Local LLM Hardware: NVIDIA, AMD and Intel — After Capacity, It's Bandwidth

When choosing hardware for local LLMs, look at memory bandwidth, not just memory capacity.

Cloud API bills and data-security concerns are pushing more developers to run AI models on their own machines, and NVIDIA, AMD and Intel are each offering developer hardware in a different form. Put the specs side by side and one thing stands out: two products can both have 128 GB of memory, yet differ a lot in how fast they can read it (bandwidth), and that speed largely decides how fast the AI produces text.

A few terms first. VRAM is memory dedicated to a graphics card. Unified memory means the CPU and GPU share one pool of memory. Bandwidth is how fast memory can move data (GB/s). Inference is using a finished model to produce answers.

At a glance

Product	Memory	Bandwidth	Reported price
NVIDIA DGX Spark (GB10)	128 GB unified	273 GB/s	$4,699 (raised in February)
NVIDIA RTX PRO 6000 Blackwell	96 GB VRAM	1,792 GB/s	about $8,565 at launch, recently $16,000 on the marketplace (out of stock)
AMD Ryzen AI Halo (Ryzen AI Max+ 395)	128 GB unified	about 256 GB/s (theoretical)	$3,999 (availability unconfirmed)
Intel Arc Pro B60 / B70	24 GB / 32 GB	not found	not found

Prices move fast and differ by source. Check the official site and retailers for current prices and stock before buying.

NVIDIA: both work, but they are different things

DGX Spark puts 128 GB of unified memory in a small box. It is good for loading big models whole, but its 273 GB/s of bandwidth is criticized as lower than Apple’s M4 Max (about 400 GB/s). Its price reportedly rose from $3,999 to $4,699 in February because of memory supply constraints.
RTX PRO 6000 Blackwell is a 96 GB workstation GPU that plugs into a regular PC. Its bandwidth is more than six times the Spark’s, so the same model runs much faster, but it costs far more. This class of card has no fixed list price, so prices vary by seller and have climbed sharply recently.

AMD: 128 GB unified memory, and setup is the catch

At CES this year AMD introduced Ryzen AI Halo, a developer mini PC with 128 GB of unified memory and ROCm preinstalled. It was slated for June, but we could not confirm that it has actually shipped.
PCs from other makers use the same chip (Ryzen AI Max+ 395). By default only a small amount of memory goes to the GPU, so you have to configure it; on Windows you can give the GPU up to 96 GB of the 128 GB.
Community reports often say the Vulkan backend works better than ROCm for models of 70B and up. That comes from blog-level benchmarks, so it is best to compare on your own workload.

Intel: big memory is for servers, developer parts are experimental

Intel’s large-memory part, Crescent Island, is a data-center inference GPU with 160 GB of LPDDR5X. Sampling is expected in the second half of this year and general availability in 2027, so it is not something an individual developer will buy.
What individuals can use is the Arc Pro B60 (24 GB) and B70 (32 GB) cards. Users report running them with llama.cpp (SYCL, Vulkan) and vLLM, along with driver errors and out-of-memory problems.
So the idea that Intel is going after developers with large unified-memory chips doesn’t fit today. What developers can get now are Arc Pro cards, and on the software side OpenVINO has only a preview llama.cpp backend.

What could change for developers

What follows is our estimate and advice based on the reports.

Model size first, then speed. If the model doesn’t fit in memory, speed is irrelevant. Check capacity against the model you want to run, then compare bandwidth.
Same capacity, different feel. Even among 128 GB unified-memory products, and against a 96 GB VRAM card, the speed at which text appears can differ a lot.
Check software support. CUDA is the easiest; reports for ROCm and Intel (SYCL, OpenVINO) often mention setup effort. Make sure the tools you use (Ollama, llama.cpp, vLLM) support the hardware.
Prices swing. There are reports that memory shortages pushed prices up. When comparing with cloud API costs, recalculate with current prices and your real usage.

What to watch

The actual release and price of Ryzen AI Halo, successors to the Spark and price trends, Intel’s Crescent Island schedule, and the state of ROCm and OpenVINO support.