Skip to content
Voltarum Local LLM Hardware Guide

Intel · Intel GPUs · 2026

Intel Arc Pro B70

Recommended

Intel's big Battlemage: 32 GB and 608 GB/s for less than AMD's R9700.

Memory
32 GB
Bandwidth
608 GB/s
Power
230 W
Largest popular model at 4-bit
Qwen3.6 35B A3B

from $1,268 (typical price)

Researched from manufacturer data and public sources; not tested by us.· Last reviewed Oct 11, 2026

What it can run

Estimate

Popular models at Q4_K_M with a 32K-token context. Memory needed from real model files; speed is the theoretical ceiling.

Change model size or context

Model weightsContext cache (KV)RuntimeUsable limit (colour = verdict)

ModelVerdictMemorySpeed ceiling
Qwen3 1.7BRuns
6.4 GB / 32 GB · with 32 GB
≤ 548 tok/s
Qwen3.5 9BRuns
8.3 GB / 32 GB · with 32 GB
≤ 107 tok/s
Gemma 4 12BRuns
9.5 GB / 32 GB · with 32 GB
≤ 85 tok/s
gpt-oss-20bRuns
14 GB / 32 GB · with 32 GB
≤ 304 tok/s
Qwen3.6 27BRuns
20 GB / 32 GB · with 32 GB
≤ 36 tok/s
Qwen3.6 35B A3BRuns
24 GB / 32 GB · with 32 GB
≤ 330 tok/s
Llama 3.3 70BNot enough memory
55 GB / 32 GB · with 32 GB
—
gpt-oss-120bPartly in system RAM
65 GB / 32 GB · with 32 GB
—
Qwen3.5 122B A10BPartly in system RAM
79 GB / 32 GB · with 32 GB
—
Qwen3.5 397B A17BPartly in system RAM
247 GB / 32 GB · with 32 GB
—

Theoretical maximum for generating text: memory bandwidth divided by the bytes read per token. Real runs usually reach about half to three quarters of it.

Where to buy

Prices are editorial references (launch price and a typical street price on the date shown), not live prices. In 2026 the memory shortage has pushed many prices well above launch.

Prices are shown only if updated in the last 24 hours and may change; the retailer’s price at checkout applies. We may earn a commission from these links at no cost to you. Affiliate disclosure

Overview

Editorial opinion

Best for: 32 GB per card on a budget, if you accept Intel's software path.

Editorial opinionWould we buy it today?

Promising value; the R9700 is the safer 32 GB choice today.

Pros and cons

Editorial opinion

Pros

  • 32 GB at a low price
  • 230 W reference card

Cons

  • Software support still catching up
  • Launched in 2026: fewer community results

Specifications

Factual data
Manufacturer
Intel
Chip
BMG-G31
Architecture
Xe2 (Battlemage)
Type
Graphics card
Form factor
PCIe card
Released
2026
Availability
On sale
Memory options
32 GB
Memory type
GDDR6
Memory bus
256-bit
Memory bandwidth
608 GB/s
Usable for models (default)
32 GB / 32 GB
INT8 AI performance
367 TOPS
Board / max power
230 W
Interface
PCIe 5.0 x16
Slots
2
Compute stack
SYCL / Vulkan
Models
Any model that fits
Several units for one model
Yes

Source: Manufacturer specifications

Software support

Editorial opinion

How well the popular local-AI tools run on it today (editorial assessment).

  • llama.cppGood
  • OllamaWorks with effort
  • LM StudioGood
  • vLLMWorks with effort
  • MLXNot supported

Measured speeds

Factual data

Measured results with their source. We never invent speeds.

No measured results yet. The table above shows the theoretical ceiling instead.

Alternatives

AMD GPUs

AMD Radeon AI PRO R9700

Memory
32 GB
Bandwidth
645 GB/s

Largest popular model at 4-bit: Qwen3.6 35B A3B

32 GB for about a third of an RTX 5090: the cheapest new 32 GB card.

from $1,550 (typical price)Recommended

NVIDIA GPUs

NVIDIA RTX PRO 4000 Blackwell

Memory
24 GB
Bandwidth
672 GB/s

Largest popular model at 4-bit: Qwen3.6 27B

24 GB in a single-slot, 140 W card: the homelab-friendly way to get 24 GB new.

from $1,500 (typical price)Recommended

AMD GPUs

AMD Radeon RX 7900 XTX

Memory
24 GB
Bandwidth
960 GB/s

Largest popular model at 4-bit: Qwen3.6 27B

24 GB and 960 GB/s: RTX 3090-class memory, bought new.

from $1,358 (typical price)Recommended