Today AMD launches its ‘first 7nm workstation graphics card’.
AMD Radeon Pro W5700
Radeon Pro W5700 in a nutshell:
- First Workstation card with PCIe 4.0
- First Workstation card manufactured in 7nm node
- First Radeon Pro with USB-C connector
- 2304 Stream Processors @ 8.9 TFLOPs
- 8GB GDDR6 256-bit
The Radeon Pro workstation series is receiving an update. The WX series will be succeeded by the W series. This is probably due to the fact that AMD is trying to unify naming schemas across the whole Radeon lineup. It is still not perfect, also constant changes are not helping, as customers are just required to constantly check what’s the latest.
The latest is Radeon Pro W5700, a Navi 10 based workstation graphics card. AMD calls it the first 7nm workstation card. It is not a 16GB model, but 8GB GDDR6.
This is not the full chip, Pro W5700 features 36 Compute Units (2304 Stream Processors), which aligns with WX 7100 and Quadro RTX 4000, to which AMD is comparing their new model.
In terms of actual performance, the maximum 32-bit floating-point throughput is 8.9 TFLOPs, this is more than RTX 4000 and much more than the Polaris-based WX series.
The W5700 features six display connectors: 5x mini-DP and USB-C (this is the first Radeon Pro card to have this connector).
A quick look at Radeon Pro Workstation series over the past 4 years:
| Radeon Pro Workstation Series | |||||
|---|---|---|---|---|---|
| VideoCardz.com | 2016 Radeon Pro WX x100 | 2017 Radeon Pro WX x100 | 2018 Radeon Pro WX x200 | 2019 Radeon Pro WX x200 | 2019 Radeon Pro Wx700 |
| Big GPU | WX 9100 | ||||
| WX 8200 | |||||
| Mid GPU | WX 7100 | ||||
| WX 5100 | W5700 | ||||
| Small GPU | WX 4100 | ||||
| WX 3100 | WX 3200 | ||||
| WX 2100 | |||||
The current mid-range workstation lineup is as follows:
| Professional Workstation Graphics Cards | ||||
|---|---|---|---|---|
| VideoCardz.com | Radeon Pro W5700 | Radeon Pro WX 7100 | Radeon Pro WX 5100 | Quadro RTX 4000 |
| Release Date | Nov 2019 | Jul 2016 | Jul 2016 | Nov 2018 |
| GPU | 7nm Navi 10 | 14nm Polaris 10 | 14nm Polaris 10 | 12nm Turing TU106 |
| Cores | ||||
| Max FP32 Compute | ||||
| GPU Clock | ||||
| Memory Clock | ||||
| Memory | ||||
| Memory Bus | ||||
| MSRP | ||||
