This year at CES, AMD had a few different big announcements, one of which was on the GPU side where they introduced their latest GPU the RX 5600 XT. The RX 5600 XT falls in between the original Navi RX 5700 and RX 5700 XT cards and above the RX 5500 XT that launched not too long ago. This is AMD filling out their Navi product stack. Well, today's the day, the RX 5600 XT is available and with that, we can also finally look at the cards and their performance. AMD sent over the ASRock RX 5600 XT Phantom Gaming OC for mt to take a look at and in addition to being excited to see how the new GPU performs I am also excited to get my first look at an ASRock video card. I’ve used many of their motherboards in the past, but it wasn’t until 2018 when they introduced their first cards. So I’m curious to see how they compare with the competition while I also see how the 5600 XT fits into the market.
Product Name: ASRock RX 5600 XT Phantom Gaming D2 6G OC
Review Sample Provided by: ASRock
Written by: Wes Compton
Amazon Affiliate Link: HERE
What is the RX 5600 XT
By now you have most likely heard about Navi, AMD's newest architecture. It is the bases of the RX 5700 cards and the RX 5500 XT as well. The RX 5600 XT is no different, this is also a Navi card which is based on AMD's new RDNA architecture. It is also built on the same 7nm process as well as the other cards. So you might be wondering what is different about the RX 5600 XT and then 5500 XT or the RX 5700. Well the RX 5500 XT is a cut down GPU which has 22 compute units which drops things down to 1408 stream processors. But when comparing the RX 5600 XT with the RX 5700 things start to get a lot more interesting. This is actually the same GPU.
You get the same 2304 stream processors. What is different is the memory which is running on a 192-bit memory interface compared to the 256-bit interface on the RX 5700. You also get less memory in general, with 6GB compared to the 8GB. Clock speeds on GPU are also lower. The boost clock on the RX 5700 on the reference card was 1725 MHz where the RX 5600 XT is up to 1560 MHz. The game clock on the 5600 XT is 1375 MHz vs 1625 MHz on the 5700 as well. You can see it all broken down below. But it is interesting that AMD didn’t cut the GPU down or “disable” some of the compute units like Nvidia will sometimes do to differentiate different models. AMD really pushed that on our call as well which is interesting because they have done it in the past. Then again in the past, those same cards did end up with community “hacks” where people reflashed them up to the higher model at a cheaper price, maybe they are looking to avoid that this time around with the RX 5600 XT.
As for where AMD pictures the RX 5600 XT in the market, in our call they did break that down as well. The 5500 XT is their middle of the park 1080p card and the RX 5700 series of cards are in their mind for 1440p. The new RX 5600 XT then lands at the high end of 1080p. Which does fit the choices they made with the memory controller on the RX 5600 XT being smaller and having 2GB of less memory than the RX 5700. 1080p can still use a lot of memory, but going higher in resolution really ramps things up even more.
So AMD is saying that the RX 5600 XT should be seeing up to 80 FPS in triple-A titles and up to 120 FPS with the traditional esports titles that are a little less demanding. This puts the card in a great position where you can play anything and hardcore competitive games like the esports titles can take advantage of high refresh monitors. We will have to put that to the test later though. As for pricing, they have a launch MSRP of $279. They do also have a game bundle as well, but not as big of one as the RX 5700 series hards had. At least right now they are offering the RX 5600 XT with a 3 month Xbox Game Pass which does get you access to a long list of games. I personally would prefer to get one game you can keep long term, but someone who doesn’t have the game library that I have might prefer the Xbox Game Pass offer more.
Hey, this is future Wes here, when we launched this review right around the same time ASRock released a new BIOS for the Phantom Gaming. That BIOS overclocked the card from the original clock speed of 1620 MHz up to 1750 MHz for the boost clock. I am redoing all of our testing with this new driver but because of the timing if you buy a Phantom Gaming you should use GPUz to check your clock speeds as well to make sure they match what we have below. Beyond that, you can see the new BIOS version in GPUz and see that I did the testing using the pre-release Jan 13th driver.
