AMD Teases Polaris 11 and Polaris 10 GPUs - Reconstructed Die Shots Reveal 232mm² Size For The Flagship: Polaris 10

AMD recently launched a brand new page dedicated to the upcoming Polaris architecture and posted some pictures teasing the new GPUs. Our collegues over at Videocardz were able to make quick work of it and produce a die shot as well every bit an judge of the judge die size of the flagship Polaris GPU. The size corresponds to the number we heard nigh three few months ago: 232mm² large dice for Polaris 10.

AMD teases die shots of Polaris 10 and Polaris 11

AMD has teased die shots of its upcoming GPUs in the by, Fiji being the near contempo example and they were successfully analyzed to give an accurate representation of not only the die but its components every bit well (anyone remember the fourth dimension HBM was still i of our "rumors"). In what appears to exist history repeating itself, AMD has published teasers of both Polaris ten advertizement Polaris 11 on its new folio. Given below are the die reconstructions courtesy of Videocardz:

The fractional but clear, reconstructed shot is that of Polaris 10 (aka Ellesmere XT). The die appears to exist roughly 232mm² which is the exact same number we saw before. This makes it highly unlikely that we are looking at sheer coincidence and probably means that Polaris 10 GPU is, in fact, of that size. We are talking well-nigh the Samsung/GloFo 14nm LPP process here, and so the size would exist optimal in terms of yield and power efficiency. The second dice (blurry and full shot) is Polaris eleven (aka Baffin XT). Polaris 11 is the smaller GPU which will be targeting the mobile GPU market and should offering the sweet spot of the Polaris efficiency claim (2.5x increment in perf/watt).

AMD Vega 10, Polaris 10, Polaris 11 Feature Not to scale.

According to the data we accept nearly the 14nm LPP procedure, and based on transistor density increase, a 232mm² GPU would be roughly equivalent to a 464mm² 28nm processor – at the same TDP levels. Since we already know that AMD is going to be focusing not just on functioning but power efficiency as well – this number could be exist much higher, in fact nosotros will talk over the number AMD is using below. We can withal safely say that this dice is more capable of meeting the 'minimum VR spec that AMD promises.

@Gchip Polaris is two.v times brighter today than when Ptolemy observed it in 169 A.D https://t.co/Qgf6C89dSN

— Raja Koduri (@GFXChipTweeter) Nov 26, 2022

The more astute of our readers would find one inconsistency with this data. A 232mm² does not fit with the positioning of the minor Polaris die, which according to our estimates should be around 150mm² and doesn't fit perfectly with our image of big Polaris which should have been near the 300mm² marking. If you were to take a expect at the AMD slide given above and calculate the die shrinks associated with the 14nm jump – the number would exist exactly 2.v. If you lot recollect, Raja Koduri has previously tweeted well-nigh Polaris existence "two.5 times brighter" in the by – so the number 2.5 times is one AMD is throwing around alot.

The performance levels of a 580mm² can be gained by a 232mm² dice (at a very rough judge) if we are talking bout a proceeds of 2.v times in performance per watt. This is the theoretical height, and so the real number would probably be lower than this for Polaris 10 accounting for leakage when clocking at high rates. As well, we do not know what kind of ability efficiency mix AMD is playing with here so information technology would be pointless to speculate further about the exact performance levels at this point.

AMD R9 400 Serial 'Preliminary' Specifications

WCCFTech	Radeon R9 480X (Radeon RX 480)	Radeon R9 480	Radeon R9 470X	Radeon R9 470
GPU	Polaris x / Ellesmere XT	Polaris x / Ellesmere Pro	Polaris 11 / Baffin XT	Polaris 11 / Baffin Pro
Fabrication Process	14nm FinFET	14nm FinFET	14nm FinFET	14nm FinFET
Compute Units	twoscore	36	20	16
GCN Generation	GCN 4.0	GCN 4.0	GCN 4.0	GCN iv.0
Stream Processors	2560 SPs	2304 SPs	1280 SPs	1024 SPs
TMUs	TBA	TBA	TBA	TBA
ROPs	TBA	TBA	TBA	TBA
Memory Motorcoach	256-fleck	256-bit	128-bit	128-bit
Memory	8GB GDDR5/10	8GB GDDR5/X	4GB GDDR5/X	4GB GDDR5/X
TDP	~110-150W	~110-150W