Table of Contents
A couple of weeks back NVIDIA released their latest Turing video card architecture. The GPUs brought new features including Ray Tracing cores, Tensor cores, and our first glimpse of GDDR6 memory. Our review of the Founders Edition cards showed significant performance gains over last year’s models, but with it came even more significant price increases putting a blemish on its release.
Now that time has passed, board partners are getting cards out to reviewers to take a look and see what they could do with the Turing GPU architecture as a base. NVIDIA has already stepped up its game with better power delivery as well as a cooler so the board partners have to put in some effort to trump NVIDIA this time around.
Today, we are looking at our first partner card in the MSI RTX 2080 Gaming X Trio. The Gaming X Trio hails from MSI’s premium line of cards (Gaming) and with it includes a slight overclock out of the box. The main difference between the FE cards and this offering is the Tri-Frozr Thermal design/large heatsink that MSI put on this card. Between its raised clocks and monster cooler, we are hoping to see better overclocking and performance from this card. Read on to see the results of our testing!
Below is a specifications table with the MSI RTX 2080 Gaming X Trio along with the 2080 Ti, 2080, and 2070 specifications. The biggest takeaway here for the MSI RTX 2080 Gaming Trio X is the boost clock being raised 60 MHz over the FE RTX 2080. Outside of that, the rest of the specifications remain the same as far as memory speed (and amount) as well as the base clock.
|MSI RTX 2080 Gaming X Trio Specifications
|MSI RTX 2080 Gaming X Trio
|RTX 2080 Ti
|GPU Base Clock (MHz)
Reference / Founders Ed.
|1350 / 1350
|GPU Boost Clock (MHz)
Reference / Founders Ed.
|1860 (1950 Actual)
|1545 / 1635
|1620 / 1710
|Frame Buffer Memory Size and Type
|8 GB GDDR6
|11 GB GDDR6
|8 GB GDDR6
|Memory Clock (Data Rate)
|Memory Bandwidth (GB/sec)
|Texture Fill Rate (Gigatexels /sec)
|420.2 / 444.7
|314.6 / 331.2
|233.3 / 246.2
|L2 Cache Size
Reference / Founders Ed.
|250 / 260W
|215 / 225W
|175 / 185W
|Transistor Count (Billions)
|Die Size (mm²)
|12 nm FinFet
|Price – MSRP (Reference / Founders)
|$999 / $1199
|$699 / $799
Our gratuitous shot of GPUz confirms what we are seeing from the specifications table. Though, the texture Fillrate field appears to be off compared to what the NVIDIA whitepaper says it should be as well as the memory bandwidth being woefully off. We have pinged the GPUz creator for comment and correction if needed.
EDITOR’S NOTE: GPUz 2.12.0 fixes the memory bandwidth issue. However, the Texture Fillrate value in GPUz is based off the base clock as opposed to how NVIDIA lists it based of the boost clock.
Below is a list of features from the MSI website. Highlights include the TRI-FROZR cooling using the third generation TORX Fan 3.0 using a combination of dispersion fan blades using a special curved blade said to accelerate the airflow for increased effectiveness and a traditional fan blade used to push airflow down on the heatsink underneath. The TORX 3.0 fan utilizes a double ball bearing setup to help ensure quiet operation and long lifespan. The TRI-FROZR cooler also uses its Zero Frozr technology on the fans where they are completely stopped when the temperatures are below 56 °C and 60 °C (large fans and small fan respectively). This allows for a completely silent graphics solution in low power situations and when needed, will spin up automatically.
In addition to the fan, the heatsink also has a few tweaks that MSI says improves its overall performance. The heatsink uses “wave-curved” fins to allow the air to pass more smoothly through the heatsink resulting in less noise. The shape of the fin array guides the airflow directly on the concentrated heatpipes while creating more surface area for the air to absorb more heat. The copper heatpipes run through the aluminum fins. At the area where they touch the nickel-plated copper base plate base and collect the heat, the pipes are as close together as possible which should spread the load and remove the heat quicker.
Last, is the Mystic Light Software. MSI has updated this RGB controlling software and will detect compatible products from both MSI and others brands in a list for users to control by themselves, or linked together. Mystic Light Sync allows users to control RGB LED implementations from other brands, its like CPU coolers, keyboards and mice, or RGB system fans, for example, can be controlled by this software.
