AMD Continues to Pleasure Apple With the Radeon PRO W6800 & W6600/M Workstation GPUs
9 June 2021
AMD navi Navi 21 Navi 22 NAVI 23 Nvidia quadro Radeon Radeon Pro RADEON PRO W6600 RADEON PRO W6800 RDNA RDNA 2 RDNA2 RTX 5000 W6600 W6800 Workstation WX WX 7100 WX 9100 WX7100 WX9100

Finally a couple of Radeon graphics cards gamers around the world might actually have the chance to find in stock.

Though for all the wrong reasons, yes folks it’s that time again for AMD to introduce its next generation workstation hardware based on the RDNA 2 architecture, featuring up to 79% more performance than previous generation hardware.

AMD have announced the Radeon PRO W6800 and the W6600 workstation cards which will all but certainly find nice little homes inside the cancerous products of the Apple corporation, because that’s sort of the kinda relationship that Apple has with AMD.

With Apple basically taking up priority when it comes to the binning and allocation of these workstation oriented products for the MacBook and the Mac Pro which have been offered with various Radeon PRO solutions over the years of which Apple probably got the bargain of the century between its partnership with AMD of which they’ll turn around and charge you thousands of dollars for pretty laughable specifications. MacFags truly are a unique species.

Much like any other RDNA 2 based graphics card, it provides exceptional performance benefits over previous generation solutions and the same certainly goes for its performance per watt characteristics.

Starting things off with the Radeon PRO W6800, it comes with the pretty bland vapor chamber blower style cooler so unanimously associated with these cards given the sorts of environments they will be placed in. However with a rather colorful sexy sleek exterior of brushed aluminum and YInMn Blue which has been a staple of Radeon Pro graphics cards since the introduction of the WX 7100 back in 2016.

The Radeon PRO W6800 comes with a rather surprising 32 GB of ECC GDDR6 memory, doubled that of the sorts of consumerist graphics cards of which its based off. In the case of the Radeon PRO W6800 as the name might suggest it’s based off the same core as the standard Radeon RX 6800 featuring 60 CU’s or rather 3840 stream processors.

And much like RDNA 2 consumer parts, compared to the competition of NVIDIA’s Quadro offerings at just $2250 USD for the Radeon PRO W6800.

Aside from that the Radeon PRO W6800 provides a maximum FP32 compute figure of 17.83 TFLOPs, doubled that for FP16 and just 1.11 TFLOPs of FP64 throughput, the Radeon PRO W6800 has a stellar I/O with a total of six mini-DisplayPort 1.4 outputs along with the standard 8+6-pin power connectors with a TBP value of 250W.

Which certainly is enough to trounce the previous generation Radeon PRO VII of which the W6800 provides 36% more FP32 computer horsepower alongside the doubled effective memory buffer for just a small 18% higher price.

While it may have the same board power value as the standard RX 6800 of which its based you can clearly see to what extent the binning process is when it comes to stability, efficiency and performance, whereas the RX 6800 has a TBP of 250W it only offers a maximum boost frequency of 2105MHz, the boost frequency of the Radeon PRO W6800 is actually higher at 2300MHz.

Continuing onwards to the Radeon PRO W6600 which utilizes the Navi 23 core which has yet to be released on a consumer based product.

previous arrow
next arrow

The Radeon PRO W6600 offers many of the same thrills in a much smaller and efficient package, being only a single slot card which is quite indicative of the sort of power it would consume of which it has a TBP value of just 100W, so it shouldn’t really come as a surprise that alongside the Radeon PRO W6600 comes a mobility variant dubbed the W6600M for mobile workstations.

Delivering 10.4 TFLOPs of FP32 compute performance, it comes with just 1792 stream processors from 28 CU’s paired with 8GB of non-ECC GDDR6 memory across a 256-bit memory interface though only providing a peak bandwidth of just 224 GB/s, it also comes with just 32 MB of Infinity Cache.

As for outputs from its single slot design, it comes with a total of four full DisplayPort 1.4 connectors and only requires a single 6-pin external power connector, more importantly however is the staggering boost frequency of the W6600 which comes in at 2900MHz.

Pricing for the Radeon PRO W6600 goes for just $649 USD.

Much like all RDNA 2 hardware these Radeon PRO workstation cards feature the wonders of the RDNA 2 architecture such as the Ray Accelerators allowing for ray tracing, and of course PCIe 4.0 x16 support which has been a staple for Radeon since the last generation along with AMD’s dedicated hardware security embedded inside which is most certainly an Israeli backdoor.

However one very important aspect noted from the Radeon PRO W6000 slide deck featuring a closer look at the Navi 23 core at the helm of the W6600.

The littlest Navi aims to punch big coming in with a die size of just 237mm2 which is considerably smaller than that of the GA106 core at the helm of the GeForce RTX 3060 of which will most likely be the main competition for a commercialized Navi 23 graphics card to humiliate next.

The following selection of slides are basically various performance scenarios comparing these new Radeon PRO W6000 workstation cards compared to previous generation solutions and the more expensive Turing based NVIDIA Quadro RTX 5000 of which these RDNA 2 based cards showcase further prowess.

previous arrow
next arrow

However I think you’d agree with me that the biggest change that comes with these Radeon PRO W6000 workstation cards are the packaging that they come in.

AMD has gone green, or more rather they’ve gone cheap with the packaging of these new Radeon PRO graphics cards being just 59% of the size in comparison to the boxes of previous generation hardware. Because the size of your box certainly matters to some, I mean how else are AMD offering the Radeon PRO W6800 and W6600 for so damn cheap if not for the “premium experience” of less cardboard?


blog comments powered by Disqus