THE COMPUTE FIGURES ARE IDENTICAL TO THE GEFORCE GTX 1660 AND THAT IS CORRECT, NOT A COPY DEFECT. Both cards run 1,408 CUDA cores at a 1,530 MHz base and a 1,785 MHz boost, so both come out at 5.03 TFLOPS FP32, 10.05 TFLOPS FP16 and 0.157 TFLOPS FP64. The SUPER is a memory upgrade and nothing else: 6 GB of GDDR6 at 14 Gbps for 336 GB/s where the plain card has 6 GB of GDDR5 at 8 Gbps for 192 GB/s, a 75 per cent bandwidth increase that NVIDIA's own launch article quantifies in exactly those terms. Confirmed side by side in NVIDIA's live compare table. NO TENSOR CORES AND NO RT CORES, as recorded on the GeForce GTX 1660 Ti row, so the FP16 figure is the shader rate and there is no TF32, BF16, FP8, INT8 or INT4 row. FP32 IS DERIVED from NVIDIA's published core count and boost clock; it publishes no TFLOPS figure for any 16-series card. REFUSED: transistor count, die size and L2 cache, none published for TU116.

NVIDIA GeForce GTX 1660 Super 6GB

Turing PCIe 2019 12nm Spec confidence: Vendor claimed
FP32
5
TFLOPS
Memory
6 GB
GDDR6
Bandwidth
336 GB/s
memory
TDP
125 W
40.2 FP32 dense TFLOPS/kW of TDP
Download datasheet

Overview

The GeForce GTX 1660 SUPER is the plain GTX 1660 with its memory replaced. The silicon is untouched, 1,408 CUDA cores at a 1,785 MHz boost for 5.03 TFLOPS FP32 and 10.05 TFLOPS FP16 on a TU116 die with no Tensor Cores and no RT Cores, but the 6 GB of GDDR5 becomes 6 GB of GDDR6 at 14 Gbps, taking bandwidth from 192 to 336 GB/s for five more watts. That is a strange-looking product and a sensible one: on this class of card almost nothing is arithmetic-bound, so buying bandwidth rather than shaders was the cheaper way to make it faster. For compute it means the SUPER and the plain 1660 will show identical peak throughput and behave quite differently on anything that streams weights, which is most inference. Its half-precision path runs at twice single precision, unlike every Pascal card, which is the one genuinely modern thing about it.

Performance

Peak theoretical throughput by precision type

PrecisionPeak
FP64
64-bit floating point
0.2TFLOPS
FP32
32-bit floating point
5TFLOPS
TF32
No verified data available
BF16
No verified data available
FP16
16-bit floating point
10TFLOPS
FP8
No verified data available
FP6
No verified data available
FP4
No verified data available
INT8
No verified data available

Every figure here is dense. The vendor documents no structured (2:4) sparsity mode for this part, so there is no second number to quote.

The GeForce GTX 1660 Super 6GB in the GPU landscape

Peak FP16 TFLOPS (dense) against TDP, single-GPU parts tracked by Flopper

06001,2001,8002,4003,0000 W250 W500 W750 W1000 W1250 W1500 WInstinct MI355XGeForce GTX 1660 Super 6GB

Higher and further left is better: more half-precision throughput for less power.

Specifications

Architecture

Turing

Form Factor

PCIe

Launch Year

2019

Process Node

12nm

Memory

6 GB GDDR6

Bandwidth

336 GB/s

TDP

125 W

Max power

125 W

CUDA Cores

1,408

Spec Confidence

Vendor claimed

Full Specifications

Compute Engine
CUDA Cores 1,408
Streaming Multiprocessors 22
Base Clock 1.53 GHz
Boost Clock 1.78 GHz
Memory
Memory 6 GB
Memory Type GDDR6
Bandwidth 336 GB/s
Interface Width 192-bit
Memory Clock 14 GT/s
Interconnect & I/O
PCIe 3.0 x16
Power & Thermal
TDP 125 W
Max power 125 W
Power Connector 8-pin
Cooling Active
Enterprise Features
ECC Memory No
Sparsity No
Compute APIs CUDA, DirectX 12, Vulkan, OpenGL
Physical & Media
Card Length 145 mm
Width 2-slot
General
Form Factor PCIe
Architecture Turing
Process Node 12nm
Launch Year 2019

Datasheet & Resources

Data Provenance

Every figure traced to a source

Primary Source

Publisher
NVIDIA
Published
No verified data available

Data Quality

Spec confidence
Vendor claimed
Clock basis
Boost
Core precisions with figures
3 of 9
Normalization
All values in TFLOPS

Compare cloud providers offering on-demand GPU instances for AI training, inference, and HPC workloads.

Browse GPU Cloud Providers

Similar GPUs

Frequently Asked Questions

How many TFLOPS does the NVIDIA GeForce GTX 1660 Super have?

The NVIDIA GeForce GTX 1660 Super delivers 5 TFLOPS FP32 and 10 TFLOPS FP16 at peak. Flopper does not currently have a verified FP8 throughput figure for it.

What is the power consumption of the NVIDIA GeForce GTX 1660 Super?

The NVIDIA GeForce GTX 1660 Super has a TDP (Thermal Design Power) rating of 125 watts.

How much memory does the NVIDIA GeForce GTX 1660 Super have?

The NVIDIA GeForce GTX 1660 Super is equipped with 6 GB of memory with 336 GB/s of memory bandwidth.

What architecture is the NVIDIA GeForce GTX 1660 Super based on?

The NVIDIA GeForce GTX 1660 Super is based on the Turing architecture, launched in 2019.

Stay Updated on GPU Releases

Get notified when new GPUs are added or specifications are updated.

Loading verification...

No spam, unsubscribe anytime.

Back to GPUs