Being the first ASRock video card to come into the office also means I don’t have any past experience with the packaging. Our card is their Phantom Gaming model which is ASRock’s gaming brand and it has that big OC logo to show that it is overclocked as I already have shown in the previous section. It has the normal red wrap around which has the model name in it and AMDs branding. ASRock kept their branding small as well in the top right corner and then the Phantom Gaming branding covers almost the entire front with a black and grey background and a touch of red. I would really prefer a picture of the card here, I say this a lot but I feel like if you are shopping in retail at all it is important to be able to see what you are getting with a picture or a window. There is a small photo of the front and the back on the back of the box where they touch on a few features like the dual fan design, metal backplate, fan that stops during light workloads, and what they call Polychrome Sync which is their lighting control software. That one also notes that the lighting uses addressable LEDs. There is a specification listing and shockingly it actually has clock speeds listed which no other company does. They also tell you which display connections there are but they are missing the card dimensions which is the other important thing to be able to check when picking a card.
Inside the box, the card comes wrapped up in the normal static protective bag. It then sits inside a thick foam tray. Then on top, there is another foam panel that helps cover every angle, this card isn’t going to get banged up in shipping unless they bend the box in two or run it over. There aren’t any accessories included or swag. You do get a small and basic installation guide which looks like it comes with all of their cards.
Card Layout and Photos
I have to admit, ASRock’s RX 5600 XT Phantom Gaming doesn’t have the exact same look that everyone else has. Most companies have gone with a color-neutral angular look in the past and recently all of the cards are going with boxy and simple designs. The Phantom Gaming mixes things up with a black, silver, and red look for the fan shroud. I would really prefer to not have the red, but I do like that this at least looks different. It has a metal looking finish, but it is made of all plastic, just like everyone else. The silver sections all have a machined look that gives that brushed aluminum look. The red and black section also uses a little texture as well but it isn’t as defined as the photos below might indicate. As for size, the height of the card is spot on, not sticking up much past the top of the PCI bracket. The card is then 9 inches long which is a little longer than the PCB which is around 7 ¼ inches long. That is still shorter than all of the three fan designs. The card, however, is thick with what is closer to a 2 ¾ wide design.
For the cooling, ASRock went with dual axial fans to push air down into the heatsink and on to the PCB as well. Then the heatsink is made out of copper heatpipes and aluminum sheet metal with the fin layout in a horizontal layout. I’ve had bad luck with axial coolers with a horizontal layout recently, this layout is normally better with the blower cards to blow air out of the end so I’m curious to see how it performs later. I will say that even just looking through the two fans you can see that the heatsink doesn’t go all the way up to the top or bottom of the fan blades leaving some cooling on the table there. You can, however, see that down on the right side where the cooler extends past the PCB they left it open to blow right through there.
Now you can see how thick this card is, XFX is calling their cards Thicc but this one is just fat. The top view lets us see that there are a total of three heatpipes split out across the card. You can also see how the cooler is shaped very tightly around each cap and the power connection, utilizing every mm of space. It's actually very interesting that they take every bit of space they can there but leave those big gaps at the top and bottom of the fan. The end of the card shows how in a lot of areas there is over an inch space between the edge of the shroud and the bottom of the heatsink. This doesn’t help with the fat look, but it does help with airflow! Then at the bottom edge, we can see those three heatpipes running from over the GPU to pull the heat out across the heatsink. There is also a bit of a gap down at the PCI bracket but with how tight the heatsink fits the PCB and components it looks like the airflow doesn’t have much of an option other than to head towards the end of the card, even without a blower fan.
So up on the top edge of the card, there are two notable things to point out. For starter, this card has an 8-pin power connection down at the end of the PCB. It is flipped around with the clip side facing the PCB and the PCB and backplate are both notched to fit the clip. This allowed ASRock to keep the heatsink extremely tight around the plug for more cooling capacity and if you look close you can see how the heatsink wraps around it. Also up on top is the backlit branding which is a 5 sided weird shape with the Phantom Gaming logo and name. This has addressable LEDs behind it that are controlled by ASRocks software.
