Catalogue

Intel Arc Pro B70 32GB

5 1 rating
Graphics Card - 32GB GDDR6 (19000MHz), Intel ARC Battlemage (BMG-G31, 2280 MHz), Boost 2800 MHz, PCI Express x16 5.0, 256Bit, DisplayPort 2.1, width 266,7 mm
-1 % off new 659,090 Ft New 665,790 Ft
Shop faster with pickup at AlzaBox or in-store.
Intel Arc Pro B70 32GB 659,090 Ft Unsealed - in stock 3 pcs
Buy

Intel Arc Pro B70 32GB

Intel Arc Pro B70 32GB is an almost new item, in perfect condition, and comes with a full warranty. However, someone opened the original packaging before you. Why? It could have been an unwanted gift. For detailed information, click here.
  • .
Intel logo

Intel Arc Pro B70 32GB graphics card

The Intel Arc Pro B70 32GB professional graphics card is designed for workstations. With its Battlemage architecture and 32GB of GDDR6 memory, it excels at handling large datasets, demanding visualisations, and AI acceleration. A computing performance of 367 TOPS and 256 XMX engines supports machine learning tasks. It uses a PCI Express 5.0 x16 interface to connect to your computer, and the card supports up to four monitors with 8K resolution. ECC memory ensures reliability for critical computations.

Intel graphics card

GDDR6 / 32GB

Operating memory
Intel graphics card

PCIe 5.0 x16

Bus interface
Intel graphics card

1× 8-pin

Additional power
Intel graphics card

Battlemage

Chip architecture
Intel graphics card

550W

Recommended PSU wattage
Intel Arc Pro B70 32GB graphics card

Key features of the Intel Arc Pro B70 32GB graphics card

  • 256-bit bus width
  • Boost core frequency reaches 2800MHz
  • AI performance of 367 TOPs
  • 4096 stream processors
  • TDP of 290W
Intel Arc Pro B70 32GB graphics card

Memory system for large datasets

The card features 32GB of GDDR6 memory, which uses a 256-bit bus. With an effective clock speed of 19Gbps, the memory bandwidth reaches 608GB/s. Thanks to these parameters, the card smoothly processes large-scale projects and voluminous data. ECC memory ensures data reliability and integrity during critical computations.

Intel Arc Pro B70 32GB graphics card

AI acceleration

AI computing performance reaches 367 TOPS thanks to 256 dedicated Intel XMX Engines. This AI graphics card accelerates machine learning and deep learning tasks. Software support for platforms like oneAPI, OpenVINO™, and Intel® Extension for PyTorch expands its use cases in data science and AI development. 

Intel Arc Pro B70 32GB graphics card

Connectivity and multimedia capabilities

The PCIe 5.0 x16 interface ensures fast communication with other components. The card supports connecting up to four monitors simultaneously via four HDMI 2.1 outputs, at resolutions of up to 7680×4320 at 120Hz. For content creators, it offers hardware acceleration for processing modern video codecs, including encoding and decoding AV1, H.265, and H.264 formats. The card's core uses 5nm manufacturing technology and contains 32 Xe-cores, 32 Ray Tracing Units, and 4096 stream processors. In Boost mode, the core frequency reaches up to 2800MHz, processing complex graphics and computational operations. An active dual-slot cooler with a single fan maintains optimal operating temperatures.

Technical specification

  • Graphics processor: BMG-G31
  • Chip architecture: Battlemage
  • AI performance: 367 TOPS
  • Core frequency: 2280MHz
  • Core frequency (Boost clock): 2800MHz
  • VRAM capacity: 32GB
  • Memory type: GDDR6
  • Memory frequency: 19GHz
  • Memory bandwidth: 608GB/s
  • To connect a monitor: 4× HDMI 2.1
  • Maximum resolution: 7680×4320 @ 120Hz
  • Codec support: AV1, H.265, H.264
  • Cooler type: Active
  • Cooler size: Dual-slot
  • Number of fans: 1
  • Additional power supply: 1× 8-pin
  • Recommended PSU power: 550W

Specifications

Compare
Chip manufacturer
Intel ARC
Model designation
Intel Arc Pro B70
Graphics processor
BMG-G31
Chip architecture
Battlemage
Number of stream processors
4,096×

Chip

Chip manufacturer
Intel ARC
Model designation
Intel Arc Pro B70
Graphics processor
BMG-G31
Chip architecture
Battlemage
Number of stream processors
4,096×
Production technology
5 nm

Performance & usage

Performance
Maximum
Use
Professional
Card profile
FHFL

Memory

VRAM capacity
32 GB
Memory type
GDDR6
Bus width
256-Bit

Frequency

Core frequency
2,280 MHz
Core clock speed (Boost clock)
2,800 MHz
Memory frequency
19,000 MHz (19 GHz)
Effective memory clock (bandwidth)
19 Gbps

