Sapphire NITRO+ Radeon RX Vega64 8GB HBM2 Limited Edition Review


Let’s check out the differences between the two, though it is a bit unfair since they are not the same category of cards.  The direct competition for the

is the GTX1080, not the GTX1080 TI, but still a nice comparison.  We will be comparing stock versus overclocked for both cards.

For reference, here are the overclock settings on the EVGA Geforce GTX1080TI FTW3 Gaming Edition.

Here are the original clocks, with the overclock settings as well for reference.

A pretty decent overclock on the 1080TI, and that is what we will be leading these results with.  Overclocks will be different on all cards, even the same GTX1080 TI FTW3 model, 2 will card will get different results, luck of the draw.

Let’s get started with the benchmarking.

Clearly, the GTX1080 TI outperformed the Vega64, it is not really a surprise being that we are running 2 separate tier cards, but the Vega64 did do relatively well.  We can see the Total Time Spy score for the overclocked GTX1080 TI was 29.56% higher than that of the overclocked Vega64.  That’s a pretty decent increase, with that increase, the Vega64 also consumed 19.05% more power and was 14.38% hotter, though still reasonable.

So let’s check it out.

Things are getting smaller as we are packing in the scores here, but clearly the GTX1080 TI is in the lead.  The GTX1080 TI’s overclocked scores pulls out ahead of the Vega64 overclocked scores by 25.42%.  With that, the AMD Overclock actually is pulling less power on average than the GTX1080 TI overclocked version by 8.53%.  The GTX1080 TI comes in though 40% cooler than the Vega64, that’s amazingly impressive.

Let’s see what Metro says.  Metro doesn’t care what you are, it will beat you down.

1920 x 1080

Eeek, that is impressive on NVIDIA’s part.  The EVGA Geforce GTX 1080TI FTW3 out performed the Sapphire Radeon NITRO+ RX Vega64 Limited Edition by 54.39%, but again, not only is the 1080TI on a tier above the Vega64

but the 1080TI FTW3 is one of EVGA’s highest clocked 1080TI’s.  The Vega64’s average wattage was 95.06% lower than that of the GTX1080 TI’s, which is a little refreshing, but of course that does not make up for performance.  On the flip side, the Vega64’s temperature was 6.99% above that of the GTX1080 TI’s.

2560 x 1440

We see almost the same thing we saw on the previous results, over double the performance in favor of the GTX1080 TI FTW3.  The GTX1080 TI FTW3 overclocked scored 69.64% above that of the Vega64 Overclocked, with that the temperature of the Vega64 Overclocked was only 4° higher than that of the GTX1080 TI FTW3, which is decent.  Here again the Vega64 consumed on average 85.60% less voltage than that of the GTX1080 TI.

3840 x 2160

This resolution was the one I was not counting on the Vega64 being above.  We can see here the overclocked GTX1080 TI performed on average 62.69% better than the Vega64 and with that, the card was 5.06% hotter than its green competition.  The saving grace here, the Sapphire Vega64 consumed 72.28% less power on average compared to that of the EVGA GTX1080 TI FTW3.

1920 x 1080

This starts to get a little interesting, while the GTX1080 TI is again out performing the Vega64, the gap has decreased a bit.  There was a performance increase of 27.96% in favor of the GTX1080 TI.  The average power consumption was 24.38% above that of the GTX1080 TI and the temperature was also 30.99% above that the GTX1080 TI.

2560 x 1440

Jumping into the results, we can see the overclocked EVGA GTX1080 TI FTW3 outperformed the Sapphire Radeon NITRO+ Vega64 Limited edition by 36.50% here.  The wattage was also lower by 25.84% and the thermals were also 22.86% improved as well.

3840 x 2160

Being that these numbers are much tinier when I saw then I thought, “Alright the card is gaining some traction” but alas, no such luck here.  The overclocked 1080 TI out performed the overclocked Vega64 by 39.12%, the wattage consumed by the Vega64 was also 4.77% higher than that of the GTX1080 TI.  Lastly, the cooling on the GTX1080 TI was 23.19% better than that of the Vega64.

EVGA did a great job on the cooling of this card, trying hard not to repeat their prior issue with the overheating VRM introduced their iCX cooling technology that contains 9 thermal sensors for more efficient cooling.  Sapphire and AMD introduced Radeon Chill, Vapor Chambers and Sapphires Turbine X.  I could have cooled the cards more but I ran the tests first on the Vega64 and when I thought all was good, I ran the same tests on the GTX1080 TI, pushing both for max overclocks on a short time crunch.  I focused more on speeds and stability than cooling and noise, only of course if cooling was an issue stopping me from obtaining a good overclock.