Around on the back of the RX 5600 XT Phantom Gaming, the card does have a full backplate. Well almost full, it does cover all of the PCB at least, but you can see that the heatsink sticks out on the end. Anyhow the backplate is a nice all-metal design and the same color theme front the front does carry over on here. They have screen printed white and red strips along with the ASRock logo and the Phantom Gaming logo also in white. I was surprised to not see a serial number sticker on the back here, that could be because this is a pre-launch sample. Beyond that, you can still access the mounting screws for the heatsink and there isn’t any other ventilation in the backplate
For connection options, ASRock stuck with the standard layout which gets you three full-sized DisplayPort’s and one HDMI. Like most cards they did the split layout which puts the HDMI in the middle of the DisplayPort’s which I’m not a big fan of, I think it can be confusing. But you do get a mix of connection options, as long as you don’t want or need a DVI which given this is still a 1080p card could be an issue for some. Beyond that, the PCI bracket does have some ventilation which given the horizontal fin layout could get some air blown its way. The bracket itself is your standard chromed look, I still keep hoping for more companies to move to a blacked-out PCI bracket. Maybe ASRock will be the one who does that eventually!
So here is a look at the RGB lighting for the 5600 XT Phantom Gaming. I’ve been preaching this for a while now, but I’m not a big fan of backlit branding. It feels like almost everything anymore just lights up branding and when your PC is all together it looks like times square. So I would really prefer the lighting here light up a few accents, like the red areas on the front or a few lines on the top like the backplate has. But I do like the use of addressable LEDs which you can see even in the default mode light up different sections different colors, not a dull old solid color that rotates through the rainbow. So you can at least get a little more creative when setting the lighting to match your own build.
Test Rig and Procedures
Motherboard: Asus Crosshair VIII HERO WiFi
Storage: Corsair MP600 2TB
Cooling - Noctua NH-U12S
Power Supply - Corsair AX1200w
Case - Primochill Wetbench
As always I like to start off my testing with a few synthetic benchmarks. 3DMark especially is one of my favorites because it is very optimized in both Nvidia and AMD drivers. It's nice to not have to worry about it being favored too much either way and the repeatability of the results makes it a nice chance to compare from card to card. I started off with the three Fire Strike benchmarks which are a little older and focused on DX11. The range in tests allows a look at different levels of detail. Going in I was curious how the 5600 XT compares to the 5500 XT and the RX 5700, then on the Nvidia side the RTX 2060 and the GTX 1660 Ti. In all three of these tests, the RX 5600 XT Phantom Gaming came in ahead of the stock clocked RTX 2060 Founders Edition with the 2060 SUPER then being ahead of it. The gap between the 5600 XT and the 5500 XT is surprising, and the RX 5700 is out ahead but in this situation, it is running with the 2060 SUPER.
Sticking with 3DMark, for the second round of tests, I went with the newer Time Spy benchmarks. There are two different details and these are DX12 based. All of the Turing based Nvidia cards have done really well in Time Spy and you can see the significant difference between the results here and in Fire Strike. The 5600 XT, even with its overclock comes in with the RTX 2060 ahead of it and the GTX 1660 Ti behind it with nothing really in the same range. Like before, the RX 5600 XT is a lot closer to the RX 5700 which here was running with the RTX 2060. Then the RX 5500 XT is a lot farther away. In the Extreme test, the 5600 XT drops slightly when compared to the RTX 2060 but still sticks between the 1660 Ti and the 2060 overall.
The last synthetic test was using Superposition which is based on the Unigine engine. For this test, I ran two different 1080p tests, one at medium detail and another at the extreme detail setting. Once again the RX 5600 XT fell in between the RTX 2060 and the 1660 Ti but it is shockingly close to the RTX 260, especially in the medium detail test.
As for Virtual Reality, I love it but it is more demanding than traditional gaming. This is partially because of the resolutions needed to render for two eyes and because they render more than what is immediately visible. But also because of post effects to get the proper “fisheye” effect for it to look proper in your eyes with the HMD. You also have to have much higher expectations for frame rates in VR, skipping frames or lower FPS can actually cause motion sickness in VR. Because of that, I ran a few tests.
My first test was again in Superposition. This time I tested the VR Maximum and VR Future tests using the Vive resolution. Like in the other Superposition test I tested at an extreme detail which this time is the VR Future test and a more reasonable detail which is the VR Maximum test. The RX 5600 XT Phantom Gaming came in with a big lead ahead of the overclocked GTX 1660 Ti but still behind the RTX 2060 by 4 FPS. In the VR Future test, however, the 5600 XT doesn’t surpass the 1660 Ti and the RTX 2060 is running up above the RX 5700.
