Teardown of GIGABYTE RTX 5060 Ti EAGLE Card Reveals Stubby PCB Design & Short PCIe Connector

[T0@st]

GIGABYTE's GeForce RTX 5060 Ti EAGLE OC graphics card model was officially unveiled on Tuesday (April 15). Two days later, the manufacturer's PR team disclosed extra in-depth details—a hype-up section stated: "(our) EAGLE series features a design inspired by the fusion of aerospace battleships and sci-fi elements, making it a preferred choice for sci-fi enthusiasts and younger users...These graphics cards are more than just components—they become battleships within the system, enhancing the overall aesthetic and immersive experience.⁠" Yesterday's press release did not delve into under-the-hood information, but reviewers have discovered that GIGABYTE's engineering department has bunged an extra short PCB design into the new-gen EAGLE's dual-fan enclosure.

Germany's HardwareLuxx received samples for evaluation purposes—directly from three brands: the aforementioned GeForce RTX 5060 Ti EAGLE OC 16 GB SKU, as well as MSI's GAMING TRIO model, and PALIT's Infinity 3 card. The site's editor—Andreas Schilling—was enchanted by the EAGLE's diminutive setup; both externally and internally. As explained at the beginning of HardwareLuxx's review, a main highlight is the brand-new product's size: "at 215 mm, the card is particularly short. Also striking is the 8-pin connector located directly behind the slot cover—an unusual position for the additional power supply. Equally striking is the short PCI Express connector. Since the GeForce RTX 5060 Ti features a PCI Express interface with eight lanes, Gigabyte eliminates the need for a longer PCB and shortens the connector accordingly." GIGABYTE has likely deployed its dinky PCB layout in new WINDFORCE (standard and overclocked) options. VideoCardz believes that the shorter design is reserved for dual-fan cards. By rule of thumb, triple-fan cooled cards are available with the regular length board and connector. Even GIGABYTE's upcoming GeForce RTX 5060 OC Low Profile 8G (182 mm) model sticks with a "full-sized" PCIe interface.



View at TechPowerUp Main Site | Source

 

[Octavean]

That's funny. You know what else is funny? Despite its diminutive PCB, they still felt the need to make a dual slot design. The price might be amusing too if it weren't so sad,....

 

[ThomasK]

I owned a GT 9500 DDR2 once, the design looked more sophisticated than this.

 

[N/A]

Half a video card for half a thousand. Makes sense.

 

[Gucky]

That's funny. You know what else is funny? Despite its diminutive PCB, they still felt the need to make a dual slot design. The price might be amusing too if it weren't so sad,....

180W is not easily dissipated. 1x fan cards usually have high temps and are loud.

 

[_roman_]

I do not know if I want the power connector near the Input/Output Bracket.

From the electronics viewpoint a smaller printed circuit board footprint is always the goal. Nothing wrong with that.

Judging from the BIll of Materials I think the graphic card could be in the 100€ range.

 

[ZoneDymo]

if only it wasnt for that dang price tag (and the inclusion of a dang 8gb model) I would praise Nvidia, its so sad how depended everything is on that, damn it Nvidia you make so ffing much money on that stupid AI crap, who cares if you even make a loss on these....

 

[Dr. Dro]

I owned a GT 9500 DDR2 once, the design looked more sophisticated than this.

Looked sounds about right, as in every other regard, that would be incorrect.

Half a video card for half a thousand. Makes sense.

Small PCBs are actively desirable for tightly integrated electronic circuits such as GPUs. It's also excellent that they made it a physical x8 instead of a x16 electrically wired to just x8. Far better signal integrity means a better behaved, more efficient and easier to cool board.

That's funny. You know what else is funny? Despite its diminutive PCB, they still felt the need to make a dual slot design. The price might be amusing too if it weren't so sad,....

TDP is still too high. Probably doable with a 5060 (something similar to the Katana 1070 from Galax). That was a 150 watt card, and even then it was very loud if you ever saw one in person.

Judging from the BIll of Materials I think the graphic card could be in the 100€ range.

It most likely is. However - you're paying for R&D, long-term driver development and maintenance, manufacturing and distribution costs, amongst others.

 

[theglaze]

5090 FE continues to amaze

 

[boomheadshot8]

500 to 600€ in france : too high considered just a bit more perf than 3070 with 8go more Vram

 

[bonehead123]

Wonder why they didn't just put the fans & sink right on top of the pcb and make it a 2.5 or 3 slot card, and save all that extra metal & plastic that was wasted for the rest of the frame....