Anyway, enough rambling, lets check out Tom Clancy’s Ghost Recon Wildlands.

1920 x 1080

A surprising turn of events here, it looks like the Vega64 surpassed that of the GTX1080 TI’s scores.  The GTX1080 TI overclock obtained a score of 80.45, while the Vega64 scored 23.22% above that at 101.58.  With that though, the Vega64’s temperatures were 32.06% higher than that of the GTX1080 TI and consumed 14.30% more power.  All of the previous results have favored the GTX1080 TI but maybe the tables have turned here.

2560 x 1440

This is pretty amazing, the Vega64 and the GTX1080 TI both score the same, to the dot in frames per second.  I don’t know if this has ever happened or if it is just incredibly rare, but it is amazing.  Of course, there are differences, the temperature on the Vega64 was 17.13% above that of the GTX1080 TI and with that, the temperature for the Vega64 was 22.90% higher.  Both temperatures are well below the state at which the cards will throttle, so this is a big win in Sapphire court.

3840 x 2160

In another amazing bout, the Vega64 overclock has again taken the crown.  We find here that the Vega64 score 47.72 frames per second, 11.55% above the GTX1080 TI’s 42.51 frames per second.  The temperature on the Vega64 however was 14.08% higher than its rival, the GTX1080 TI and the wattage was also 18.58% higher on the Vega64.  Let’s see on which The Division will run better on.

1920 x 1080

Looks like the streak was broken.  We find on the AMD Overclock, Tom Clancy’s The Division scored a decent 121.8 frames per second but the NVIDIA Overclock came in with a 24.74% gain in performance.  With that win, the NVIDIA Overclock also 26.57% cooler and utilized 10.48% less power than its competition.

2560 x 1440

The GTX1080 TI blew up here, it scored 113.8 framer per second, that’s 27.83% faster than the Vega64’s 86.3.  Things heated up here, the Vega64 heated up to 81°C while the GTX1080 TI was 21.92% cooler at 65°C.  Lastly, the Vega64 consumed 20.47% more power than the GTX1080 TI.

3840 x 2160

At 4K, we can see the GTX1080 TI is still playable at 63.6 frames per second, that’s 24.54% above the Vega64.  The Vega also came in 10.13% hotter than the GTX1080 TI and with that consumes 21.14% more power than the GTX1080 TI.

This comparison by no means was fair, again these are different tiered cards and as I mentioned before, a much better comparison would have been made if I would have benchmarked it again a standard GTX1080.  I did however want to benchmark them to give you an example of the performance you can expect.  We will get more into this on the conclusion.

Let’s check out one more feature of the card using TRIXX Card Utility.

TRIXX Card Utility

TRIXX previously would overclock the video card, change fan speeds, change color of RGB LED’s and more, but it seems Sapphire has stripped it all of its abilities save the RGB feature.  Let’s check out TRIXX Nitro Glow.

This portion of the card, on the back lights up.

The white portions surrounding the outer 2 fans also lights.

The retention bracket itself does not light up but it is a silk screened logo, so it reflects the light.

The software is very basic and simple to use.  The first of the selection is “Pure blue color”.  With Pure blue selected and with all of the options, you can adjust the brightness from 100%, 75% and 50%.  Pure Blue color, Rainbow (random colors), Color change by PCB temperature and Color change by fan speed are basic options with no other options other than brightness, so there is nothing to show there.

Color change by PCB temperature will change the color of the fan outlines as well as the back of the card depending on the temperature of the PCB.  Color change by fan speed changes the color of the fan outlines as well as the back of the card depending on the fan speed, and remember, the fans turn off as well and go extremely fast depending the application, so there will be many different lights you see.

Custom Color, allows you to select predefined colors, select colors from the color prism or actual type in a number reflecting the color you prefer and specifically find the color you want.  You can of course play with these numbers under the R, G, B, # sections to find the color for you.

You can also turn off the lights if they bug you, but we all need a little light in our lives, I prefer blue or Rainbow, what color do you like?

Just in case you want to see what this looks like, I got something for you.

With all this done, its time for the final chapter, Final Thoughts and Conclusion.

Iggy Castillo
I have spent many years in the PC boutique name space as Product Development Engineer for Alienware and later Dell through Alienware's acquisition and finally Velocity Micro. During these years I spent my time developing new configurations, products and technologies with companies such as AMD, Asus, Intel, Microsoft, NVIDIA and more. The Arts, Gaming, New & Old technologies drive my interests and passion. Now as my day job, I am an IT Manager but doing reviews on my time and my dime.
Iggy Castillo
Share Feedback We Want to Hear From You