Bionic_Squash

silicon enthusiast, has broken humour #intel #AMD #Nvidia

[Hassan Mujtaba @hms1193]

Sparkle Squeezes Intel's 32 GB Arc Pro B70 Into A Single Slot, Unlocking 256 GB VRAM Workstations https://t.co/iHNwoYHfHQ

https://t.co/iHNwoYHfHQ

[VideoCardz.com @VideoCardz]

AMD FSR 4.1 to skip RDNA 3.5 integrated graphics

https://t.co/QVOu54KsQM

[Bionic_Squash @SquashBionic]

Welp, i guess no FSR4 on APUs until Medusa

[Andreas Schilling 🇺🇦 @aschilling]

AMD currently says "no" to FSR 4.1 on RDNA 3.5, despite the fact that dGPUs on RDNA 3 and RDNA 2 will get it in July / beginning 2027.

https://t.co/OppTsnsrNK

[Bionic_Squash @SquashBionic]

Lmao wtf is Gigabyte doing

[uchy @uchy_13]

1Uで40ノードの集積の化け物みたいなサーバ。
どのような場面で必要になるのか全く想像出来な��

[Bionic_Squash @SquashBionic]

@FelixCLC_ .......i do

[vLLM @vllm_project]

Intel's AutoRound post-training quantization is now integrated natively into vLLM-Omni, bringing W4A16 (4-bit) to Omni multimodal, diffusion image, and video models. It cuts Qwen3-Omni-30B from 66GB to 25GB with no quality cliff. Quantize once offline, then serve with the same command as the BF16 model.

📈 Highlights:
- weights shrink to ~1/4 of BF16, dropping FLUX.1-dev from 4 GPUs (TP=4) to a single one
- accuracy held on OmniBench, with ~1.3% drift on text-to-image
- on Intel XPU B60, the freed memory unlocks CFG Parallel for 1.55-1.67x faster guided diffusion

🙌 Thanks to @IntelAI, @HaihaoShen, and the vLLM-Omni team for the collaboration!

🔗 https://t.co/LRvkwAnNcL

[𝐷𝑟. 𝐼𝑎𝑛 𝐶𝑢𝑡𝑟𝑒𝑠𝑠 @IanCutress]

At @gskillgaming:

➡️ 0.5 TB of RDIMMs (8 x 64 GB)
➡️ Xeon 698X
➡️ DDR5-10000 !!!
➡️ 512-bit bus
➡️ 602 GB/sec read

🤯🤯

[Bionic_Squash @SquashBionic]

@Procras36715620 @Alex_Intel_ @jaykihn0 @G_melo_ding @SteakandChickn NVL 2+0+4 was also internal, but yes cost is another factor.
They say for that target market, considering current memory situation. WCL-R makes much more sense.

[Bionic_Squash @SquashBionic]

@Alex_Intel_ Almost every decision made under justin hotard was bad, glad he left in 2025

[Bionic_Squash @SquashBionic]

@Alex_Intel_ @jaykihn0 @G_melo_ding @SteakandChickn Well the cancellation happened like 2 weeks ago

[Bionic_Squash @SquashBionic]

@Alex_Intel_ @jaykihn0 @G_melo_ding @SteakandChickn It's 2+0+4 NVL, intel says demand for WCL is more than expected and intel believes WCL-R is a more competitive product. Both WCL-R and NVL 6C overlap in target market

[Bionic_Squash @SquashBionic]

@anshelsag @highyieldYT @SteakandChickn No no, not Mediatek.
It's NVDLA NPU

[Intel @intel]

Increase performance per watt and enhance efficiency in 5G telco and cloud-native applications.

Today, Intel launches #IntelXeon 6+, the first data center CPU built on Intel 18A to help AI go to work.

Read the press release: https://t.co/6lb57CMKbz

[Bionic_Squash @SquashBionic]

@Alex_Intel_ Maybe, PRQ is Q1 27

[Bionic_Squash @SquashBionic]

Intel: DMR, Crescent island, WCL
AMD: FPGAs, MI400 series
Nvidia: Vera, Rubin, Spectrum, bluefield and VPU
Google, Meta, Microsoft: TPUv8, MTAI 400, MAIA 200
Arm: AGI CPU
Fujitsu: Monaka CPU
Samsung: LP5X-PIM
Sambanova: SN50
Broadcom: Thor ultra
IBM: Z and Inference chip

[Bionic_Squash @SquashBionic]

Damn this year's hotchips is packed, so excited for this one

[Bionic_Squash @SquashBionic]

@Alex_Intel_ Nah, Hotchips is for sharing technical details

[bobduffy 🖱️🎨 @bobduffy]

Want a sneak peek of what Intel is bringing to Computex?🚀
AI Playground Home Agent is officially available for early testing in our latest v3.1.1-beta!
@intel is bringing AI execution out of the cloud and putting it directly in your control, right from your home hardware. Here is what you can do with it today:
📱 True Remote Access: Connect and run your local AI from any device.
💰 Zero Token Costs: Free, unlimited use without subscription fees.
🛠️ MCP Integration: Local AI access and control for 3rd party Model Context Protocol supported apps.
🎨 All-in-One Creation: Advanced chat, image generation, image editing, and video generation powered by 60+ curated open-source models.

Intel and our OEM partners will be sharing more details at Computex, but you don't have to wait.
👇 Grab the beta, test it out on your Intel Core Ultra, Arc Graphics, or RTX hardware, and let us know what you think!
🔗 https://t.co/xZGluMZM7P

#AIPlayground #IntelCoreUltra #IntelArc #GenerativeAI #OpenSource #Computex2026 #ModelContextProtocol

[Bionic_Squash @SquashBionic]

Damn this year's hotchips is packed, so excited for this one

[Hot Chips @hotchipsorg]

The Hot Chips 2026 preliminary advance program with presentation titles and affiliations is now available at https://t.co/XV5mcOO5br

Speaker names and information about keynotes will be added soon.