Nvidia’s Tiny RTX 4000 SFF 20GB Delivers RTX 3070 Performance at 70W

There are a great deal of compact fashionable workstations that pack some fairly succesful CPUs, however in addition they do not have the area to accommodate a normal high-performance workstation-class graphics card. This normally limits them to entry-level GPUs with mediocre efficiency. For individuals who need a compact SFF workstation with extra graphics capabilities, Nvidia has launched a brand new ProViz-focused workstation. RTX 4000 SFF Island Generation (opens in new tab) graphics card. Some of the fascinating choices of current years packs a high-end GPU right into a low-profile type issue with solely 70W of energy consumption.

The Nvidia RTX 4000 SFF Ada card makes use of the corporate’s AD104 graphics processing unit with 6144 CUDA cores enabled (out of a complete of 7680 cores). This is identical GPU because the RTX 4070 Ti, however with fewer lively cores and the enhance frequency is capped at round 1560 MHz to cut back total card energy. Then again, the graphics card comes with 20GB of GDDR6 reminiscence with ECC that connects to the GPU utilizing a 160-bit interface, so there’s loads of reminiscence for workstation use.

The GPU comes with two NVENC encoders and two NVDEC decoders enabled, however Nvidia hasn’t touched on the total capabilities of those models. They need to be just like NVENC and NVDEC utilized in different Ada playing cards, and you may see the video encoding efficiency and high quality in our newest GPUs.

The GA104 chip on this configuration delivers the best single precision efficiency of 19.2 TFLOPS, making it theoretically akin to the GeForce RTX 3070. The very best RT efficiency has the best FP8/INT8 tensor efficiency of 44.3 TFLOPS and 306.8 TFLOPS/TOPS.


(Picture credit score: Nvidia)

Almost 20 FP32 TFLOPS is likely to be dwarfed by the overwhelming efficiency of Nvidia’s RTX 6000 Ada Technology or GeForce RTX 4090, however the RTX 4000 SFF is a low-profile dual-slot graphics card that may match, even match, nearly any desktop PC. there isn’t any spare auxiliary PCIe energy connector. Apparently, the RTX 4000 Ada’s 153/306.8 INT8 TFLOPS (sparse and sparse, respectively) efficiency may be very near Nvidia’s GeForce RTX 3090 Ti, which is each dearer and way more energy hungry.

Swipe to scroll horizontally
Nvidia RTX 40 Sequence Specs
Row 0 – Cell 0 GPU FP32 CUDA Cores FP32 TFLOPS INT8 TFLOPS Reminiscence Configuration TBP MSRP
GeForce RTX 4070 Ti AD104 7680 40 TFLOPS 160/320 TFLOPS 12GB 192-bit 21 GT/s GDDR6X 285W 799 {dollars}
GeForce RTX 4070 AD104 5888 (?) ? ? 12GB 192-bit 21 GT/s GDDR6X 250W(?) ?
RTX 4000 Island Technology AD104 6144 19.2 TFLOPS 153/307 TFLOPs 20GB 160bit 16 GT/s GDDR6 ECC 70W $1,250
GeForce RTX 3090 Ti GA102 10,752 40 TFLOPS 160/320 TFLOPS 24GB 384bit 20 GT/s GDDR6X 450W $1,999
GeForce RTX 3070 GA104 5888 20.31 TFLOPS 81/160 TFLOPS 8GB 256-bit 14 GT/s GDDR6 220W $499

#Nvidias #Tiny #RTX #SFF #20GB #Delivers #RTX #Efficiency #70W

Leave a Reply

Your email address will not be published. Required fields are marked *

Canva is rolling out a number of new features, including several AI-powered tools Previous post Canva is rolling out a number of new features, including several AI-powered tools
IntegrityNext raises $109M for a platform to audit supply chains for ESG compliance Next post IntegrityNext raises $109M for a platform to audit supply chains for ESG compliance