Intel Arc A370M Laptop GPU Transforms into ITX-Sized Desktop GPU

[AleksandarK]

Taiwanese tech maker Advantech has converted Intel's Arc A370M mobile GPU into a desktop graphics card named the EAI-3100. The new card utilizes the same Arc A370M mobile GPU based on the Xe-LP architecture chip as found in laptops but adds more robust cooling to enable desktop-level performance. Specifically, the EAI-3100 implements a large aluminium heatsink spanning the entire PCB, paired with a 40 mm fan active cooling fan. This allows the card to operate at up to 60 Watt TGP (total graphics power), a noticeable increase over the A370M's 35-50 Watt mobile power range. Despite the improved cooling, Advantech has not factory overclocked the EAI-3100, leaving its graphics clock speed unchanged at 1,550 MHz. The card also retains the same PCIe 4.0 x8 interface as the mobile A370M. An 8-pin PCIe power connector has been added, giving headroom for user overclocking attempts.

In terms of gaming performance, the A370M and, by extension, the EAI-3100 deliver playable frame rates at 1080p resolution with medium image quality settings. The card is comparable to NVIDIA's mobile RTX 3050 GPU. As Intel continues optimizing Arc drivers, more gains are expected. The EAI-3100's dual-slot, 6.61-inch design allows compatibility with most desktop PC cases. Between its small size and the A370M's solid 1080p capabilities, this transformed card represents an interesting budget option for gamers seeking a discounted route to Arc's architecture. Despite the diminutive size, this custom cooling solution keeps the A370M at appropriate temperatures for sustained operation, possibly delivering more than the laptop form factor SKU. For video output, the card features two HDMI 2.0b and two DP 1.4a ports.



View at TechPowerUp Main Site | Source

 

[Toothless]

Little AV1 encoder! Hopefully it comes out at a proper price for it.

 

[Fouquin]

Little AV1 encoder! Hopefully it comes out at a proper price for it.

Affordable pricing isn't in Advantech's vocabulary. Their primary customer is industrial and automation. These won't be cheap.

 

[AusWolf]

What's the point? Actually launching the A310 would be way more useful, in my opinion.

 

[Toothless]

Affordable pricing isn't in Advantech's vocabulary. Their primary customer is industrial and automation. These won't be cheap.

Pls no killing dreams, ty

 

[Denver]

It's not worth the headache, but it would be nice to have more decent GPUs in this compact form factor. The last one was the R9 Nano.

 

[SOAREVERSOR]

It's not worth the headache, but it would be nice to have more decent GPUs in this compact form factor. The last one was the R9 Nano.
View attachment 332433
View attachment 332432

That was a flop.

 

[Rexter]

But, why not half-height single slot?

 

[docnorth]

What's the point? Actually launching the A310 would be way more useful, in my opinion.

They are kind-hearted, they don't want to kill gt 1030 and gtx 1630 (although 1630 is already dead I hope)...

 

[Denver]

That was a flop.

Nah. In my opinion it was a very respectable compact GPU, 90% of the Fury X's performance and less than half the power consumption. There is nothing like it today, I wish HBM interactions had become cheaper and become the default choice.

 

[AusWolf]

They are kind-hearted, they don't want to kill gt 1030 and gtx 1630 (although 1630 is already dead I hope)...

The 1630 was dead on arrival. Half the performance of the already weak 6500 XT for the same price... lol!

 

[SOAREVERSOR]

Nah. In my opinion it was a very respectable compact GPU, 90% of the Fury X's performance and less than half the power consumption. There is nothing like it today, I wish HBM interactions had become cheaper and become the default choice.

HBM is not needed for casual time wasting stuff. Of which gaming is the gold standard of casual time wasting. HBM is still around for things that aren't wastes, aka professional stuff. As it should be and always will be. Time wasting bullshit should get bullshit, as gaming does. Sun still rises and sets!

 

[FrostWolf]

But, why not half-height single slot?

Exactly what I was thinking. A half-height single or double-slot would make far more sense at the entry level, because this is the space it will sell in.

I'd like an alternative to entry Quadro cards, which are currently the only reasonable ones out there (there is one Radeon option, they aren't made as well by the vendors that supply them). It's also not clear whether the PCIe power connector is optional; that seems an odd addition for the segment they're selling to.

 

[Assimilator]

Unnecessarily full-height dual-slot cooler, unnecessary extra power connector, another manufacturer that has no clue as to what the market for these would be if they thought about it for 5 seconds.

 

[Denver]

HBM is not needed for casual time wasting stuff. Of which gaming is the gold standard of casual time wasting. HBM is still around for things that aren't wastes, aka professional stuff. As it should be and always will be. Time wasting bullshit should get bullshit, as gaming does. Sun still rises and sets!

You seem to be a troll devoid of any arguments, so your post suits you well.

But I'm polite and I'll hit you with rational arguments. Games is a (billion dollar) entertainment market like any other, and produces millions of jobs throughout the chain. Professional stuff that produces what? Does it render films that are entertainment? is it used in streaming infrastructure that is also entertainment? or help produce games that are also entertainment? Shoot wherever you want, it's all entertainment.

