Summary of "Was Ray Tracing a Scam?"
Was hardware-accelerated ray tracing a scam?
The speaker (Steve from Hardware Unboxed, referred to as “Harbor Unboxed” in the subtitles) argues that NVIDIA’s early ray tracing (RTX 20-series era) marketing promises didn’t match real-world results. While ray tracing can look impressive, the video claims it failed to become a mainstream “must-buy/must-enable” gaming feature due to limited game support, weak performance, and limited competitive value.
Key claims about NVIDIA’s ray tracing promises vs reality
-
2018 NVIDIA claims: At a major presentation (CEO Jensen Huang/Jensen Wong is referenced), NVIDIA emphasized that ray tracing would apply “the laws of physics,” removing months of “faking” lighting (e.g., shadow maps, reflection probes). NVIDIA also described the RTX 20 series as a huge generational leap.
-
Contradicting outcome: The speaker says ray tracing hasn’t transformed gaming even after ~8 years, and that many games either:
- don’t support ray tracing, or
- support it only as an optional toggle rather than something games rely on by default.
Performance and hardware limitations (RTX 20-series focus)
The RTX 20-series is described as not powerful enough for real-time ray tracing at acceptable frame rates.
Example expectations mentioned:
- In Control at 1080p with medium RT settings, performance often doesn’t reach 60fps.
What NVIDIA suggested at the time:
- Accept lowering resolution (e.g., 1080p is cited as common per Steam survey data)
- Lower other graphics settings
- Use DLSS
Why the speaker argues these solutions don’t help enough:
- Lowering overall visuals to enable a “premium” feature undermines ray tracing’s purpose.
- DLSS 1.0 (what was available then) is described as poor, while better versions (e.g., DLSS 2.0) arrived much later.
- Upscaling doesn’t eliminate the fundamental performance gap, and upscaling artifacts become more noticeable at the lower resolutions forced by RT demands.
Sparse early ray tracing game support (and underwhelming cases)
The speaker claims that 12 months after the RTX 20 release, only two games were strong candidates to enable ray tracing:
- Metro Exodus
- Control
Other early ray tracing examples (e.g., Battlefield V) are described as producing noisy and largely pointless reflections—effects that look impressive but don’t meaningfully improve gameplay.
Why ray tracing is argued to be especially bad for multiplayer/competitive play
The video emphasizes that many multiplayer shooters prioritize high FPS, and ray tracing can:
- reduce performance
- reduce visibility, making it harder to spot enemies
Examples cited:
-
Call of Duty: Modern Warfare: Ray-traced shadows are described as helpful in single-player, but worthless in multiplayer where the long-term mode is competitive.
-
Fortnite: The speaker argues the best experience is often the performance mode (reduced quality). Hardware RT may look great, but the performance hit—and reduced spotting/visibility—can make it a disadvantage.
The speaker includes gameplay/visibility comparisons showing that:
- Hardware RT can make players very difficult to see at distance or in dark scenes.
- Performance mode can offer better visibility and competitiveness, even if visuals look worse.
“Future” argument and timing skepticism
The speaker questions the “RT is the future” framing, arguing that:
- Multiple years later, RT is still optional, not baseline.
- Given the long hype window, multiplayer is expected to have adopted RT by default—but it hasn’t.
Poll/usage statistics: do gamers actually enable RT?
The speaker references:
- a prior viewer poll (~2 years earlier)
- a recreated version later (with fewer options)
Reported results:
- ~15% enable ray tracing at maximum settings
- ~7% use medium-to-low RT
- Many either use it sometimes or not at all; combined, the speaker estimates nearly 80% fall into “sometimes or never.”
Conclusion: A technology marketed as revolutionary is not broadly adopted, weakening the “must-have” marketing position.
“Scam” conclusion: what the speaker says went wrong
The video’s overall framing is that:
- Early RTX RT-capable GPUs were effectively not powerful enough for later, more demanding ray/path tracing experiences.
- Many RTX-era games are characterized as underwhelming tech-demo-style content rather than transformative mainstream titles.
- The core issue is a misalignment between marketing and customer outcomes, especially for gamers who bought RTX hardware specifically for ray tracing.
Main speakers/sources (as stated in the subtitles)
- Steve — host of Hardware Unboxed (referred to as “Harbor Unboxed” in subtitles)
- Jensen Wong / Jensen Huang — referenced as NVIDIA CEO in the 2018 presentation
- Tim — referenced as someone who made a related video about how RTX 2060 aged after ~6 years of ray tracing development
- Viewer community polls — via Hardware Unboxed audience voting
Category
Technology
Share this summary
Is the summary off?
If you think the summary is inaccurate, you can reprocess it with the latest model.