NVIDIA sous le CHOC : La RX 10700 XT détruit la RTX 5080 !

NVIDIA sous le CHOC : La RX 10700 XT détruit la RTX 5080 !

NVIDIA in SHOCK: The RX 10700 XT destroys the RTX 5080!

Title history
  • NVIDIA sous le CHOC : La RX 10700 XT détruit la RTX 5080 ! — observed on 20261031-2319 UTC (YouTube)

by MAKERTRONIC TECH · View the channel on PSI

Applied Sciences & Engineering Computing & Cybersecurity UComputing and Information TechnologyUKComputer hardware
🎙 MAKERTRONIC TECH 👥 6K 📅 October 1, 2026 ⏱ 12 min 👁 37K 📄 news review 🧭 2026-10-31
Available in: English (current) Français

Keywords

RX 10700 XTRDNA 5RTX 5080AMDSony

Summary

This video from the French hardware channel MAKERTRONIC TECH discusses the latest leaks and rumors surrounding AMD’s upcoming Radeon RX 10700 XT, which is part of the RDNA 5 architecture (codenamed Alpha Trion). The presenter claims that preliminary data suggests this card could outperform the NVIDIA RTX 5080 by about 6.5–7% in raw rasterized rendering, thanks to a redesigned execution pipeline and a notable IPC improvement (5–10% at the same frequency). He explains that AMD is using TSMC’s 3nm (N3P) process, which allows higher clock speeds (3.1–3.4 GHz) and better energy efficiency. The GPU is said to feature only 64 compute units, yet it would rival or beat the RTX 5080, which is a major architectural shift. The video also highlights three key pillars of RDNA 5: Radiant Score (ray tracing), Neural Array (AI and FP8/FP6/FP4 support), and Universal Compression (30% data transfer reduction). Memory-wise, the RX 10700 XT is expected to come with 18 GB of GDDR7 and a 192-bit bus, yielding a bandwidth of 768–864 GB/s. The presenter also outlines the rest of the lineup: the 100900 XT (36 GB, 184-bit bus), the 10070 GRE (15–16 GB), and the 10060 XT (12 GB). Finally, he discusses the alliance between AMD and Sony, which aims to use these chips in the PlayStation 6, and concludes that AMD may be returning to the high-end market, although the price of memory (GDDR7) remains a concern.

234 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video offers a detailed, if speculative, overview of the rumored RX 10700 XT and RDNA 5 architecture. Its value lies in aggregating multiple leaks (compiler tracks, preliminary performance estimates) and explaining technical concepts like compute units, IPC, pixel pipelines, and memory bandwidth in an accessible way. The argumentation is coherent, but it is built almost exclusively on unverified information; the presenter adopts a cautious tone with repeated conditionals (‘should’, ‘would’, ‘allegedly’), which mitigates but does not eliminate the risk of misleading viewers. The technical reasoning about how a smaller number of compute units could achieve high performance is plausible, but without official benchmarks or confirmed specs, the claims remain highly uncertain. The video also includes a brief cost-benefit analysis, noting that the RTX 5080 is priced above €2000, which influences the competitive positioning.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is weak: the video does not cite any primary sources, official documents, or independent reviews. It relies on anonymous leaks and the presenter’s interpretation of compiler data, which is not peer-reviewed. The use of terms like ’leaks’ and ‘allegedly’ is honest, but the lack of verifiable references reduces the reliability of the information. The title is deliberately provocative (‘destroys’) and overstates the expected performance advantage, which is presented more moderately within the video itself. The content alignment is thus only partial, and the title may mislead viewers about the certainty of the claims. No external comments are considered in this section, as the instructions separate public feedback analysis.

260 words

Title / Content Match

The title is sensationalistic: it claims the RX 10700 XT 'destroys' the RTX 5080, while the video only mentions a theoretical 6.5–7% lead in raw performance, which is not a 'destruction'. Still, the content does focus entirely on this GPU comparison, so it is loosely aligned.

Quality & Reliability

5/10

This video relies entirely on unverified leaks and rumors about AMD's RDNA 5 architecture and the RX 10700 XT. No official sources are cited, and the presenter repeatedly uses conditional language. The technical explanations are plausible but speculative, and the performance claims are based on preliminary data from compiler tracks and the like, which are not confirmed by AMD or Sony.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Contribution & Novelties

The video provides a consolidated summary of the latest RDNA 5 and RX 10700 XT leaks, focusing on the surprising performance expectations relative to NVIDIA’s RTX 5080. It also highlights the AMD–Sony partnership as a potential game-changer for console and desktop GPUs. However, the information remains speculative, and the video does not provide any original analysis beyond interpreting the leaks.

Pour aller plus loin :

  • RDNA architecture (Wikipedia) — Background on AMD’s previous GPU architectures, useful to understand the evolution towards RDNA 5.
  • GDDR7 SDRAM (Wikipedia) — Technical details on the memory type mentioned in the video.
  • Project Amethyst (Search results) — If this exists; otherwise mention as a concept without URL. Since I’m not certain, I’ll skip the URL and just mention it as a concept.

127 words

Radar Profile

The radar profile is moderate: quantity of information is relatively high, but quality and reliability are low due to the speculative nature of the content. Technical level is average, with some deeper explanations, but the lack of verified sources keeps the overall fiabilité score low.

Reliability 4/10

💬 Skeptical. On the 30 comments analyzed, many viewers express doubt about AMD's ability to truly beat NVIDIA, citing years of unfulfilled promises, while a few are hopeful about performance and pricing.