Then for the second round of VR tests, I used VRMark where I ran the Cyan and Blue room tests. In the blue room test as you can see the 5600 XT fell behind even the 1660 Ti. But in the more realistic cyan room test, it pulled back ahead of those cards and edged up 13 FPS from the RX 5700. The RTX 2060 on the other hand still has almost 7 FPS on the 5600 XT in the more realistic cyan room test. That said the 5600 XT still showed that it is a great VR option.
Now we finally get into the in game performance and that is the main reason people pick up a new video card. To test things out I ran through our new benchmark suite that tests 9 games at three different resolutions (1080p, 1440p, and 4k). Most of the games tested have been run at the highest detail setting and a mid-range detail setting to get a look at how turning things up hurts performance and to give an idea of if turning detail down from max will be beneficial for frame rates. In total, each video card is tested 48 times and that makes for a huge mess of results when you put them all together. To help with that I like to start off with these overall playability graphs that take all of the results and give an easier to read the result. I have one for each of the three resolutions and each is broken up into four FPS ranges. Under 30 FPS is considered unplayable, over 30 is playable but not ideal, over 60 is the sweet spot, and then over 120 FPS is for high refresh rate monitors.
So how did the ASRock RX 5600 XT Phantom Gaming perform? Well AMD promoted the RX 5600 XT as their ultimate 1080p card where the RX 5500 XT is a standard 1080p and the 5700 series of cards are for 1440p. So my main focus is on 1080p where the card did have two of our tests come in over 120 FPS, which given how none of the games tested are less demanding esports titles. 11 are in the 60-119 FPS range which is great. But the three that still fall into the over 30 FPS but under 60 FPS range are still a little concerning. When I dive into the detailed results I’m going to look for those three specifically to see what is going on there. Even with this not being a 1440p card, every game tested would be considered playable with the 5600 XT and with a majority of those even being over 60 FPS. Then at 4k things really take a dive. 10 tests were playable but none of them would be called smooth and nearly half were unplayable at all in the sub 30 FPS range. No surprises there!
Next, I want to take a look at the actual results. Specifically, I wanted to get an idea of how the RX 5600 XT compares with the RTX 2060 and the GTX 1660 Ti. I also want to look out for any weird results and look for those three 1080p results that fell under 60 FPS and see what games/detail they were and how close they were to making the 60 FPS cut. So start with those three sub 60 FPS results. The first one was in Metro Exodus with the ultra detail setting where it came in at 54.75 FPS. This was right above the RTX 2060 and in fact, even the 2060 SUPER didn’t break 60 FPS on most of the cards tested as well. The second one was Total War: THREE KINGDOMS at the ultra detail with an FPS of 57.7. Here the RTX 2060 did make the cut, coming in with almost 6 FPS more on this test. The last one was Ghost Recon: Wildlands at the ultra detail setting with an FPS of 57.49. The RTX 2060 was right there as well at 54.38 FPS but the 5600 XT did come out ahead on that one. SO overall all three of the sub 60 FPS results were VERY close. I think we can still consider this a premium 1080p card given that.
Above I also have the breakdown of the average FPS of all three cards for our 1080p results. I stick with just 1080p here because that is where this card is focused. But the breakdown shows exactly what I was seeing in all of the graphs below and previously in the synthetic results as well. The RX 5600 XT dominates the GTX 1660 Ti by a CRAZY amount with it being ahead over 12 FPS on average. But at the same time, it does still come in behind the RTX 2060 which is to be expected given pricing even after Nvidia’s pricing changes. Now keep in mind that is with the 2060 being a non-overclocked model where the other two are overclocked. This is especially true now that I am retesting with the new BIOS which added an even bigger overclock to the Phantom Gaming. The 1 FPS difference between the RTX 2060 and the RX 5600 XT doesn’t actually show that the 5600 XT came out ahead in our tests more often. But the 2060 did make up for it with a bigger lead in the games it did excel at.
Now some people don’t need a video card for gaming, they need the processing power for rendering or 2D/3D production or in some cases people who game also do work on the side. So it is also important to check out the compute performance on all of the video cards that come in. That includes doing a few different tests. My first test was a simple GPU Compute benchmark using Passmark’s Performance Test 9 and the RX 5600 XT Phantom Gaming came in right where you would expect, 332 points below the RTX 2060 and 600 points above the overclocked GTX 1660 Ti.
