Category - Artificial Intelligence

Google vs NVIDIA: Why TPUs are becoming a real threat to GPU supremacy

It may come as a surprise at first glance, but NVIDIA’s toughest competitor in the AI market is currently not AMD or Intel, but Google. While the latter is often perceived by the public as a cloud service provider or software company, it has been consistently developing its own chip strategy for years and with growing success. With the [...]


From 95% to 0%: NVIDIA’s diplomatic capitulation in the Chinese AI market

It is a drastic cut that is rarely admitted so openly in this form: NVIDIA CEO Jensen Huang speaks in no uncertain terms about the fact that his company has not only lost relevance in the Chinese market, but has been completely eliminated. What was once considered a golden growth pit for the Californian chip manufacturer is now just a diplomatic [...]


Gartner warns: No survival in the PC market without NPU

For a long time, it was just a marketing buzzword that was slapped onto every half-warm notebook campaign: “AI PC”. But now we seem to be at the point where marketing is becoming reality – at least if you believe the latest figures from Gartner. And unlike many other analysts, Gartner has so far mostly followed the development of [...]


Zhaoxin KH-50000 and KX-7000N, China’s digital battle cry

While the West continues to juggle price explosions in the GPU segment and hot AI air, China is drawing a card in the background that is potentially more far-reaching than any new GPU line-up: Zhaoxin, the joint venture between VIA Technologies and the Shanghai government, has unveiled two new processors, the KH-50000 and the KX-7000N, which are [...]


AMD SP7: 1,400 watts for a CPU, curse or progress?

The headline sounds almost like science fiction, but it is bitter reality: with the upcoming SP7 platform, AMD is planning server processors that will consume up to 1,400 watts of power – per socket, mind you. This doubles the TDP compared to the current SP5 platform, which is already at 700 watts with Zen 4-based EPYC processors. This leap [...]


NVIDIA Blackwell Ultra: The ExaFLOP accelerator for the AI factory age

With Blackwell Ultra, NVIDIA is setting a new benchmark in AI acceleration. While the original Blackwell chip was already considered a radical departure from previous GPU designs, Blackwell Ultra goes even further, with twice the transistor density of Hopper, an aggressive dual-die architecture and impressive performance scaling up to 15 PetaFLOPS [...]