Retail Packaging and Accessories
MSI’s retail packaging is fairly nondescript with a black and NVIDIA green theme prevailing along with a picture of the card inside. The front of the box contains identifying information such as the model itself and its own internal branding, in this case, one can see it is part of their Gaming X Trio line of cards. The back of the packaging goes into more details on the specifications and features.
Inside the box, the card itself sits securely in form fitting foam and an anti-static bag to prevent any issues during shipping. The card comes with a couple of accessories including a large strut to support the monster GPU (we suggest using this if the card isn’t vertical in the slot). It also contains a 6-pin to 8-pin adapter along with the usual manuals as well as some MSI gaming branded coasters to put drinks on.
Meet the MSI RTX 2080 Gaming X Trio
The MSI RTX 2080 Gaming X Trio marks the return of MSI’s three-fan graphics cards which we have seen for a couple of generations now. What is different about this implementation is the fact that the smaller fan of the three is now on the left side of the card (close to the outputs) with the two larger fans in the middle and right side of the card. The Trio series of cards features a blend of gunmetal grey and black with a brushed metal backplate intended to match the front of the card. It also has five locations on the card for RGB lighting controlled by their Mystic Light software.
The card itself is huge. It measures in at 12.8″ x 5.5″ x 2.2″ and weighs a hefty 3.4 lbs. The era of double+ slot cooling has seemingly returned with a lot of partner cards strapping on some pretty beefy coolers on the Turing GPUs. Thankfully, MSI includes a strut which screws into the case/frame which will support the card for any non-vertically oriented systems.
The TRI-FROZR cooler includes the wave-curved heatsink mentioned previously, which is designed to optimize heat dissipation and to reduce noise. The heatsink itself includes multiple copper heat pipes measuring up to 8 mm, as well as the orientation of the heat pipes close together to transfer more heat from the copper base through the heat pipes. The TRI-FROZR heatsink also directly cools the MOSFETs through a connected cooling plate. Additionally, MSI uses a unique thermal compound to provide maximum conductivity.
Overall, the cooler is pretty large taking up 2.5 slots, but looks to do a good job cooling the TU104 GPU hidden underneath with all the features it packs in. We saw some good results in our thermal testing at default.
A Closer Look
In this section, we take a closer look at the cards including outputs, power lead requirements, a look at the PCB and heatsink base, and power delivery section as well as some of the IC’s used to drive the card.
The outputs here are the same as on the FE cards with a total of five outputs. There are three DisplayPorts (v1.4), one HDMI (2.0b), and one USB Type-C port. The card supports HDCP 2.2. As far as power requirements, the GPU is said to use 260 W and with the Gaming X Trio requiring two 8-pin PCIe power connectors. This gives the card 375 W of “in specification” power which should be plenty with overclocking.
After taking off the heatsink, we expose the PCB to see a different power design than the FE cards. In this case, the MSI RTX 2080 Gaming X Trio uses a 10+2 phase setup as compared to the 8+2 setup on the FE cards we reviewed earlier. This should allow the GPU and memory to receive more power and cleaner power through the setup.
The MOSFETS are FDMF3170 models from OnSemiconductor rated for up to 70 A. Driving the power bits are two controllers. The memory uses a uP9512Q 4-phase (up to) buck controller while the GPU uses a Monolithic Power System (MPS) MP288Q multi-phase PWM controller. The memory chips used on this sample are Micron based D9WCW which translates to MT61K256M32JE-14:A. They are specified to run at 1750 MHz (14 Gbps GDDR6 effective).
Below is a photo of the card and its RGB LEDs sitting on the test system.
Below are images of just how big this card is relative to the Founder’s Edition. Be sure to have room in your case to fit this video card in there!
