Quadro RTX 4000 VS RTX 2070: Hashrate Memory| Power Consumption (Who Win)

Quadro RTX 4000 VS RTX 2070: Hashrate Memory| Power Consumption (Who Win): Quadro RTX 4000: The smallest Quadro with Turing is a slower RTX 2070 – In terms of computing power and price, Nvidia today introduced the smallest version of the Quadro RTX series. The professional graphics card Quadro RTX 4000 uses the same Turing GPU as the consumer model GeForce RTX 2070. The chip and memory, however, have a slightly lower clock rate.

Contents hide

1 Nvidia Quadro RTX 4000 with a chip from the GeForce RTX 2070

2 Areas of application

3 Price and availability

3.1 Related posts:

Nvidia Quadro RTX 4000 with a chip from the GeForce RTX 2070

The TU106 GPU is fully equipped on both the Quadro RTX 4000 and the GeForce RTX 2070 with 2,304 FP32 ALUs, 288 tensor cores and 36 ray tracing cores. With 7.1 TFLOPS, however, the FP32’s performance is below the 7.5 TFLOPS of the GeForce RTX 2070 (test). Mathematically, this results in a clock rate of around 1,540 MHz, the RTX 2070 clocks at 1,620 MHz.

Read This Now: GTX 1660 and 1660 Ti mining benchmarks: tuning, overclocking and hashrate

The Quadro RTX 4000 also lags behind the RTX 2070 in terms of memory bandwidth. Nvidia speaks of “up to 416 GB / s“, Which results in a clock rate of up to 6,500 MHz for the 8 GB GDDR6 memory with the 256-bit memory interface. On the other hand, the RTX 2070 achieves 448 GB / s at 7,000 MHz.

The new Quadro graphics cards in comparison

	RTX 8000 panel	RTX 6000 panel	RTX 5000 panel	RTX 4000 panel	GeForce RTX 2070
architecture	Turing
GPU	TU102	TU102	TU104	TU106	YOUR 106
FP32-ALUs	4.608	4.608	3.072	2.304	2.304
FP32 performance	16,3 TFLOPS	16,3 TFLOPS	11,2 TFLOPS	7,1 TFLOPS	7,5 TFLOPS
Tensor Core	576	576	384	288	288
RT cores	72	72	48	36	36
GigaRays/s	10	10	8	6	6
Clock rate (calculated)	~1.770 MHz	~1.770 MHz	~1.820 MHz	~1.540 MHz	1.620 MHz
Memory expansion	48 GB GDDR6	24 GB GDDR6	16 GB GDDR6	8 GB GDDR6	8 GB GDDR6
Memory clock	7.000 MHz	7.000 (?) MHz	7.000 (?) MHz	up to 6,500 MHz	7.000 MHz
Memory interface	384 Bit	384 Bit	256 Bit	256 Bit	256 Bit
Memory bandwidth	672 GB/s	672 GB/s (?)	448 GB/s (?)	“Up to 416 GB / s”	448 GB/s

In the absence of details on the definition, the direct comparability of the information on power consumption is questionable. But Nvidia speaks of 125 watts for the “Graphics Power Consumption“The Quadro RTX 4000. The value is well below the 175 watts or 185 watts (Founders Edition) that Nvidia calls for the GeForce RTX 2070. A lower power consumption of the Quadro model can be expected solely from the reduced clock rate.

The requirements for the cooling system, which only occupies one slot in the Quadro RTX 4000, should be lower.

Read This Now: What is mining and what is it for - End of mining in 2019

Areas of application

Nvidia sees the professional graphics cards of the Quadro RTX family among other things for “Experts from the fields of manufacturing, architecture, mechanical engineering and media design” before. The new computing units, RT cores and tensor cores of the Turing architecture should also create new possibilities in this segment in terms of real-time ray tracing and artificial intelligence.

Price and availability

Nvidia calls one “estimated street price“Of 900 US dollars for the Quadro RTX 4000. The graphics card will be available from December on both the Nvidia website and from authorized sales partners such as PNY and from workstation providers such as Dell, HP and Lenovo.

The Quadro RTX 4000 has a lot in common with the civilian GeForce RTX 2070 . The manufacturer used the Turing TU106 graphics chip with 2304 CUDA cores, 288 Tensor cores and 26 RT units, as well as 8 GB of GDDR6 256-bit memory.

Read This Now: Download TeamRedMiner 0.7.3: setup, installation - Easily Done

Model	Nvidia Quadro RTX 4000	Nvidia Quadro RTX 5000	Nvidia Quadro RTX 6000	Nvidia Quadro RTX 8000
Graphic layout	Turing TU106	Turing TU104	Turing TU102	Turing TU102
CUDA cores	2304	3072	4608	4608
Computing power	7,1 TFLOPS	11,2 TFLOPS	16,3 TFLOPS	16,3 TFLOPS
Tensor Cores	288	384	576	576
Ray-Tracing	6 GigaRays/sec	6 GigaRays/sec	10 GigaRays/sec	10 GigaRays/sec
Video memory	8 GB GDDR6 256-bit	16 GB GDDR6 256-bit	24 GB GDDR6 384-bit	48 GB GDDR6 384-bit
Memory bandwidth	do 412 GB/s	448 GB/s	672 GB/s	672 GB/s
price	$900	$2300	$6300	~$10 000

The new card works with lower clock speeds (around 1540 MHz for the core and up to 6500 MHz for memory), which allowed to reduce the demand for electricity and cooling requirements – the TDP factor is only 125 W.

The Quadro RTX 4000 uses single-slot cooling

The Quadro RTX 4000 is expected to go on sale in December with an estimated price of $ 900. The new card should be of particular interest to architects, designers and engineers.