France raises solar-plus-storage grid-connection cap to 25 MW by year-end 2026
France plans to lift the 17 MW injection ceiling for hybrid PV-BESS projects up to 25 MW installed by end-2026, alongside smart-meter and heat-pump subsidy changes.
Component sourcing news
What changed today, why it matters for your BOM, and what to watch. Every factual claim is traceable to the source it came from; our own analysis is labelled and kept separate.
EN · single language Editorial policy
Source sanity (S1)
Fetched in last 24 h100
Clean prose8743
Rejected before any model call1018
Bad-source rate10%
Gate (spec v2.9)
Pass133
Hold (human review)604
Reject203
Release rate14%
Publishing volume
Today / daily limit1 / 10
Sensitive-topic escalations0
Auto-publishOff (shadow period)
Intel Razor Lake-AX targets 700 TOPS: what 3.88x Panther Lake means for mobile AI BOMs
Intel said Razor Lake-AX will deliver 700 TOPS at INT8/FP8, 3.88x Panther Lake's 180 TOPS, using a Xe3 iGPU up to 32 cores and bringing back on-package LPDDR5X/LPDDR6 memory.
What Infineon's 200mm SiC wafer deal with Wolfspeed means for EV and AI power BOMs
Infineon locks in multi-year 200mm SiC substrate supply from Wolfspeed, supporting its Kulim fab ramp and broadening multi-source sourcing for SiC power devices.
·4 min read
Intel Nova Lake-HX and -H die shots surface: six-tile mobile package mapped
Bare-die photos of Nova Lake-HX and -H lay out a six-tile mobile package using TSMC N2P, Intel 3-E and Intel 18A-P nodes, with new core counts and package sizes.
TI RES80A-Q1 divider signals where EV-grade passives are headed for HV BMS
TI rolls an AEC-Q200 50 MΩ divider rated to 2,000 V DC for EV battery management, underscoring how automotive specs now reshape passive-component sourcing.
·4 min read
Modded Preset C Brings Up to 35% Faster DLSS Frame Generation to RTX 30-Series GPUs
dashdogy's Universal MFG Unlocker v1.5.0 enables NVIDIA's new Preset C on Ampere cards, but RTX 30 numbers and image-quality trade-offs remain unverified.
SP Industrial launches six enterprise SSDs: PCIe Gen5 E1.S to 7.6 TB and 2.5" SATA for AI racks
SP Industrial adds six 3D eTLC enterprise SSDs across E1.S, U.2, M.2 2280, and 2.5" SATA, with a 7.6 TB PCIe Gen5 E1.S and dual 1/3 DWPD endurance tiers.
Hybrid bonding process maturity: what 1um pitch and 450nm demos mean for HBM and chiplet roadmaps
A 3D InCites technical review of hybrid bonding defects, CMP, warpage, and annealing shows Applied Materials at 450nm pitch with 98% yield and imec at 200nm pitch on 300mm W2W.
What NVIDIA's 165W RTX PRO 4500 Blackwell means for 2U server inference BOMs and L4 refresh cycles
NVIDIA's RTX PRO 4500 Blackwell Server Edition matches L40S throughput at less than half the power in a single-slot 165W card, reshaping edge and mainstream inference BOMs.
·6 min read
What ChipAgents and NVIDIA's Build-A-Chip-Bench means for autonomous RTL-to-GDS tooling
ChipAgents and NVIDIA launch Build-A-Chip-Bench to score end-to-end digital chip design. Only ~10% of agent runs hit PPA targets, exposing the gap to tapeout-ready AI.
Supermicro's OCP Push: 48V ORv3 Racks, 250 kW CDU, and 102.4 Tbps CPO Switch Reshape AI Datacenter BOMs
Supermicro bundles ORv3 racks, a 250 kW in-rack CDU, and a 102.4 Tbps CPO switch into one OCP-accepted stack, with implications for AI rack sourcing.
·6 min read
Aetina AIE-KT78/68 brings Jetson Thor Blackwell compute to deterministic edge AI control
Aetina launches AIE-KT78 and AIE-KT68 edge AI systems on NVIDIA Jetson Thor, with QSFP28, 8× GMSL2, and an isolated EtherCAT port for cobots, humanoids, and AMRs.
·4 min read
About the data on this page. Charts on this site are native SVG data visualisations built from verified data held on this site; there are no decorative illustrations and no hot-linked third-party images.
Get the daily brief
One email per publishing day. You can also add XAIPART as a preferred source in Google.