The blender benchmark is where I was most curious though, this is the best real world example of GPU compute because Blender is open source and used often. My benchmark uses the real Blender Benchmark program and it runs two renders and times them so a lower score is better. The RX 5600 XT came in at 402 seconds which was basically in line with the 1660 TI and 1660 SUPER with the original clock speeds but dropped down to 360 with the updated BIOS. This helped it close to the RTX 2060 but wasn’t enough to even things out.
In Basemark I ran both the DirectX12 and OpenGL test. All of the AMD cards have fallen far behind on the OpenGL performance and the RX 5700 XT wasn’t really any different. The DX12 results were better, with the 5600 XT coming in right with the RX 5700’s. This put it out ahead of the 1660 Ti but still behind the RTX 2060 however.
The last compute test is with the new Geekbench 5 where I run both the Vulkan and OpenCL tests. You would think that AMD would excel with the Vulkan test but they haven’t been. The RX 5600 XT did do better in the OpenCL test, but not enough to catch up to the GTX 1660 Ti, it was just a touch below the two RX 5700’s however. The RTX 2060 on the other hand though was still way out ahead, running with the RX 5700 XT.
Cooling, Noise, and Power
My last round of tests are all focused on areas that I find important, but don’t always have anything to do with gaming performance like most people are worried the most about. Here I test power usage, noise from the fans, and cooling performance. Once you have decided which GPU model you want, beyond clock speeds, these are the areas where different models from the same company and cards from multiple companies can stand out against each other.
To start things off, I’m going to look at power usage. Stock for stock it should normally be about the same. But when it comes to overclocks, each company has their own idea of what amount of voltage is needed for a safe overclock. More means it can be more stable and with a lot of the auto overclocks these days more voltage can also mean some GPUs will run at a higher clock speed. But it also means more power usage and often more heat as well. To test this I use a kill-a-watt on our test system to measure total power draw in wattage and run two different tests. 3DMark Fire Strike during the combined test that loads the video card and CPU similar to a demanding game. Then I do a GPU only test using AIDA64’s stress test only on the GPU which puts even more load on the video card but doesn’t have any extra CPU power being pulled. In the 3Dmark test, the RX 5600 XT Phantom Gaming pulled 314 watts which right with the RX 5700 and just 10 watts below the RTX 2060. In the AIDA64 test, the wattage dropped down to 270 which put the 5600 XT up with the overclocked RTX 2060 SUPERs. I should point out that with the original BIOS before the “2060 price drop 5600 XT overclock” the card pulled 218 which is a huge difference.
For the next test, I look at the noise output of the cooler. For this, I set up our decibel meter 18 inches away from the test bench and I run three tests. I do a manual speed test with the card set to 50% and 100% fan speeds, which by the way the AMD driver still freaks out when doing this required a reboot to get the fan speeds to drop back down. Then I do a third with the card under an AIDA64 load for over a half-hour where I test the noise level of the card in a worst-case scenario with the stock fan profile. ASRock did have the Phantom Gaming up on the high end of the chart at 100% fan speed, the 50% result was a little better with a few of the cards below it higher at 50%. This was still way off from the noise that both of the reference RX 5700 series cards put out, they were horrible. The RPM chart explains why the Phantom Gaming was so high up the charts as well with the 100% fan speed RPM being right up in the same area of the chart with its 4000 RPM fans. Then the last and most important test is the under load test. The Phantom Gaming was in the bottom half this time around showing that even though the fans can make all of that noise, they don’t need too with the stock fan profile.
My last set of tests are the most important, thermal testing. For this, I use the AIDA64 Stress Test again. I run the test for a half hour or longer, waiting on the temperatures to level off. I do the test twice, once with the stock fan profile and then again with the fans running at 100% fan speed. This gives us a look at what to expect for temps in a worst-case when gaming and then 100% lets us see how much room is left. So with the stock fan profile, the Phantom Gaming reached 69 degrees which is a lot more like a reference card than an aftermarket. The 61 degree’s that this card did with the original BIOS was great but the update clearly changed the dynamics of things. At 100% fan speed temperatures dropped down to 46 degrees which kept it around the same place in the charts. But the 23-degree delta between the two isn’t bad at all but it also tells us that they may have benefited by adjusting the fan profile a little with the new overclock.
