devtake.dev

Hiro Tanaka

Hardware and silicon reporter. Covers CPUs, GPUs, AI accelerators, panel tech, and the gaming peripherals that ride the same supply chain.

AMD EPYC server processor in cinematic studio lighting.
Hardware·

AMD Zen 7 'Florence' leak: 288 cores, a separate cache die, and a 25% IPC swing

Leaked slides describe Zen 7 EPYC scaling to 288 cores per socket, with all L3 moved off the CCD onto a stacked cache die. Tape-out is October 2026, launch late 2028.

Render of Anker's Thus chip alongside a pair of Soundcore earbuds.
Hardware·

Anker built its own AI chip. It runs neural nets inside flash memory cells.

Anker's Thus chip embeds compute inside NOR flash, claims 150x more on-device AI for noise cancellation, and ships in Soundcore earbuds on May 21. Here's why it matters.

Steam Controller hardware press image.
Gaming·

$99 Steam Controller leak: a YouTuber broke embargo and Reddit thinks it's too much

Tech YouTuber Techy Talk pulled an early Steam Controller review minutes after posting. The leaked $99 price, gyro, and magnetic TMR sticks have already split fans.

Cerebras Systems brand image from the Cerebras website.
Hardware·

Cerebras files for an IPO again, this time with $510M in revenue and a $10B OpenAI deal in its pocket

Cerebras filed an S-1 on April 17 listing as 'CBRS,' targeting roughly $23B at the prior private mark. The OpenAI inference deal is the line item that changed the story.

NASA Artemis II mission image showing the Orion spacecraft's lunar trajectory and Earth communication concept
Hardware·

Artemis II beamed 4K video from the Moon over a laser. The backup station was a $5M Australian kit.

NASA's Orion O2O terminal hit 260 Mbps from lunar orbit to White Sands, Table Mountain, and a sub-$5M Australian station. Crewed lasercom cleared its first real test.

Intel corporate newsroom image showing Intel branding
Hardware·

Intel's foundry finally has a whale: Musk says Tesla's Terafab will run on 14A

Elon Musk said Tesla plans to manufacture at Terafab in Austin on Intel's 14A process. It's the first major external customer Intel CEO Lip-Bu Tan needs to keep the foundry alive.

Two Google TPU 8 chips, the 8t training and 8i inference variants, rendered side by side with the Axion branding
Hardware·

Google splits its AI chip: the TPU 8t trains, the TPU 8i serves, and Axion handles the rest

At Cloud Next 2026, Google unveiled TPU 8t and TPU 8i, an 8th-generation split that matches Nvidia on agentic inference. Ironwood hit GA the same day.

Framework Laptop 13 Pro product hero image showing the new aluminum chassis
Hardware·

Framework's Laptop 13 Pro is a ground-up redesign: aluminum, 20-hour battery, $1,199 DIY

Framework opened pre-orders for the Laptop 13 Pro on April 21. Panther Lake or Ryzen AI 300, LPCAMM2, a 74Wh battery, and Framework's first touch display.

TrendForce chart showing Q1 2026 memory contract price increases across DRAM and NAND
Hardware·

Why your next RAM kit costs double: the AI memory crunch, explained

DDR5 contract prices jumped 100% since 2025, HBM is eating 23% of DRAM wafers, and the fabs that could fix it don't come online until 2028. Here's the shape of the crunch.

Amazon Fire TV home screen on a television
Hardware·

Amazon is done with Android on Fire TV. Vega OS is the only future.

Amazon's new Fire TV Stick HD ships with Vega OS, no sideloading allowed. Sources say every future Fire TV will follow. Here's what changes for users and devs.

GitHub repository page for the PLFM_RADAR Aeris-10 project
Hardware·

A $5,000 open-source radar that sees 20 km, built by one engineer in Morocco

Nawfal Motii's Aeris-10 phased-array radar beats $250,000 commercial systems at 3% of the cost. Hardware, firmware, and FPGA bitstream are all on GitHub.

Meta and Broadcom partnership announcement header showing custom AI silicon
Hardware·

Meta just committed 1 gigawatt of custom AI chips with Broadcom through 2029

Meta and Broadcom expanded their partnership to co-develop four generations of MTIA chips on a 2nm process. Here's what the deal includes and why Nvidia should care.