Interface

PC connection
PCI Express x16 5.0
Bus interface
PCIe 5.0 x16
Number of image outputs
4
Maximum number of connected monitors
4

Technology

DirectX version
12 Ultimate
OpenGL version
4.6

Cooling system

Cooling type
Active
Cooling size
Dual-slot
number of fans
1 ×

Dimensions

Width
266.7 mm
Depth
99.06 mm

Consumption

TDP
290 W
Auxiliary power supply
1× 8-pin
Recommended power sources
550 W

AI

Total number of TOPS
367
Number of TOPS (GPU)
367
Software platform
Intel OpenVINO
INT8 TOPS
367
Supported precision
FP32

Tops

Total number of TOPS
367

AI specifications

Primary AI use
LLM Inference
Certified AI frameworks
PyTorch
More Specifications

Intel Deutschland GmbH, Dornacher Str. 1, 85622 Feldkirchen, Germany [email protected] 


Intel Corporation is the world's largest manufacturer of semiconductor circuits and related components, powering a large share of the world's computers, devices, and machines with their powerful and reliable processors. One of the most established and successful companies to come out of Silicon Valley, they've been a driver of the technology revolution ever since.

Intel employs over 96,500 employees worldwide and is one of the biggest and richest companies in the world. For decades, their slogan was "Intel Inside"; however, they replaced with the more dynamic "Leap ahead" in 2005, signalling a renewed commitment to innovation.

Note:

This product is not an independently functional unit and may require professional installation.

Specifications can be changed without notice. Images are for illustrative purposes only.

There are no posts in this discussion.
More Information

Mesterségesintelligencia-gyorsítók

A modern AI-modellekre jellemző, hogy elkülönítik a tanulási (training) és a futtatási (inference, üzemeltetési) módot. A tanulás rendkívül nagy teljesítményű hardvert igényel, és jelentős energiafogyasztással jár, míg a modellek futtatása lényegesen kevésbé erőforrás-igényes. A kellően optimalizált modellek olyan kicsik és hatékonyak lehetnek, hogy akár klienseszközön is futtathatók – legyen szó mobiltelefonról, AR szemüvegről, okosautók processzoráról vagy robotokról. Az ilyen jellegű AI-feldolgozást edge AI-nak, illetve edge AI computingnak nevezzük („peremhálózati feldolgozás”, azaz a számítások az adatforráshoz közel történnek).

The Lifecycle of AI Hardware: How to Make a Purchase That Won't Become Obsolete

AI hardware pays off based on factors other than its technical lifespan—the decisive factor is the point at which its operation becomes less cost-effective than the available alternative. The article offers advice on how to assess a system’s economic lifespan, what to focus on when comparing new-generation accelerators, which parts of the infrastructure (not just GPUs) typically reach their limits first, and how to factor in purchasing, leasing, and phased upgrades into the decision-making process. It also includes tips on how to continue making meaningful use of older hardware after it has been taken out of primary production.

TCO kalkulátor: mennyi idő alatt térül meg a saját infrastruktúra a felhőhöz képest?

A gyakorlati megtérülési számítás megmutatja, mikor éri meg egy vállalatnak az AI-működés egy részét a felhőből saját infrastruktúrára áthelyezni. A cikk az NVIDIA DGX Spark 4TB valós árával, TCO-modellel és olyan költségekkel dolgozik, amelyeket a jóváhagyás során gyakran alábecsülnek.

AI Agents and Automation in Practice: From Pilot to Production

AI agents are currently attracting companies with quick pilot projects, in which it’s relatively easy to verify initial benefits. The true value only becomes apparent when the agent proves itself in day-to-day practice and the company can safely integrate it with its data, systems, and internal policies. This article explains how the demands on performance, monitoring, and infrastructure surrounding the model itself change during the transition from testing to production. It will help clarify when a workstation or a smaller AI server is sufficient for a company and when it needs a more robust solution.

Az AI-modellek robbanásszerű növekedése

A mesterséges intelligencia iránti jelenlegi érdeklődési hullám az OpenAI generatív intelligenciája, a ChatGPT megjelenésével, valamint a DALL·E és a Midjourney generatív intelligenciák 2023 eleji hozzáférhetővé válásával kezdődött. Ezt a megjelenést a kritikusok néha „túl korainak” nevezik – de mit jelent az, hogy valami túl korai a technológiák világában?
Reviews Intel Arc Pro B70 32GB
Accessories for Intel Arc Pro B70 32GB
Try our cookies Alza.cz a. s., Company identification number 27082440, use cookies and other data to ensure the proper functioning of the website and, with your consent, also, among other things, to personalize advertising and the content of our websites. By clicking on the “I understand“ button, you agree to the use of cookies and the transfer of data regarding the behavior on the website for displaying targeted advertising on social networks and advertising networks on other websites.
Detailed settings
Alzak icon
P-DC1-WEB15
30 év tapasztalat az e-kereskedelemben
3 million megrendelés évente
a vásárlók 98% visszatérõ
Több infó