When doing the stock fan profile test, I do also get a few thermal images of the RX 5600 XT Phantom Gaming to see how things are looking. The fan side of the card didn’t show any significant hotspots, including down along the bottom edge where a lot of cards like to vent. ASRock did a great job getting good cooling performance with a dual axial fan design and those horizontal fins that often don’t cool as good as they should. You can see that the left fan does have a slightly warmer area on the bottom than the right fan has, but it wasn’t warm enough to be concerned at all. Up on top, you can see that the PCB is the hottest part of the card, especially down on the VRM side of the card which is a little bit of a concern. The backplate image shows similar with it being cooler on end near the PCI bracket. You can also see how the card is venting out at the end with that haziness, it's good to see the air being pushed that far.
Overall and Final Verdict
New GPU launches are always interesting and exciting and the RX 5600 XT is no different. AMD went with what is basically the RX 5700 with a smaller memory controller and less memory while also running the GPU clock speed much lower. The tradeoff with that is you do still get the higher number of processing cores which I thought might translate to a little bit of a hidden gem when it came to compute testing, but that wasn’t the case. What ended up happening is the lower clock speed kept the RX 5700 out in front in all of the testing, but the RX 5600 XT ended up blowing away the GTX 1660 Ti and hanging with the RTX 2060 FE. In fact, I could argue that even though he 2060 FE had a 1 FPS lead on the average the 5600 XT was a faster card with it being faster in more tests after the updated BIOS that was pushed. This translated to great 1080p performance which AMD hinted at and frankly, it is at least 1440p capable as well. You can plan on taking advantage of high refresh monitors if you are playing less demanding esport focused titles as well.
I was also excited about this one because this is the first ASRock card I’ve had the chance to check out. I ended up being surprised by the cooling performance of the card. This is partially because of just how thick the cooler is, as I mentioned before if XFX card is THICC, this card is just plain fat. So keep that in mind, this isn’t going to work in 2 slot cases. I was concerned that noise was going to be an issue when I did my 100% fan speeds tests and the dual fans ramped up to 4k RPM and were loud. But the load testing which shows actual fan noise levels when under load for extended periods were actually on the other end of the chart. The cooler design isn’t going to be for everyone, but it is at least unique and unlike what anyone else is doing. I could go without the red accents though and I would prefer to see the RGB lighting being used with accents not just to light up the branding. Having addressable LEDs there was a nice touch. When looking at the cooler design I was unhappy with big gaps left at the top and bottom of the card. With the original BIOS these issues didn’t end up being an issue but with the updated BIOS the stock fan profile and cooler did run warmer than I would have liked. There was still headway with a fan profile adjustment being all that is needed. But that will most likely mean a lot more noise.
With me updating this review now a few days part the launch I can actually get a much better look at how the RX 5600 XT and specifically the ASRock RX 5600 XT Phantom Gaming fit into the market. When I wrote the original review I was never given pricing for the ASRock card and just AMDs MSRP. Along with that we also have the new BIOS update to take into account as well. I will say that I do also have the EVGA 2060 KO in the office and in fact already tested but waiting on me to get the 5600 XT testing fixed so you can expect a second look at this situation soon taking into that card as well. But for now, let's look at things in comparison to the regular RTX 2060. We can toss the GTX 1660 Ti out the window, Nvidia’s price change and AMDs last-minute overclock made that card completely irrelevant currently until it sees its own pricing adjustments.
Anyhow, so the MSRP of the RX 5600 XT is $279 but I can confirm that the two fan ASRock Phantom Gaming that I tested today is selling for $299.99. In fact, it is listed with an MSRP of $319.99 and on sale so hopefully, that doesn’t mean it may go back up in price. Like before the 5600 XT launch, the RTX 2060 doesn’t seem to have any availability at $299 but you can get cards in the $320 and $330 range. Given that issue and with the RX 5600 XT actually being faster in a lot of the games I tested the RX 5600 XT is looking like a good deal right now. This could change later if Nvidia’s cards do start to reflect the lower price more, especially with EVGA saying the $299 2060 KO will be back in stock soon. One thing is for sure, this is the new sweet spot. Either direction you go, you get top-notch 1080p performance and even some 1440p capabilities at a price that actually makes sense.
Live Pricing: HERE