Tomshardware Premium
-
PremiumSplave's Cave: Overclocking the Asus RTX 5090 Astral and setting the 3DMark Port Royal world record
-
-
PremiumAmazon launches Trainium3 AI accelerator, competing directly against Blackwell Ultra in FP8 performance
-

Access Bench to view a treasure trove of data and compare products
-
PremiumThe Senate's new SAFE bill is set to curb access to advanced chips to China, but that won't slow down the AI war
-

Leave your feedback on Tom's Hardware Premium
-
Subscribe to Tom's Hardware Premium
- ✓ Premium-exclusive features and interviews
- ✓ Unlock expert news analysis and roadmaps
- ✓ Harness the power of our Bench database
Premium Features
View all-
-
PremiumSplave's Cave: Overclocking the Asus RTX 5090 Astral and setting the 3DMark Port Royal world record
By Allen 'Splave' Golibersuch Published
-
PremiumWhy copper markets are feeling the pinch due to aggressive AI data center expansion
By Chris Stokel-Walker Published
-
PremiumThe web's infrastructure has a concentration problem, exposing us all to crushing outages
By Chris Stokel-Walker Published
-
PremiumExplosive AI buildout brings into question water supply concerns
By Jon Martindale Published
-

American startup Substrate promises 2nm-class chipmaking with particle accelerators, at a tenth of the cost of EUV
By Anton Shilov Published
-
Premium20 years of Noctua: How one brand became a leader in the PC cooling business
By Niels Broekhuijsen Published
-
Roadmaps
View All-
-
PremiumHuawei Ascend NPU roadmap examined — company targets 4 ZettaFLOPS FP4 performance by 2028, amid manufacturing constraints
By Anton Shilov Published
-
PremiumInside the AI accelerator arms race: AMD, Nvidia, and hyperscalers commit to annual releases through the decade
By Anton Shilov Published
-
PremiumThe future of DRAM: From DDR5 advancements to future ICs
By Anton Shilov Published
-
PremiumDesktop GPU roadmap: Nvidia Rubin, AMD UDNA & Intel Xe3 Celestial
By Hassam Nasir Published
-
News AnalysisSilicon Motion announces new devices: PCIe 6.0 SSDs, 256 / 512 TB drives, and next-gen 16K LDPC
By Anton Shilov Published
-
PremiumHBM roadmaps for Micron, Samsung, and SK hynix: To HBM4 and beyond
By Anton Shilov Published
-
PremiumInside the future of 3D NAND: The roadmap to 500 layers
By Anton Shilov Published
-
PremiumNvidia enterprise GPU and CPU roadmaps: Rubin, Rubin Ultra, Feynman, and silicon photonics
By Anton Shilov Published
-
News AnalysisMicron details new U.S. fab projects — HBM production coming to America
By Anton Shilov Published
-

TSMC to spend $42 billion on expansion in 2025 — ambitious plans detail nine production facilities
By Anton Shilov Published
-
News Analysis
View all-
-
PremiumThe Senate's new SAFE bill is set to curb access to advanced chips to China, but that won't slow down the AI war
By Jon Martindale Published
-
PremiumNexperia’s standoff puts a core part of the chip supply chain under strain
By Luke James Published
-
PremiumNvidia’s $2B Synopsys stake strengthens its push into AI-accelerated chip design
By Luke James Published
-
PremiumHBM undergoes major architectural shakeup as TSMC and GUC detail HBM4, HBM4E and C-HBM4E
By Anton Shilov Published
-
PremiumNvidia’s TiDAR experiment could speed up AI token generation using hybrid diffusion decoder
By Luke James Published
-
PremiumChina claims domestically-designed 14nm logic chips can rival Nvidia's 4nm
By Luke James Published
-
Latest
-
-

Musk confirms Tesla AI5 and AI6 will be made at both Samsung and TSMC
By Luke James Published
-
PremiumAMD’s record quarter shows strength in client and AI, as Ryzen leads the charge
By Luke James Published
-
PremiumExplosive AI buildout brings into question water supply concerns
By Jon Martindale Published
-
PremiumNvidia wants China's market share to secure the future of CUDA in the region
By Jon Martindale Published
-
PremiumChina and America's AI war isn't just about compute, it's about energy
By Jon Martindale Published
-

American startup Substrate promises 2nm-class chipmaking with particle accelerators, at a tenth of the cost of EUV
By Anton Shilov Published
-
PremiumBlackwell GPU's exclusion from high-level trade talks highlights deepening AI ecosystem rift between nations
By Luke James Published
-
PremiumOpenAI's Microsoft contract negotiation is a necessary step toward a future IPO
By Jon Martindale Published
-
PremiumMusk says Samsung's Texas fab outclasses TSMC's US-based fabs
By Anton Shilov Published
-
PremiumJet engine shortages threaten AI data center expansion as wait times stretch into 2030 — the rush to power AI buildout continues
By Luke James Published
-
PremiumAMD reportedly establishes $280 million silicon photonics hub in Taiwan — new R&D center could accelerate company's co-packaged optics roadmap
By Anton Shilov Published
-
Premium20 years of Noctua: How one brand became a leader in the PC cooling business
By Niels Broekhuijsen Published
-