GPUs Latest news

AMD’s RDNA 4 laptop GPUs revealed: Radeon RX 9080M with 64 compute units and 16 GB memory

While AMD has been rather slow with the release of new graphics card architectures in the past, there finally seems to be some movement in the mobile sector. According to recent leaks published by @All_The_Watts and distributed via well-known sources such as Videocardz, six new RDNA-4-based laptop GPUs are in the starting blocks. If the usual schedules are to be believed, the unveiling should take place in time for Computex 2025.

Overview of the new models

The planned lineup is led by the Radeon RX 9080M. This is to be based on Navi-48 silicon and feature 64 compute units (CUs) and 16 GB of GDDR6 memory. AMD is therefore apparently positioning this GPU as a competitor to NVIDIA’s 70 and 80-class mobile GeForce models. However, it remains pure speculation at present as to what real performance values will ultimately emerge. Right behind it is the Radeon RX 9070M XT, also with a Navi-48 substructure, but with a slimmed-down 48 compute units. Here too, the memory remains identical at 16 GB, which is an interesting choice when you consider the typical gradations in previous generations. One level lower are the RX 9070M and RX 9070S, this time based on Navi 44. With 32 compute units and 8 GB of memory, these models should cover the mid-range segment – at least on paper. The range is rounded off by the RX 9060M and RX 9060S, which are probably at home in the lower mid-range with 28 compute units and also 8 GB of memory.

RDNA 4 in mobile use – a belated debut

Considering that AMD continues to rely on an interim solution in the form of RDNA 3.5 for APUs (Strix Halo and Medusa Point send their regards), the introduction of dedicated RDNA 4 GPUs in the laptop sector seems almost overdue. Apparently, AMD now wants to make up lost ground in the mobile segment with brute force, after NVIDIA and Intel had largely divided the last few quarters between themselves. The fact that AMD has set the minimum standard at 8 GB VRAM with the Radeon 9000M series could indicate a realization from the recent past: Less than 8 GB is simply no longer marketable in 2025, neither for games nor for productive software.

https://x.com/All_The_Watts/status/1915769679777534412

Open questions and initial assessments

The question of what efficiency AMD can achieve with the new chips remains interesting. While RDNA 3 occasionally offered mixed efficiency values in the desktop segment, every wasted watt second is a clear competitive disadvantage in the laptop sector. In addition, the coolability of the new GPUs remains exciting: depending on the implementation, the mobile Navi-48 chips in particular could require quite ambitious thermal management. There are also no definitive statements on production as yet. Rumour has it that an optimized version of TSMC’s 4nm technology will be used. Whether this will be enough to catch up with or even surpass the competition in terms of energy efficiency will only become clear after tangible benchmarks – and until then, everything remains pure guesswork as usual.

Computex 2025: Clear the stage for AMD?

Everything points to AMD officially unveiling the new mobile Radeon models at Computex 2025. The desktop version of the Radeon RX 9060 will probably also be unveiled at the same time, which will create a certain thematic bracket. It will be interesting to see how aggressive AMD can be in terms of pricing and availability, as the competition is known to never sleep – and has brought its own guns into position with Blackwell and Lunar Lake at the latest. Until then, however, the critical observer remains on standby with skepticism: too often early hopes for new generations have recently been confronted with harsh realities. So it remains to be seen whether AMD will actually deliver more than just a repetition of old promises in a new guise

Source: All_The_Watts via X, Videocardz

[the_ad_group id=”7834″

Kommentar

Lade neue Kommentare

e
eastcoast_pete

Urgestein

2,470 Kommentare 1,614 Likes

Solange die mobilen dGPUs nicht schon katastrophal schlechte Effizienz haben (wofür es bei RDNA4 keine Anzeichen gibt!) sind sehr niedrige Stromverbrauchswerte für dGPUs in Laptops eher nicht so wichtig. Typischerweise werden mobile dGPUs nicht oder nur für kurze Zeit exklusiv auf Akku betrieben. Allerdings ist gute Leistung, wenn man an der Strippe hängt, zentral.
Das ist sowohl für Gaming als auch Workstation Laptops sehr wichtig, denn eine mobile dGPU muss ihren Aufpreis dadurch rechtfertigen können, daß sie deutlich mehr Leistung bereitstellen kann als selbst eine gute iGPU.
Wenn AMD so wie in der Meldung geschrieben beide dGPUs mit 16 GB VRAM ausstatten wird, wär das IMHO sehr schlau. Nur die beiden Top mobilen Nvidia 5000er dGPUs haben 16 GB oder mehr VRAM, und Laptops mit einer mobilen 5080 oder 5090 sind auch entsprechend teuer. Wenn AMD jetzt noch das nahtlose Umschalten von einer Ryzen AI iGPU auf die RDNA4 dGPUs und zurück gut hinbekommt, hätten sie IMHO sehr gute Karten - in jeder Hinsicht 😄.

Antwort Gefällt mir

Corro Dedd

Urgestein

2,015 Kommentare 816 Likes

Ich würde mich jedenfalls freuen, wenn Framework neue GPU Module damit anbieten würde. So eine 9070M XT wäre meiner Meinung nach genau das richtige für das FW16, dürfte deutlich mehr Dampf haben, als das Modul mit der 7700S, aber nicht deutlich mehr Strom verbrauchen.

Antwort 1 Like

Danke für die Spende



Du fandest, der Beitrag war interessant und möchtest uns unterstützen? Klasse!

Hier erfährst Du, wie: Hier spenden.

Hier kannst Du per PayPal spenden.

About the author

Samir Bashir

Werbung

Werbung