The RTX 4090 Clocks at 3.1GHz While Consuming 492W Of Power

NVIDIA’s RTX 4090 is just a few days away from us. The Lovelace kingpin is set to retail by the 12th of October with reviews going live 1 day before. The RTX 4090 is quoted as a 2-4x improvement against the RTX 3090 Ti, which is the theoretical maximum. Initial tests show that the RTX 4090 is 90% faster than last-gen in synthetic benchmarks. DLSS 3.0 will play a key role in achieving that 2-4x target in games. 

4090 Clocked At 3.1GHz

Wxnod on Twitter shared 2 images of the RTX 4090 in GPU-Z. The 4090 manages to attain a clock speed of 3.1GHz which is the highest we’ve seen, so far. Other than that, it does consume more (+40W) power than the stock TDP (450W). This is expected behavior since the 4090 can consume upwards of 660W

In the first test, the RTX 4090 clocks at 3.135GHz and the GPU temperature stands at 68.2C. This temperature figure is still relatively low and explains the beefy and bulky design used for these GPUs. The 4090 does consume nearly 500W of power which is slightly on the higher end. For those concerned about their electricity bills, this equates to 1 unit of electrcitiy (1kWh) after just 2 hours of use. 

RTX 4090 At 3.1GHz | Wxnod

Another picture shared by wxnod showcases the 4090 clocked at 2.73GHz. This is almost 400MHz lower than the last test, albeit the GPU does consume less power. 

RTX 4090 At 2.7GHz | Wxnod

The RTX 4090

The RTX 4090 is powered by the AD102 GPU, which is unparalleled in terms of performance. The GPU features a massive 24GB of Samsung-made GDDR6X memory running at 24Gbps for 1152GB/s of effective bandwidth. Packed with 16384 CUDA Cores, this powerhouse is home to 75 billion transistors in total, more than 2.5x than the RTX 3090 Ti. A price tag of $1599 for the 4090 has had its fair share of criticism from the community. Read more about the 4090 here

What Does AMD Offer

NVIDIA’s numbers are good for a generation-to-generation leap. However, AMD is aiming at 4.0GHz for their top-notch silicon. Navi 31 in synthetic testing is almost on par with the RTX 4090 (in Time Spy Extreme). AMD is not looking to beat NVIDIA in raw performance, but in the price-to-performance category. It will be interesting to see if AMD can pull a Ryzen on the GPU market this time.

SKUChipFP32/CUDAMax ClockCacheMemory BusVRAMMemory SpecSpeed (Gbps)TDP
RTX 4000 Titan?AD102-450181763.0GHz+?96MB?384/382-bit48GBGDDR6X24~800W
RTX 4090 TiAD102-350184323.0 GHz?96MB382-bit24GBGDDR6X24600W?
RTX 4090AD102-300-A1163842.52GHz96MB384-bit24GBGDDR6X21450W+ (TGP) / 660W (Max TGP)
RTX 4080 TiAD10214848?2.7 GHz?80MB?320-bit20GBGDDR6X23420W
RTX 4080 (Variant 1)AD103-300-A197282.505GHz64MB256-bit16GBGDDR6X22.5320W(TGP)/ 516W (Max TGP)
RTX 4080 (Variant 2)AD104-400-A176802.61GHz48MB192-bit12GBGDDR6X21285W (TGP) /366W (Max TGP)
RTX 4070 TiAD104-300?76803.0GHz??192-bit12GBGDDR6X21300W?
RTX 4070 AD104-275?7168??160-bit10GBGDDR6X21250W
ABOUT THE AUTHOR

Abdullah Faisal


With a love for computers since the age of five, Abdullah has always sought to delve into the depths of information, and uses it as his guiding light. He believes success is of utmost importance as history is written by the victor.
Back to top button