Report an Error

NVIDIA A16 PCIe

Graphics Processor
GA107 x4
Cores
2560 x4
TMUs
80 x4
ROPs
48 x4
Memory Size
16 GB x4
Memory Type
GDDR6
Bus Width
128 bit x4
The A16 PCIe is a professional graphics card by NVIDIA, launched on April 12th, 2021. Built on the 8 nm process, and based on the GA107 graphics processor, the card supports DirectX 12 Ultimate. A16 PCIe combines four graphics processors to increase performance. It features 2560 shading units, 80 texture mapping units, and 48 ROPs, per GPU. Also included are 80 tensor cores which help improve the speed of machine learning applications. The card also has 20 raytracing acceleration cores. NVIDIA has paired 64 GB GDDR6 memory with the A16 PCIe, which are connected using a 128-bit memory interface per GPU (each GPU manages 16,384 MB). The GPU is operating at a frequency of 885 MHz, which can be boosted up to 1695 MHz, memory is running at 1812 MHz (14.5 Gbps effective).
Being a dual-slot card, the NVIDIA A16 PCIe draws power from an 8-pin EPS power connector, with power draw rated at 250 W maximum. This device has no display connectivity, as it is not designed to have monitors connected to it. A16 PCIe is connected to the rest of the system using a PCI-Express 4.0 x16 interface. The card measures 267 mm in length, 112 mm in width, and features a dual-slot cooling solution.

Graphics Processor

GPU Name
GA107
Architecture
Ampere
Foundry
Samsung
Process Size
8 nm
Transistors
unknown
Die Size
unknown

Graphics Card

Release Date
Apr 12th, 2021
Availability
2021
Generation
Tesla
(Axx)
Production
Active
Bus Interface
PCIe 4.0 x16