THAT would have been a true "Stubby" card for sure

 

[hsew]

Hope to see low-profile versions of this soon.

 

[Hxx]

Missed opportunity to make it a one fan design to carter to the sff crowd

 

[mechtech]

Holy stubby!!!

Holy thermal paste!!

 

[Chaitanya]

Holy thermal paste!!

Thermal putty.

 

[Dr. Dro]

Missed opportunity to make it a one fan design to carter to the sff crowd

Most modern SFF cases conform to the Nvidia SFF-Ready specifications nowadays, and those fit any "normal" size 3-fan models

Build Small, Play Big – Introducing SFF-Ready Enthusiast GeForce Cards & Compatible Cases

Size guidelines ensure you can build with confidence, knowing your GeForce RTX card will fit inside a SFF case.

www.nvidia.com

AMD GPUs are also available at that form factor

 

[csendesmark]

So cute

 

[lexluthermiester]

180W is not easily dissipated. 1x fan cards usually have high temps and are loud.

That depends on the cooler and fan size. As a rule, single slot cards aren't any louder than dual slot cards.

 

[cal5582]

That's funny. You know what else is funny? Despite its diminutive PCB, they still felt the need to make a dual slot design. The price might be amusing too if it weren't so sad,....

man id really like a replacement for my dualslot 4070 ti. but i dont think im going to find one.

 

[Yashyyyk]

180W is not easily dissipated. 1x fan cards usually have high temps and are loud.

TDP is still too high. Probably doable with a 5060 (something similar to the Katana 1070 from Galax). That was a 150 watt card, and even then it was very loud if you ever saw one in person.

Most modern SFF cases conform to the Nvidia SFF-Ready specifications nowadays, and those fit any "normal" size 3-fan models

Build Small, Play Big – Introducing SFF-Ready Enthusiast GeForce Cards & Compatible Cases

Size guidelines ensure you can build with confidence, knowing your GeForce RTX card will fit inside a SFF case.

www.nvidia.com

AMD GPUs are also available at that form factor

We had 180W R9 Nano almost 10 years ago, this will be same TDP in a >30% larger card, and probably worse than the similar (lol) single single fan RTX 4070

Likewise Katana was about 1/4 TDP and 1/4 size of 5090 FE, but 8 years ago. See this text from Techspot's review; they say it's not too loud and it beats larger 1070 FE lol

Besides thinking that it would run warmer and louder than most GTX 1070s, I wasn't sure what to expect from the Katana's performance in this category so I was pleasantly surprised by the card's 82 degree result, which was a degree under the Founders Edition card despite providing slightly better performance. Of note, the card wasn't too noisy out of the box but maintaining that 82 degrees after our overclock required us to increase the fan speed to a pretty loud setting.

Another example is 5070 FE design, we had Zotac 1080Ti Mini 8 years ago, same TDP, same volume. Admittedly ignoring thermals / noise, you would think over 8 years later we would get better designs not same/stagnation.

Apparently the 5070 cooler is worse than the 4070 (S?) FE since one of the fans don't even work properly lol


I have the Gigabyte ITX 1080, it's really not loud, maybe my temps are closer 80 on certain games

There is no GPU I can upgrade to with more VRAM at the same size besides 3060 12GB, the Zephyr 4070, and Colorful's 4060Ti 16GB (for some reason cooling +8GB VRAM on 4060Ti 16GB requires an extra fan with other models? )

Perhaps the problem is the shrinking die space over the years...but I still don't think OEMs are trying hard enough in the first place

It's really sad

It would be interesting if you had some sort of size-TDP-temp-noise normalized testing to refer to


I think Nvidia's SFF-Ready is a joke, it doesn't differentiate between large 3 fan 4060s and Nvidia's own excellent 5090 cooler, why does 5x less TDP require more volume to cool lol?

https://www.reddit.com/r/sffpc/comments/12ne6d7

 

[Dr. Dro]

We had 180W R9 Nano almost 10 years ago, this will be same TDP in a >30% larger card, and probably worse than the similar (lol) single single fan RTX 4070

Likewise Katana was about 1/4 TDP and 1/4 size of 5090 FE, but 8 years ago. See this text from Techspot's review; they say it's not too loud and it beats larger 1070 FE lol

What you're not accounting for is that the R9 Nano was easy to cool because the Fiji core is a monstrous slab of silicon. It's probably thrice as large as this chip, if not larger, which made heat dissipation a lot easier. Heat density, not merely the TDP in itself, is what makes hardware difficult to cool. I believe it beating the FE of the time, though, since Galax used a large, pure copper heatsink on the Katana.