But in fact HBM was designed by AMD as a new and efficient memory format for GPUs in general, no wonder it was launched in a gaming GPU. The companies that manufacture HBM just haven't been able to solve the challenge of making it cheap; each interaction is faster but also more expensive.

 

[Fouquin]

The last one was the R9 Nano.

There have been a bunch of GPUs in this form factor since the Nano. RTX 3060, RTX 4060 (also available from Palit, Inno3D, Colorful, MSI) and RTX 4060 Ti most recently received the ITX treatment. There's also the half height RTX 4060 for something even smaller. The RX 6600 XT did get an ITX variant but it's not as small as the R9 Nano.

Unnecessarily full-height dual-slot cooler, unnecessary extra power connector, another manufacturer that has no clue as to what the market for these would be if they thought about it for 5 seconds.

They already know their market, they don't need to ask us. These are intended as entry level edge AI hardware accelerators for their embedded systems with a 5 year minimum service life. The large cooler is for allowing a configurable TGP between 35W and 60W, so the customer can choose what their priority is for their system; power efficiency or performance.

 

[NewMoon]

I honestly see this as a competitor to the 1650 and 6400 (power efficiency wise, but computing power wise, no idea). But for multimedia use, this is a good brawler to the lower end of the market who want to take advantage of it. Mileage will vary on the user's preference tho

 

[Lew Zealand]

Unnecessarily full-height dual-slot cooler, unnecessary extra power connector, another manufacturer that has no clue as to what the market for these would be if they thought about it for 5 seconds.

This right here, especially the freakin 8-Pin power connector for a 60W card. This is how to completely miss the narrative.

 

[Assimilator]

They already know their market, they don't need to ask us. These are intended as entry level edge AI hardware accelerators for their embedded systems with a 5 year minimum service life. The large cooler is for allowing a configurable TGP between 35W and 60W, so the customer can choose what their priority is for their system; power efficiency or performance.

The RTX A2000, which draws more power than this turd, gets away with no external power connector and a half-height cooler. Yet you're here arguing that it makes sense to marry a much larger cooler, and provision for drawing up to 225W of power, to this GPU that caps out at 60W. Which a PCIe slot easily provides.

 

[Dr. Dro]

That was a flop.

The R9 Nano wasn't a flop and people who have them tend not to sell them so easily as the market prices that it commands tend to be laughably low - it's worth just keeping it. I looked it up out of curiosity as I recall this being the case some time ago, there's no active listings on AliExpress and there's one in working order listed for $79 USD on eBay right now.

But in fact HBM was designed by AMD as a new and efficient memory format for GPUs in general, no wonder it was launched in a gaming GPU. The companies that manufacture HBM just haven't been able to solve the challenge of making it cheap; each interaction is faster but also more expensive.

The biggest issue that HBM faces is that you can only test the extra fragile steps after the entire package has been assembled. If the memory is bad, you have to throw out the entire processor, GPU core and all, because you can't disassemble the processor after the packaging process. That's why HBM will never be cheap until yields exceed 95%.

 

[Fouquin]

The RTX A2000, which draws more power than this turd, gets away with no external power connector and a half-height cooler. Yet you're here arguing that it makes sense to marry a much larger cooler, and provision for drawing up to 225W of power, to this GPU that caps out at 60W. Which a PCIe slot easily provides.

There is no argument. Simple statements of fact.

If you look at their documentation you will see quite clearly that the 8-pin power is optional. It is likely expected to be used in situations where the mainboard is not guaranteed to deliver power, such as in some video slot machines.

 

[lexluthermiester]

For video output, the card features two HDMI 2.0b and two DP 1.4a ports.

NVidia and AMD take notice. THIS is how output port selection is done right. Get on it and get it right!

 

[trsttte]

NVidia and AMD take notice. THIS is how output port selection is done right. Get on it and get it right!

Not exactly, right would be DP2.0 (which the arc cards support) and HDMI 2.1 but well, better than most i guess

But yeah, 2 of each is better than what most cards do with 3 DP and 1 HDMI. Gigabyte often does the right version (2+2)

 

[lexluthermiester]

Not exactly, right would be DP2.0 (which the arc cards support) and HDMI 2.1 but well, better than most i guess

Fair point. However, I was referring to the 2 x DP and 2 x HDMI.

 

[Rexter]

There have been a bunch of GPUs in this form factor since the Nano. RTX 3060, RTX 4060 (also available from Palit, Inno3D, Colorful, MSI) and RTX 4060 Ti most recently received the ITX treatment. There's also the half height RTX 4060 for something even smaller. The RX 6600 XT did get an ITX variant but it's not as small as the R9 Nano.



They already know their market, they don't need to ask us. These are intended as entry level edge AI hardware accelerators for their embedded systems with a 5 year minimum service life. The large cooler is for allowing a configurable TGP between 35W and 60W, so the customer can choose what their priority is for their system; power efficiency or performance.

View attachment 332453

Ok that does make this product make more sense actually. I can see now it's purpose built for that.
But there is still a huge gaping hole in the low-mid end low profile gpu segment, where this would be a great contender.