AMD Unveils Radeon RX 6000M GPUs for Next-Gen Gaming Notebooks

AMD Unveils Radeon RX 6000M GPUs for Next-Gen Gaming Notebooks

AMD has announced a suite of new mobile GPUs at Computex this week. The new cards, which are all based on its RDNA2 GPU family, will offer improved gaming performance and sizable Infinity Caches.

AMD retains the same tactic in mobile that we saw the company use on desktop, namely using less memory bandwidth and compensating for the decrease with a large L3 cache. The Radeon 6800M, 6700M, and 6600M will offer 192-bit, 160-bit, and 128-bit memory interfaces, respectively. They’re backed with a 96MB, 80MB, and 32MB L3 cache.

AMD Unveils Radeon RX 6000M GPUs for Next-Gen Gaming Notebooks

A few bits of information the PCMag chart above doesn’t cover. Board TGPs (Total Graphics Power) are 145W+, 80-135W, and 50-100W, respectively, and OEMs will have the option to tune their targets. The clock speeds above are accurate for a given power output, but AMD will sell these chips into a range of systems. TGP measures the total output of the graphics card, not just the GPU core. Second, memory clocks: All of these GPUs include GDDR6 at up to 16Gbps.

One difference that you need to keep in mind when comparing mobile GPUs from Nvidia and AMD is between the two companies’ new messaging. For years, both Nvidia and AMD maintained different products for mobile and desktop, with mobile products differentiated by the use of the “M” suffix or the word “Mobile.” Today, Nvidia products all carry the same branding, desktop or mobile.

AMD’s Radeon 6800 Mobile is roughly equivalent to the 6700 XT, with somewhat lower clocks and what looks like better overall binning. We can’t really compare with “145W+”, but the gap between the 6700M and the 6800M in other respects is modest. The 6700M offers most of the VRAM, memory bandwidth, and clock of the 6800M, and most of the Infinity Cache as well.

The stepdown from the 6700M to the 6600M is much sharper. The 6600M only offers 32MB of Infinity Cache and a 128-bit memory bus, though the GPU clock stays high. Again, we see that AMD chose to emphasize high clocks and narrow chips for RDNA2. It’d be interesting to know more about the thinking behind the design.

When we reviewed the 6700 XT earlier this year, our IPC comparison between the 6700 XT and 5700 XT suggested that the 6700 XT was very power efficient in comparison to its predecessor, but not at the clock speeds AMD was targeting. When we cut the 6700 XT down to 1.85GHz, or approximately the clock speed on our 5700 XT, power consumption dropped sharply — by over 100W. AMD is undoubtedly binning more aggressively more mobile compared with desktop, but we’re a little surprised the company didn’t opt for wider, slower chips to better compete with Nvidia.

AMD Unveils Radeon RX 6000M GPUs for Next-Gen Gaming Notebooks

AMD’s claims for performance wins are also pretty narrow, especially for benchmarks the company opted to publish itself. The gap between the 6600M and the RTX 3060 at 1080p looks a little larger, according to PCMag, but AMD isn’t arguing in favor of dramatic upsets here.

AMD’s Ryzen 5000 CPUs have been quite popular in desktop and mobile systems in 2021, so we’ll see if the advent of new RDNA2-powered GPUs leads to an uptick in all-Team Red systems.

Continue reading

Intel Launches New Xe Max Mobile GPUs for Entry-Level Content Creators
Intel Launches New Xe Max Mobile GPUs for Entry-Level Content Creators

Intel has launched a new consumer, mobile GPU — but it's got a very specific use-case, at least for now.

MIT Creates Battery-Free Underwater GPS
MIT Creates Battery-Free Underwater GPS

GPS radio signals dissipate quickly when they hit water, causing a headache for scientific research at sea. The only alternative is to use acoustic systems that chew through batteries. A team from MIT has devised a battery-free tracking technology that could end this annoyance.

Intel Details XPU Strategy, Launches New Server GPU, OneAPI Gold
Intel Details XPU Strategy, Launches New Server GPU, OneAPI Gold

Intel made a pair of announcements today regarding its OneAPI initiative and the launch of its first server GPU based on Xe graphics.

Nvidia Unveils Ampere A100 80GB GPU With 2TB/s of Memory Bandwidth
Nvidia Unveils Ampere A100 80GB GPU With 2TB/s of Memory Bandwidth

Nvidia announced an 80GB Ampere A100 GPU this week, for AI software developers who really need some room to stretch their legs.