As for the Blackwell FE cooler... well, it's flow-through and a lot of engineering was put into it. It's also known to be a lot noisier than the other traditional axial models.

 

[Yashyyyk]

What you're not accounting for is that the R9 Nano was easy to cool because the Fiji core is a monstrous slab of silicon. It's probably thrice as large as this chip, if not larger, which made heat dissipation a lot easier. Heat density, not merely the TDP in itself, is what makes hardware difficult to cool. I believe it beating the FE of the time, though, since Galax used a large, pure copper heatsink on the Katana.

As for the Blackwell FE cooler... well, it's flow-through and a lot of engineering was put into it. It's also known to be a lot noisier than the other traditional axial models.

Perhaps the problem is the shrinking die space over the years...but I still don't think OEMs are trying hard enough in the first place

if Zephyr, a small niche company can make single (lol) single fan 4070, then my opinion is more and more that other OEMs are incompetent

nowadays, undervolt+PL often gets rid of >25% TDP for <5% performance loss. The same way there is "headroom" to OC or sell the cards already set at diminishing return TDP or v/f curve, there is (legroom?) to go smaller / more efficient

Here is another example:

Inno3D (while still compact), used X2 on 4070Ti Super (285W) but not on 5070Ti (300W), do they really need an extra fan for 15W? Same die size btw

Not much different than 4060Ti 8GB having single fan cards but 4060Ti 16GB only having the Colorful model...

 

[k_9virus]

that's $500+? lol wtf?

 

[Yashyyyk]

Here is another one of my copypastas of sorts

It's depressing to me to see very tangible improvements in CPU coolers but not in GPU coolers...Imagine if over 10 years your CPU cooler options became 2x larger / cooled 2x less...

and people still called it a "good design"" because it fits the case.

(I am referring to the SFF coolers, the improvement from like Scythe BS3 or Noctua L9 to now AXP120 67 or Noctua L12S, same size. Though I would imagine there has been improvements in CPU coolers in general over time
And if you use copper the full copper thermalright coolers are quite smaller than those for only a little worse cooling performance)

Imagine if like 2015 Alienware x14 or some other older smaller laptop cooled better than a 2025 16" laptop (solely talking about cooling, not performance), and/or there was no 14" option at all in 2025 to buy

 

[billEST]

We had 180W R9 Nano almost 10 years ago, this will be same TDP in a >30% larger card, and probably worse than the similar (lol) single single fan RTX 4070

Likewise Katana was about 1/4 TDP and 1/4 size of 5090 FE, but 8 years ago. See this text from Techspot's review; they say it's not too loud and it beats larger 1070 FE lol




Another example is 5070 FE design, we had Zotac 1080Ti Mini 8 years ago, same TDP, same volume. Admittedly ignoring thermals / noise, you would think over 8 years later we would get better designs not same/stagnation.

Apparently the 5070 cooler is worse than the 4070 (S?) FE since one of the fans don't even work properly lol


I have the Gigabyte ITX 1080, it's really not loud, maybe my temps are closer 80 on certain games

There is no GPU I can upgrade to with more VRAM at the same size besides 3060 12GB, the Zephyr 4070, and Colorful's 4060Ti 16GB (for some reason cooling +8GB VRAM on 4060Ti 16GB requires an extra fan with other models? )

Perhaps the problem is the shrinking die space over the years...but I still don't think OEMs are trying hard enough in the first place

It's really sad

It would be interesting if you had some sort of size-TDP-temp-noise normalized testing to refer to


I think Nvidia's SFF-Ready is a joke, it doesn't differentiate between large 3 fan 4060s and Nvidia's own excellent 5090 cooler, why does 5x less TDP require more volume to cool lol?

https://www.reddit.com/r/sffpc/comments/12ne6d7

the problem is not the PCB card , the problem is like processor is the size of GPU : you have for exemple 10 mm 180 W in new gpu , 200 mm 180 W old : you can put the new in small pcb BUT its more hard to dissipate : 3 solution water , better air or normal air ... the less expensi is normal 30 cm air , more over 30 cm air can be use for other GPU :all problem of price and margin