Founder, Country Manager, TinyML , EdgeML & AI Acceleration Sales Executive, SoMs & SBC Global Business Development, Sales & Marketing Semiconductors Specialist
Thanks you Nitin Dahad for the opportunity to present the latest on EdgeAI and the ability to run NVIDIA TAO models on any edge device using Edge Impulse Platform at Hardware Pioneers Max couple of months ago.
Renesas Electronics' new RA8 #microcontroller family is based on the Arm #CortexM85 with #Helium #machinelearning acceleration. Watch this video to learn how it works ⬇️
📉 +$443 Trade Recap: AMD Short to Pre-market Low then ORB Long to the Camarilla Pivots🚀 Trade Date: 6/14/24 1st trade, AMD had a nice reversal off pre-market low to S1, saw a quick rejection and went short (small size as I was not that confident) to VWAP/S2 and PM/S3 low targets. We sold off nicely on the retest of PM low but Doji'd out. AMD was holding VWAP for a potential squeeze higher, and I exited before B/E. Better to get out for some profit than wait for a stop out at B/E if you sense it! 2nd trade, beautiful bullish 1st 5-min candle, 2nd candle also had a great wick and we were above VWAP. Saw momentum flowing for an ORB higher and went long (heavy size) for HOD break towards S1 then to PDC/R1, large orders at $160, and a final push to R2 goal. We ended up testing R3 but by that time we were extended and wicked on decent volume for the rejection. #AMD #CamarillaPivots #ORB #OpeningRangeBreakout
🚀🚀Today Hugging Face accelerate v0.28.0 has been released! 🚀🚀 A brief summary of our release notes: * We are deprecating a few arguments in the `Accelerator` to make room for a `DataLoaderConfig`. This allows for easier future-proofing of more arguments to tweak what happens behind the scenes. These are arguments like `split_batches` and `dispatch_batches` * After discovering that while not synchronizing the gradients each batch can have some speedups in raw DDP, with FSDP this causes a large memory spike. As a result we make this configurable now * We've added support for XLA GPUs and gradient accumulation on XLA devices * Downstream FSDP + QLORA support has been started by allowing for a more configurable mixed precision setup with FSDP * `accelerate launch` now supports `mpirun` for multi-CPU training thanks to the folks at Intel Corporation!
Motorola Edge 50 Review: Mid-Range Marvel with Premium Features Read Here: https://lnkd.in/gx8UfVCj #MotorolaEdge50Review #MotorolaEdge50 #MotorolaReview #SmartphoneReview - #EdgeSeries #technews #mymobileindia
Motorola Edge 50 Review: Mid-Range Marvel with Premium Features Read Here: https://lnkd.in/gx8UfVCj #MotorolaEdge50Review #MotorolaEdge50 #MotorolaReview #SmartphoneReview - #EdgeSeries #technews #mymobileindia
Discover the secrets of professional content creation with SanDisk PRO TEAM ambassador, Jerry Ghionis! In this video, Jerry dives into the amazing features of the NEW SanDisk Desk Drive SSD and talks about how it can upgrade a photographer’s workflow. https://lnkd.in/gyHATTd5
Thought for sellside to ask AMD and Lisa Su .. about how much Mi300 revenue will be coming from Chinese customers in CY24? What is the timing of the Mi400 ? As I understand, NVDA will have its next generation GPU chip in customer hands for testing by summer of 2024.. so effectively Mi300X will already by outdated by the 2H ramp unless Mi400 is the real revenue generator next year. (Missing sellside days)
Technology evangelist and writer
1moAmir Sherman it’s always a pleasure talking to you!