Test System and Benchmark Methods
Our test system is based on the latest mainstream Intel platform, Z370, and uses the i7-8700K 6c/12t CPU. The CPU is overclocked to 4.7 GHz on all cores/threads with cache set to 4.3 GHz. The clock speeds used provides a good base to minimize any limitations the CPU may have on our titles, particularly when using the lower resolutions, and should be attainable with a good air cooler or better. DRAM is in a 2×8 GB configuration at 3200 MHz with CL15-15-15-35-2T timings which is a middle of the road option that balances performance and cost.
|Test System Components
|ASRock X370 Taichi
|Intel i7 8700K @ 4.7 GHz / 4.3 GHz Cache
|EVGA CLC 240
|2×8 GB G.Skill Trident Z 3200 MHz CL15-15-15-35
|Toshiba OCZ TR200 480 GB (OS + Applications)
|EVGA 750W G3
|NVIDIA RTX 2080 and RTX 2080Ti (411.51 drivers)
Thanks goes out to EVGA for providing the CLC 240 CPU Cooler and 750 W G3 Power Supply to cool and power the system, G.Skill for the Trident Z DRAM, and Toshiba OCZ for the 480 GB TR200 SSDs storage running the OS, benchmarks, and games. With our partners helping out, we are able to build matching test systems to mitigate any differences found between using different hardware. This allows for multiple reviewers in different locations to use the same test system and compare results without additional variables.
Below are the tests we run with a brief description of the settings. We have made some significant changes since the last update adding a few new titles and dropping some of the older games. More details can be found in the GPU Testing Procedure article which we have updated with our latest benchmarks.
- UL 3DMark Time Spy – Default settings
- UL 3DMark Fire Strike (Extreme) – Default settings
- Shadow of the Tomb Raider – DX12, “Highest” preset (will add RTX when it has been patched)
- The Division – DX12, Ultra preset, VSync Off
- Ashes of the Singularity: Escalation – DX12, Crazy preset, GPU focused
- Far Cry 5 – Ultra defaults
- F1 2018 – Very High defaults, TAA, and x16 AF, Australia track, show FPS counter
- World of Tanks: Encore Benchmark – Ultra defaults
- Final Fantasy XV Benchmark – High defaults
Our first set of benchmarks hail from Underwriters Laboratories who acquired Futuremark back in 2014. Earlier in 2018, a rebrand occurred and since that time, Futuremark is now UL. The benchmarks have not changed, just the name. We chose to stick with 3DMark Fire Strike (Extreme) and 3DMark Time Spy as these tests give users a good idea of performance on modern titles.
3DMark Fire Strike (Extreme) is a DX11-based test which UL says the graphics are rendered with detail and complexity far beyond other DX11 benchmarks and games. This benchmark runs at 1920×1080. 3DMark Time Spy is a DX12 benchmark designed for Windows 10 PCs. It supports new API features such as asynchronous compute, explicit multi-adapter, and multi-threading and runs at 2560×1440 resolution.
In our first tests, we are able to see the slightly higher clock MSI card is slightly faster 3DMark Fire Strike (Extreme) by around 1.6% scoring 12776. 3DMark Time Spy shows similar improvements over the FE scoring 10611 here.
Moving on to the gaming benchmarks, we have updated our testing suite to bring more modern titles into the mix. Gone are GTA V, Crysis 3, and Rise of the Tomb Raider, which were replaced with Shadow of the Tomb Raider, World of Tanks: enCore benchmark, F1 2018, Final Fantasy XV benchmark, and Far Cry 5. We kept The Division and Ashes of the Singularity (though we updated to AOTS: Escalation). The games should provide a good view of the overall performance as many of these are DX12 games.
Sadly, we will not be able to test some of the DLSS features as we are having issues with downloading the file (working with NVIDIA to get it sorted). Ray Tracing will also not be tested here as none of the titles out that we have, currently support the technology. In the future, SoTR will have it along with many other titles so we will circle back when appropriate. As of the time of this writing, Windows now supports Ray Tracing API, we are waiting on games to be patched and released at this point.
The results below are all in 1920x1080p which will skew the perception of the cards when comparing them to the others with the resolution being so “low”. The RTX 2080 and RTX 2080 Ti were not intended for 1080p gameplay but for higher resolutions, in particular, 4K. Those results are found towards the end of this section.
In our World of Tanks enCore results, we are able to again see very little difference between the FE and MSI card both hitting 228 FPS. In F1 2018, we see a two FPS difference which is about a 1.5% gain.
