Dell is releasing its version of NVIDIA Spark in the mini PC form factor named as Dell Pro Max with GB10. Specs are all the same except this will be marketed and serviced by Dell, it has petaflop (1,000 TOPS) of FP4 AI performance, GB10 Grace Blackwell Superchip. Paired with 128GB of unified LPDDR5x memory, it’s got the… Read More »
NVIDIA’s newly launched enterprise cards has just got their price announced, check the pricelist of Pro Blackwell graphics card ranging from 96GB VRAM of the RTX PRO 6000 to the 24GB VRAM of the RTX Pro 4000. NVIDIA is launching cards that cater to wide range of audiences, you can compare with RTX 5090 here. These will start… Read More »
NVIDIA today also launched a big brother of the Spark (Digits), The DGX Station is a new type of computer made for working with AI, it has server grade parts and it would be insanely expensive ($20,000+) to be called a desktop but it is here now. It’s the first to use the NVIDIA GB300 Grace Blackwell Ultra… Read More »
DGX Spark formerly Digits are now available to order at 3,999 for the Nvidia Founders edition and $2999 for the Asus branded Spark named as Ascent GX10. Asus one comes with 1TB whereas Nvidia branded one comes with 4TB NVME space, all models will come with Ubuntu based DGX OS preinstalled. The problem with Spark is that it… Read More »
You can now own the full fat die silicon in the form of RTX PRO 6000 Blackwell Workstation Edition, launched today at GTC 2025. As compared to the RTX 5090 it offers a bit more performance and ECC memory. The RTX PRO 6000 is designed for professional workstations, while the RTX 5090 is a consumer gaming GPU. Compared… Read More »
Aoostar upcoming Mini PC is powered by the Radeon 890M integrated graphics featuring the RDNA 3.5 architecture and has 16 computing units, operating at 2.9GHz, this has enabled this mini PC to reach RTX 2060 equivalent power in this compact design, you can definitely play AAA gaming titles of the previous years on this. The inbuilt flip display… Read More »
Apple just released the MAC Studio upgrades for the year 2025, powered by both M4 Max and M3 Ultra. Yes, it will be available in two flavors: M4 Max will be up to 128GB unified RAM and M3 Ultra gets up to 512GB Unified RAM. The M3 Ultra Studio doubles the memory bandwidth (819 GB/s vs. 546 GB/s)… Read More »
One of the biggest challenges in running local LLM models right now is having enough VRAM to keep the models weights in VRAM for fast generation (tok/s). Industry is starving right now to get bigger VRAM, both AMD and NVIDIA are gatekeeping their higher VRAM cards for enterprises, these AI MAX Strix Halo systems are all you will… Read More »
With Strix Halo already here, Local LLM runners are now waiting for the next big thing in local i.e the NVIDIA DIGITS, dedicated hardware to run local models, it will be based on Ubuntu 22/24 and will come bundled with NVIDIA software ecosystem already installed. In this post we will compare their architectures, specifications, and projected LLM speeds… Read More »
The upcoming RTX 5070 Ti is built on NVIDIA’s advanced “GB203” silicon, which is the same 4nm chip as the elusive and overpriced RTX 5080. 5070ti however should be available widely, specs wise it includes 8,960 CUDA cores, derived from 70 out of 84 SMs, along with 280 Tensor cores and 70 RT cores. The card is also… Read More »