- Intel Arc Pro B50 Double Xe Cores Over A50, offering stronger performance
- The GPU of the Intel workstation brings a memory of 16 GB against the 8 GB of Nvidia
- 168 mm low profile card design highlights
The latest graphics card of Intel’s workstation, the Arc Pro B50, impressed the way it accumulates with NVIDIA RTX A1000.
Combined as an entry -level professional GPU, the B50 is of compact but ambitious size, transporting features that push it in a serious workstation territory.
The Arc Pro B50 is built on the XE2 architecture of Intel, the same base as the Battlemage consumption line, which also feeds GPUs in laptops of play.
A workplace card with modest ambitions
This chip doubles the Xeurs and the tracing units of the preceding Pro A50 rays and is equipped with new engines with XMX matrix.
The Arc Pro B50 works on the GPU BMG-G21, a smaller chip that has deliberately capped for cost reasons.
The map offers 16 XE2 cores and 128 XMX motors, supported by 16 GB of GDDR6 memory on a 128 -bit interface, more than the 8 GB of the A1000.
This additional capacity provides a larger stamp for professional workloads such as CAO models, video editing or execution of small models of AI locally.
With a bandwidth of 224 GB / S and 170 peaks of the FP8 calculation, it targets efficiency rather than raw power.
In this regard, the B50 leads the A1000, which operates at a 192 GB / s lower bandwidth and only delivers 107.8 above in int8 workloads.
Unlike consumer conceptions, the B50 is available in a low -profile shape factor measuring only 168 mm, cooled by a single radial fan.
Its TDP 70W means that it does not require an auxiliary power, and there are four mini DisplayPort 2.1 outputs, marking an upgrade on the old 1.4A connectors of the A1000.
These additions are targeting workloads such as rendering, automatic learning inference and content creation.
Intel has also obtained more than 50 certifications with the main software suppliers, a step necessary to compete with the workstation market.
In the references, B50 of Intel surprised many observers. Geekbench AI showed it well in advance on the NVIDIA A1000 between precision and habit tests.
Procyon AI benchmarks have also highlighted faster results in the vision of the computer and the generation of text.
The Arc Pro B50 also made a modest 7% advantage on the RTX A1000 in Photoshop, with even greater earnings during the isolation of GPU tasks.
Premiere Pro showed similar results, the B50 outperforms its rival by almost 20% overall.
The results of the effects were closer, although the B50 has demonstrated clear forces in 2D workloads.
The most surprising result may have come to Blender, because the rendering is traditionally an area of ​​domination of Nvidia, but the B50 posted a 20% lead on the A1000.
This generational jump, more than 130% compared to the old A50, shows that Intel approached the past gaps.
The card also excelled in MLPERF markers, generating tokens almost twice faster than the entry -level NVIDIA professional option.
However, in areas like Revit and Inventor, the B50 often dragged RTX A1000, sometimes by large margins.
In Inventor’s graphics test, the Nvidia card took place more than four times faster. However, in Solidworks, the B50 reversed the result, displaying an advantage of 33%.
The Intel Arc Pro B50 is not the best GPU for each task, or will move Nvidia in all markets, but for a sales card at $ 350, it offers high value for professional users.
Via Puget Systems / Hardwareluxx
