Skip to main content
  1. Home
  2. Computing
  3. News

This GPU just beat the RTX 4090 — and Nvidia didn’t make it

Add as a preferred source on Google
The board of the RTX 4090 Super graphics card.
Teclab

Modders are doing what Nvidia won’t. The team at Teclab put together a Frankenstein graphics card, which it calls the RTX 4090 Super, that was able to beat the RTX 4090 by 13% in benchmarks.

You can’t buy this graphics card, of course, but it’s an interesting look into how splicing together the best components can deliver big performance gains. The heart of the RTX 4090 Super is the RTX 4090 GPU, which is Nvidia’s AD102 die. Teclab changed everything else about the graphics card, though.

Recommended Videos

For starters, it borrowed the memory modules from an RTX 4080 Super. The RTX 4080 Super actually has faster memory than the stock RTX 4090, which allowed the team at Teclab to reach 26 Gbps speeds on the GDDR6X modules. The stock RTX 4090 uses 21 Gbps modules. This change alone accounted for the 13% jump in performance without any additional overclocking.

O "monstro" em detalhes! 4090 "SUPER"

That’s just the start, though. Teclab took the AD102 die and placed it on the board of a Galax RTX 3090 Ti HOF OC Lab Edition. The board of this card is not only capable of handling high voltages thanks to to its 28 power stages, but it also comes with dual 16-pin connectors to feed the GPU with more power. Combine that with a high-end water block, and Teclab was able to overclock the GPU and optimize level of detail in the Superposition benchmark, leading to a 40% lead over the stock RTX 4090 in the benchmark.

Although Teclab achieved a score worthy of a top placement on the Superposition leaderboards, you won’t see the result online. With a new board, custom memory, and only the GPU tied back to the RTX 4090, it’s hard to call this an RTX 4090 at all.

The RTX 4090 Super installed in a motherboard.
Teclab

It’s an interesting showcase of what Nvidia could have achieved with an RTX 4090 Super, however. At the time when Nvidia introduced its RTX 40-series Super cards, the RTX 4090 was selling for around $2,000. A Super refresh of the RTX 409o running upwards of 13% faster would surely inflate the price even more, which is exactly what we saw with the RTX 4080 Super when it released.

The RTX 4090 Super experiment shows some promise for the next generation of Nvidia GPUs as well. Nvidia’s upcoming RTX 50-series cards are said to come with GDDR7 memory. Micron, the manufacturer of these memory modules, says that they’ll offer up to a 30% performance improvement in games. Even if that 30% number is optimistic, Teclab’s monster RTX 4090 Super shows that memory speed can have a big impact on performance.

Jacob Roach
Former Lead Reporter, PC Hardware
Jacob Roach is the lead reporter for PC hardware at Digital Trends. In addition to covering the latest PC components, from…
The Mac Pro nearly received an M3 Extreme chip twice as powerful as M3 Ultra
High production costs likely killed Apple’s M3 Extreme plans
Apple's Mac Pro on a table at a press event.

Apple discontinued the Mac Pro earlier this year, ending a 20-year run for a computer that once represented the very best of the company’s desktop lineup. However, Apple reportedly had much bigger plans for the machine before ultimately replacing it with the Mac Studio.

According to Bloomberg’s Mark Gurman, Apple developed an M3 Extreme chip that could have offered twice as many CPU and GPU cores as the M3 Ultra. The processor was intended to sit above the Ultra tier and could have finally given the Mac Pro the performance advantage it badly needed. Apple eventually abandoned the chip due to concerns over production costs and limited demand for such an expensive machine.

Read more
Hidden prompts can secretly rewrite an AI’s memory, and researchers say that’s a serious problem
Researchers discover AI attack that rewrites an assistant's long-term memory
Chatbot on a smartphone.

Large language models are getting better at remembering us. Whether it's your preferred writing style, recurring tasks, shopping habits or project deadlines, AI assistants are increasingly storing long-term memories to make future conversations feel more personal and useful. But according to new research, that same feature could become one of AI's biggest security vulnerabilities.

Researchers from New Mexico State University have demonstrated a new attack called GhostWriter, capable of secretly planting false memories inside AI agents. Rather than stealing information outright, the attack manipulates what an AI remembers, potentially causing it to make dangerous decisions long after the original attack has taken place.

Read more
This experiment shows how easy it is to poison an open-weight AI model for under $100
This research raises new doubts about trusting open weight AI models.
Computer, Electronics, Laptop

Open-weight AI models have been having a moment lately. Just this month, Moonshot's massive Kimi K3 model landed close behind Claude Fable 5 and GPT 5.6 Sol in several benchmarks, all while remaining fully open-weight and downloadable by anyone.

However, Katie Paxton-Fear, a cybersecurity lecturer at Manchester Metropolitan University and staff security advocate at Semgrep, managed to poison an open-weight model and proved how easily that openness can be turned against you (via The Register).

Read more