NVIDIA presented its bets for low-end and therefore cheaper laptops with its RTX 2050 and MX550, MX570, AMD has yet to reveal too much and Intel must have seen how more details are leaking about their bets ARC DG2-128 and DG2-96. The novelty is precisely its consumption, something fundamental which indicates a first point of view on the performances which one can hope for. Will that be enough?
Without a doubt, the one that will hit the hardest is AMD with its low-end RX 6000 GPUs and Rembrandt APUs with RDNA 2 iGPUs, but its rivals will not be quiet. NVIDIA was the first, Intel could be the last and not therefore the worst, because its strategic diversification is just as interesting as that of the green team.
Intel ARC DG2-128: two different models
Intel’s bet diversifies into two GPUs for laptops and one for PC, all of which will be manufactured by TSMC at 6 nm and with that we hope they will be competitive to get a very advanced node. For this reason, the leaked data on the consumption of the variants is particularly interesting, as there are curiosities that must be taken into account.
We could divide the segments into laptops and PCs, but we will do this by model, although, as we will see, all three have their peculiarities:
- Intel DG2-128 EU (1024 shaders): 6 GB 16 Gbit / s 96 bit 75 W (desktop)
- Intel DG2-128 EU (1024 shaders): 4 GB 16 Gbit / s 64 bit 35-50 W (laptops)
- DG2-96 EU (768 shaders): 4 GB 16 Gbit / s 64 bit ~ 35 W (laptops)
Logically the 128 EU version is the full low end chip and it is quite surprising that it is shared for PCs and laptops with such an overwhelming TDP difference and with different configurations. This is not new, but we find the same number of Shaders on both platforms, with different VRAMs, data buses for GDDR6 memory and above all more than double the consumption compared to the portable version.
Because of this, the performance must be extremely uneven, at least on paper, which will be interesting for PCs and not so much for laptops, where it looks like Intel might have a harder time competing.
ARC DG2-96
This is the most economical option and therefore it has fewer benefits. The funny thing here is that it shares consumption again with the 128 EU version despite 25% less shaders, which could be due to a higher frequency to alleviate the performance deficit and not lose so much camber.
Likewise, and although it only has the difference in cores to its credit, the ARC DG2-128 EU version will have a curious power management system that Intel has integrated into its processors and that we all know about. under the name of PL2. Now it looks like it will be in the brand’s GPUs as well and could be one of the reasons Intel is now referring to PL1 = PL2 and had to make the call Maximum turbo power and the name on the other hand of Base processor power o PBP.
For this reason and as it has been disclosed we will have a TGP in the DG2-128 EU 64 bit 50 watt GPU in PL2. And what about the prices and final names? In the price there is none and there will not be in the laptop versions as this data is left to the OEMs, as for the desktop version it is said with an MSRP of $ 299. Concerning the commercial nomenclatures, we know that the version DG2-128 will be called ARC A350, while its desktop version is named ARC A380. So far nothing has been revealed about EU version 96.