage.
AI imageNVIDIA Vera announced: 192-CPU HPC platform targets scientific simulations
NVIDIA announced a liquid-cooled HPC platform combining 192 Vera CPUs with an InfiniBand network. The system targets scientific simulations, with hardware expected in early 2027.
AI imageNVIDIA Reports MLPerf Leadership for Vera Rubin
NVIDIA says Vera Rubin submitted results in every MLPerf Training v6.1 test and stood out in MLPerf Endpoints agent benchmarks. The Kimi K3 comparison is still awaiting validation.
AI imageNVIDIA Rubin HPC announced: More than 50,000 CUDA cores in one GPU
NVIDIA announced the Rubin HPC GPU at SC26 for science, engineering and analytics workloads. The product will feature more than 50,000 CUDA cores in a single GPU and is expected to arrive in 2028.
AI imageUS plans export licenses for remote access to AI chips
The US administration is preparing a temporary rule that would require licenses to provide restricted countries with remote computing power from AI chips and high-capacity data centers. The scope also includes chip transfers to data centers in certain countries.
AI imageNVIDIA’s 260,000-GPU plan: South Korea to monitor deliveries
South Korea said NVIDIA will continue its 260,000-GPU supply plan and joint AI projects following Jensen Huang’s departure from the CEO role. Delivery progress will be closely monitored.
AI imageCalls for New Restrictions on AI Chip Exports to China
Following developments involving former Nvidia CEO Jensen Huang, the U.S. Congress is once again debating new restrictions on AI chip exports to China. No regulation has been announced as having taken effect.
AI imageAMD EPYC Embedded 9006 Unveiled: 256 Cores and SP7 Platform
AMD has announced the Zen 6-based EPYC Embedded 9006 processors for embedded systems, part of the Venice family. The series scales up to 256 cores and supports PCIe Gen6 and DDR5-8000 on the SP7 platform.
AI imageMarvell to showcase CXL memory and photonic connectivity at SC26
Marvell announced it will showcase CXL memory, photonic connectivity and networking technologies at SC26, taking place in Chicago from 15–20 November 2026. Demonstrations will include memory modules, switches and storage solutions.
AI imageStaera unveils Lightwire: Data moves between AI chips through light
Staera Semiconductor has unveiled Lightwire, an LED-based optical interconnect technology built into chips, and announced $4.6 million in funding for its development. Commercial launch is targeted for 2030.
AI imageHyperAccel to Showcase Eight-LPU AI Server at SC26
South Korean company HyperAccel will showcase its server with eight LPUs and a live LLM inference demo running on operational hardware in Korea at SC26, November 17–19. The company claims energy and cost efficiency based on its own tests.
AI imageAMD's Siren Halo APU Rumored to Feature Dedicated GDDR7 Memory
According to hardware leaks, AMD's rumored Siren Halo APU may use LPDDR6 for system memory alongside dedicated GDDR7 memory for its integrated RDNA 5 GPU. The information has not yet been confirmed.
AI imageAMD Expands Instinct MI400 Series with MI450X; Sets MLPerf Record with MI455X
On 12 November 2026, AMD expanded its Instinct MI400 series GPU portfolio with the MI430X and new MI450X models, while announcing MLPerf inference results for the MI455X that it says outperform Nvidia Vera Rubin.
AI imagePenguin Solutions, DDR4 destekli CXL bellek kartı SMART CXL-8F2W-D4'ü tanıttı
Penguin Solutions, yapay zekâ veri merkezlerindeki bellek darboğazına çözüm olarak DDR4 belleği CXL üzerinden kullanılabilir kılan sekiz DIMM'li SMART CXL-8F2W-D4 kartını tanıttı.
AI imageKIOXIA and Dell Build 157 PB Storage Solution in a Single 42U Rack
KIOXIA and Dell created an object storage solution with 157 PB of raw flash capacity in a standard 42U server rack, using 640 245.76 TB SSDs. The system is designed for AI and data-intensive workloads.
AI imageMetaRoCE Showcased at OCP Summit 2026 with an Open RDMA Approach
Meta presented its MetaRoCE approach at OCP Summit 2026 and showcased a network card in the OCP NIC 3.0 form factor at its booth. The presentation highlighted multi-path networking, congestion control, and resilience to packet loss.
AI imageFirst photos of Intel Xeon 7 Diamond Rapids’ 256-core package
The first photos of Intel’s 256-core Xeon 7 Diamond Rapids processor package were published after it was displayed at the Tencent Global Digital Ecosystem Conference in Shenzhen. The package is designed for the LGA 9324 socket on the Oak Stream platform.
About this category
The Hardware category covers hardware news that directly affects 3D, architectural visualization, and AI work: graphics cards from NVIDIA, AMD, and Intel; professional workstation cards; processors; render workstations; AI accelerators; and AI PCs. We share new products’ VRAM, performance, and pricing details, and explain what they mean for rendering and running AI models locally.
Industry developments such as chip production, supply shortages, export restrictions, and price changes also feature here. A factory shutdown or a new restriction, for example, could affect the price and availability of the graphics card you plan to buy a few months from now. We compile our coverage from official manufacturer announcements and reliable hardware publications.
Rather than focusing on gaming benchmarks, we prioritize real workloads: render times in V-Ray, Blender, D5, or Unreal Engine; memory capacity for large scenes; and the hardware required to run AI models locally. This helps you make upgrade decisions based on your own workflow.