Relative Performance

2%
GeForce 210
3%
GeForce 9400 GT
3%
Radeon HD 4550
4%
Radeon HD 5450
5%
Radeon HD 6450
5%
GeForce GT 520
6%
GeForce GT 220
8%
GeForce GT 430
8%
Radeon HD 5570
9%
Radeon HD 4670
9%
GeForce GT 440
10%
GeForce GT 240
11%
GeForce 9600 GT
12%
Radeon HD 5670
13%
GeForce GT 640
13%
GeForce 9800 GT
14%
Radeon HD 6670
14%
Radeon HD 4830
15%
Radeon HD 4770
16%
Radeon HD 4850
16%
GeForce GTS 250
17%
GeForce GTS 450
17%
Radeon HD 5750
20%
Radeon HD 7750
20%
Radeon HD 5770
20%
GeForce GTX 260
20%
Radeon HD 4870
20%
Radeon Vega 8
21%
GeForce GTX 550 Ti
21%
GeForce GTX 650
22%
Radeon HD 5830
23%
Radeon HD 6790
23%
Radeon HD 4890
24%
GeForce GTX 460
24%
Radeon RX Vega 11
25%
GeForce GT 1030
25%
GeForce GTX 275
25%
Radeon HD 7770 GHz Edition
25%
GeForce GTX 280
26%
GeForce GTX 465
27%
Radeon HD 6850
27%
GeForce GTX 285
29%
Radeon HD 5850
30%
Radeon HD 7790
30%
GeForce GTX 650 Ti
31%
Radeon RX 550
31%
GeForce GTX 470
32%
Radeon HD 6870
34%
Radeon HD 5870
36%
GeForce GTX 560 Ti
36%
Radeon HD 4870 X2
36%
GeForce GTX 750 Ti
37%
Radeon HD 6950
38%
GeForce GTX 295
38%
GeForce GTX 650 Ti Boost
39%
Radeon HD 7850
39%
GeForce GTX 480
41%
Radeon HD 6970
42%
Radeon R7 265
43%
Radeon RX 460
43%
Radeon R7 370
43%
GeForce GTX 660
43%
GeForce GTX 570
44%
Radeon RX 560
46%
GeForce GTX 950
47%
Radeon HD 7870 GHz Edition
48%
Radeon R9 270X
48%
GeForce GTX 660 Ti
49%
GeForce GTX 580
49%
GeForce GTX 1050
49%
Radeon HD 7950
50%
Radeon HD 5970
51%
GeForce GTX 760
55%
GeForce GTX 670
57%
GeForce GTX 960
58%
Radeon R9 380
59%
Radeon R9 285
59%
Radeon HD 7970
60%
GeForce GTX 680
61%
GeForce GTX 1050 Ti
63%
GeForce GTX 770
64%
Radeon HD 6990
66%
Radeon R9 280X
67%
GeForce GTX 590
67%
Radeon HD 7970 GHz Edition
74%
GeForce GTX 780
77%
GeForce GTX 1650
81%
Radeon RX 470
83%
Radeon R9 290
87%
Radeon RX 570
88%
Radeon R9 390
89%
Radeon R9 290X
90%
GeForce GTX 970
91%
GeForce GTX TITAN
92%
GeForce GTX 780 Ti
93%
Radeon R9 390X
94%
Radeon HD 7990
94%
Radeon RX 480
95%
GeForce GTX 690
98%
GeForce GTX 1060 6 GB
99%
Radeon RX 5500 OEM
100%
A16 PCIe
100%
Radeon RX 580
101%
Radeon RX 5500 XT
102%
GeForce GTX 980
103%
GeForce GTX 1650 SUPER
104%
Radeon R9 FURY
108%
Radeon RX 590
114%
Radeon R9 295X2
115%
GeForce GTX 1660
116%
Radeon R9 FURY X
118%
GeForce GTX 980 Ti
121%
GeForce GTX TITAN X
129%
GeForce GTX 1660 SUPER
132%
GeForce GTX 1070
132%
GeForce GTX 1660 Ti
140%
Radeon RX Vega 56
149%
GeForce GTX 1070 Ti
151%
Radeon RX 5600 XT
153%
Radeon RX Vega 64
155%
GeForce RTX 2060
155%
GeForce GTX 1080
161%
Radeon RX 5700
174%
GeForce RTX 2060 SUPER
180%
GeForce RTX 2070
184%
GeForce RTX 3060
185%
Radeon RX 5700 XT
187%
Radeon VII
192%
TITAN X Pascal
197%
GeForce GTX 1080 Ti
200%
GeForce RTX 2070 SUPER
206%
Radeon RX 6600 XT
212%
GeForce RTX 2080
225%
GeForce RTX 2080 SUPER
230%
GeForce RTX 3060 Ti
234%
Radeon RX 6700 XT
253%
GeForce RTX 2080 Ti
270%
GeForce RTX 3070
289%
GeForce RTX 3070 Ti
297%
Radeon RX 6800
341%
Radeon RX 6800 XT
351%
GeForce RTX 3080
371%
Radeon RX 6900 XT
393%
GeForce RTX 3080 Ti
397%
GeForce RTX 3090
Based on TPU review data: "Performance Summary" at 1920x1080, 4K for 2080 Ti and faster.
Performance estimated based on architecture, shader count and clocks.

Clock Speeds

Base Clock
885 MHz
Boost Clock
1695 MHz
Memory Clock
1812 MHz
14.5 Gbps effective

Memory

Memory Size
16 GB
Memory Type
GDDR6
Memory Bus
128 bit
Bandwidth
231.9 GB/s

Render Config

Shading Units
2560
TMUs
80
ROPs
48
SM Count
20
Tensor Cores
80
RT Cores
20
L1 Cache
128 KB (per SM)
L2 Cache
2 MB

Theoretical Performance

Pixel Rate
81.36 GPixel/s
Texture Rate
135.6 GTexel/s
FP16 (half) performance
8.678 TFLOPS (1:1)
FP32 (float) performance
8.678 TFLOPS
FP64 (double) performance
271.2 GFLOPS (1:32)

Board Design

Slot Width
Dual-slot
Length
267 mm
10.5 inches
Width
112 mm
4.4 inches
TDP
250 W
Suggested PSU
600 W
Outputs
No outputs
Power Connectors
8-pin EPS

Graphics Features

DirectX
12 Ultimate (12_2)
OpenGL
4.6
OpenCL
3.0
Vulkan
1.2
CUDA
8.6
Shader Model
6.6
Copyright © 2004-2021 www.techpowerup.com. All rights reserved.
All trademarks used are properties of their respective owners.