facebook rss twitter

Review: Nvidia GeForce RTX 3080 Founders Edition

by Tarinder Sandhu on 16 September 2020, 14:01

Tags: NVIDIA (NASDAQ:NVDA)

Quick Link: HEXUS.net/qaen56

Add to My Vault: x

GA102 as GeForce RTX 3080

Turing to Ampere

 
RTX 3090
RTX 3080
RTX 3070
RTX 2080 Ti
RTX 2080 Super
RTX 2080
RTX 2070 Super
Launch date
Sep 2020
Sep 2020
Oct 2020
Sep 2018
July 2019
Sep 2018
July 2019
Codename
GA102
GA102
GA104
TU102
TU104
TU104
TU104
Architecture
Ampere
Ampere
Ampere
Turing
Turing
Turing
Turing
Process (nm)
8
8
8
12
12
12
12
Transistors (bn)
28.3
28.3
17.4
18.6
13.6
13.6
13.6
Die Size (mm²)
628.4
628.4
392.5
754
545
545
545
PCIe
4.0
4.0
4.0
3.0
3.0
3.0
3.0
Base Clock (MHz)
1,400
1,440
1,500
1,350
1,650
1,515
1,605
Boost Clock (MHz)
1,695
1,710
1,725
1,545
1,815
1,710
1,770
Founders Edition Clock (MHz)
1,695
1,710
1,725
1,635
1,815
1,800
1,770
Shaders
10,496
8,704
5,888
4,352
3,072
2,944
2,560
GFLOPS
35,581
29,768
20,314
13,448
11,151
10,068
9,062
Founders Edition GFLOPS
35,581
29,768
20,314
14,231
11,151
10,598
9,062
Tensor Cores
328
272
184
544
384
368
320
RT Cores
82
68
46
68
48
46
40
Memory Size
24GB
10GB
8GB
11GB
8GB
8GB
8GB
Memory Bus
384-bit
320-bit
256-bit
352-bit
256-bit
256-bit
256-bit
Memory Type
GDDR6X
GDDR6X
GDDR6
GDDR6
GDDR6
GDDR6
GDDR6
Memory Clock
19.5Gbps
19Gbps
14Gbps
14Gbps
15.5Gbps
14Gbps
14Gbps
Memory Bandwidth
936
760
448
616
496
448
448
ROPs
112
96
64
88
64
64
64
Texture Units
328
272
184
272
192
184
160
L2 cache (KB)
5,120
5,120
4,096
5,632
4,096
4,096
4,096
SLI
Yes
No
No
Yes
Yes
Yes
Yes
Power Connector (FE)
12-pin
12-pin
8-pin
8-pin + 8-pin
8-pin + 6-pin
8-pin + 6-pin
8-pin + 6-pin
TDP (watts)
350
320
220
250
250
215
215
Founders Edition TDP (watts)
350
320
220
260
250
225
215
Suggested MSRP
$1,499
$699
$499
$999
$699
$699
$499
Founders Edition MSRP
$1,499
$699
$499
$1,199
$699
$799
$499

Here are the GeForce RTX 30-series GPUs laid bare next to their 20-series brethren. We're focussing on the RTX 3080 in particular. Like the RTX 3090, it uses the 28.3-billion transistor GA102 die that, if you recall, comprises 84 SMs across split across seven GPCs, and you see it in all its floorplan glory on the previous page.

The RTX 3080 implementation removes one GPC and further trims two others to hold 10 instead of the maximum 12 SMs. That means a total of 68 SMs (4x12 plus 2x10). The resultant core calculations are straightforward. RTX 3080 holds 8,704 cores that look so impressive compared to the paltry 3,072 on RTX 2080 Super, which is the GPU it effectively replaces at the $699 price point.

But do recall just how these cores are counted. Half of the them are straight-up FP32, while the other half can do FP32. Unlike Turing there is no further, dedicated INT pipe. Even so, Nvidia is correct in saying that at a peak boost speed of 1,710MHz RTX 3080 has nearly 30 TFLOPS of raw FP32 power, compared to 11.1 on RTX 2080 Super. This is fantastic if your GPGPU workloads smash out FPs, where RTX 3080 is a monster, but less relevant in gaming.

Even though RTX 3080 has more SMs than RTX 2080 Super - 68 vs. 48 - it has fewer Tensor cores due to there being half as many in every SM. There's no such reduction in RT cores, however, so they remain aligned to SMs on a 1:1 basis. ROPs, however, see a hearty increase due to them being decoupled from the backend and located in the GPC instead.

SLI goes by the wayside on this model, with Nvidia increasingly moving away from it on anything but its best-in-class GPU, and keeping up with interfaces, PCIe 4.0 is present and accounted for here

Nvidia increases the memory footprint to 10GB and, as discussed ramps up bandwidth by using GDDR6X operating at 19Gbps on, this time around, a 320-bit bus. Frequency and width combine to offer 760GB/s of bandwidth, or 53 percent more than its direct predecessor.

The visceral numbers suggest that not only is RTX 3080 far quicker than RTX 2080 Super - up to twice as fast, if you believe Nvidia's best-case projections - it ought to be substantially quicker than RTX 2080 Ti, which is a GPU that cost almost twice as much on its release two years ago.

Benchmark performance is likely not in doubt because it will top the charts compared to the existing competition. We're more concerned with the 320W TDP. Nvidia, too, realises that it needs innovation to cool such densely-packed heat. Enter the all-new Founders Edition card.