First DDR5 Kit with EXPO ULL just arrived. Let’s see how the fine tuned sub-timings compared to the manuell optimization will work out.
"Sub-1nm" chip technology by @IBMResearch. IBM nanostack architecture: Two transistors stacked and bonded together. Each transistor consists of three nanosheets with a thickness of 5 nm thick or the equivalent to about 15 rows of atoms. newsroom.ibm.com/2026-06-25-i...
My colleague did review the Noctua NL-LC1, the first all-in-one cooler from the air cooling and fan specialist. In a nutshell: Impressive! www.hardwareluxx.de/index.php/ar...
The SiPearl Rhea1 is now in a 12-week functional bring-up phase. It's using 61 billion transistors and is manufactured in N6 at TSMC. sipearl.com/rhea1
It's not a new pictures, but I can't stress enough just how amazing IBM's 300 mm wafer with quantum chips looks!
Does it show how old I am if I like a design like this?
Today’s coverage focuses on 12V-2x6 power delivery: First, we’ll take a look at Sapphire #PhantomLink ecosystem. This BTF-compatible connector is designed to simplify power delivery to graphics cards. www.hardwareluxx.de/index.php/ar...
Why? Why not? Will have some time with four Arc Pro B70 in a Battlematrix workstation.
VSORA Jotunn8 - 8 compute chiplets - 8 HBM3E - TSMC CoWoS-S packaging - 500 W TDP The package is just a dummy right now. Production is ramping up right now. We will have to wait until MLPerf Inference v6.1 to see how it compares to the competition.
Four extra cores and a little software magic: Intel Core Ultra 5 250K Plus and Core Ultra 7 270K Plus review: www.hardwareluxx.de/index.php/ar...
This excellent video by Scott Manley explores how the thermal balance of spacecraft works and what it would take to design a spacecraft capable of dissipating the heat generated by space-based data centers. www.youtube.com/watch?v=FlQY...
Imec starts installing of its ASML EXE:5200 High NA EUV system in its 300 mm clean room in Leuven. Qualification should be finished in Q4 2026.
Since I’m currently looking into 200S Boost again for various reasons, I came across this fantastic German translation. "Die" (Die-2-Die) can be translated to "Stirb" in German which means "death" or "pass away". So it's a Death-to-Death-Interface.
Comparison of the logic boards of the new MacBook Neo with the iPhone 16 Pro which is using the same SoC as well as the MacBook Air with M3. Source: youtu.be/PbPCGqoBB4Y
Even though DDR prices are not going up anymore, this shows escalation stage 1 and stage 2 when it comes to memory being not affordable anymore.
Meta #MTAI Roadmap: MTAI 300: - 1x Compute - 2x Network - in production MTAI 400: - 2x Compute - 2x Network - 1x SoC-Chiplet - path to being deployed MTAI 450: - 2x Compute - 2x Network - 1x SoC - faster D2D compared to MTAI 400 - early 2027
Core Ultra 7 270K Plus vs Core Ultra 7 265K Core Ultra 5 250K Plus vs Core Ultra 5 245K But be careful: Plus SKUs did use DDR5-7200 while the non refresh SKUs did use DDR5-6400. It's legit in terms of MC verification, but keep it in mind when seeing these numbers.
Intel just announced the Core Ultra 200S Plus SKUs. Two additional efficiency cores and +900 MHz for the D2D interface. www.hardwareluxx.de/index.php/ne...
Today I saw my first meteorite burn up in the atmosphere. What a spectacle!
These are tools, that should not really exist, but for certain reasons, in some cases there is at least a desire to use them. We had a look at the Aqua Computer Ampinel and Thermal Grizzly WireView Pro II – do show the features and limitations: www.hardwareluxx.de/index.php/ar...
So we now have the Compute Tile (Intel 18A, first wafer) and Base Tile (Intel 3, second wafer), but we are still missing the I/O Tile (Intel 7) for Clearwater Forest.
#Ampinel by #AquaComputer working hard to balance a faulty 12V-2x6 cable at 600 watts. You can see how the MOSFETs of the balancing electronics are increasing the electrical resistance to regulate inequalities to a certain degree.
Intel just released some new wafer shots of Xeon 6+ aka Clearwater Forest with a different esthetic compared to what we have seen before. Thats presumably because more metal layers are present to this one?