Moving on to Far Cry 5, in this test, the MSI offering against runs just a tad faster, by 2 FPS versus the FE counterpart while in The Division, we do not see a difference with the card averaging 129 FPS.
Moving on to Shadow of the Tomb Raider and Final Fantasy XV, the overall result doesn’t change here with the MSI card matching the FE for all intents and purposes.
Little difference in Ashes of the Singularity at this resolution as well, but it is a fairly CPU heavy title and tends to show less scaling than more GPU bound titles at almost any resolution.
2560×1440 and 4K UHD Results
Below are the higher resolution results starting with 2560×1440 and the gaining in popularity 3840×2160 (4K UHD). These resolutions are more fitting for the cards we are testing as the 1080p results with these cards can have a glass ceiling on them from the CPU (even at 4.7 GHz).
Across the 2560×1440 results, we aren’t seeing much of a difference here either between FE card and the MSI offering with its slightly higher clock speeds.
Moving up to 4K UHD we start to see a bit of separation but nothing to write home about. These cards in this testing environment with these titles perform close to the same with the MSI RTX 2080 Gaming X Trio perhaps having a slight lead.
Overclocking with the MSI RTX 2080 Gaming X Trio was as straightforward as the FE cards. The overclocking results we see in the graphs above are with increasing the memory speeds as well. In the FE review, I let the OC Scanner currently only found in the EVGA software do its magic which only raised the core. I tried that this time around as well and it ended up with a “+96” value. I plugged that into the core and it worked without issue. Adding +96 to the core resulted in 1611 MHz base clock with 1956 MHz boost. When testing, this resulted in a clock speed of 2070 MHz which is seemingly around average if not a bit above. The memory topped out at 2002 MHz (8000 MHz GDDR6). Any more clocks on both and we started to see artifacts as well as starting to hit power limit which would then drop boost bins. Along those lines, the power limit is set to 10% here which with an MSI stated TDP of 260 W, this will allow the card to accept around 286 W before it hits that limit.
The latest edition of MSI Afterburner is pictured below. While it will control (but doesn’t officially support) the Turing based video cards, it doesn’t have the NVIDIA OC Scanner integrated so those wishing to use the MSI software with the MSI card and the OC Scanner are out of luck until a new version is released which includes it. That said, the software is able to adjust the fans, voltage, and overclock the cards manually.
Temperatures and Power Use
We test power consumption by running through the game benchmarks of Shadow of the Tomb Raider and F1 2018 at both stock speeds, and while overclocked. We monitor temperatures throughout this testing with the peak temperature what is listed in the data below. In order to more accurately simulate real gaming conditions, the benchmarks are extended (time) to allow the card to settle.
Power consumption for the MSI RTX 2080 Gaming X Trio has surprisingly come in a bit less than the FE card we have tested by a couple of percent. In stock form, the card peaks at 381 W at the wall in F1 2018 while SoTR hits 357 W. Overclocking the memory and core results in F1 2018 hitting 407 W and SoTR 383 W. Idle use sits at 72 W with this driver. It is reported that the latest driver, 411.70, lowers idle power use. To that end, we saw a 4 W drop on system power use when using the new driver, about a 6% reduction in idle use to 68 W.
Temperatures on this massive cooler proved to be notably better than the FE edition with the card peaking at 67 °C while overclocked. The fans barely ramped up at that speed, but still kept the card cool during our testing. The TRI-FROZR design features Zero Frozr technology which turns the fans off below a specific temperature threshold so idle browsing and light work workloads, the card is completely silent.
Now that we have seen how the card performs its time to talk the end game. Priced at $849.99 at Newegg the MSI RTX 2080 Gaming X Trio is listed above the FE card pricing by around $50 (MSRP for the MSI card is $829, note). What users get for the additional $50 (or around 6% more) over NVIDIA cards is a much larger, quieter, and more efficient heatsink along with what some will say improved aesthetics. If a user happens to be looking for RGB LEDs, a partner card such as this one will have to be the way to go. This price point places the card right in the middle of most other partner RTX 2080s with some coming in less expensive and some more. For example, the EVGA RTX 2080 XC Ultra is priced at $849 as well using a dual fan cooling solution slightly increased core clocks with the MSI solution offering the faster out of box solution by specifications (we will have a review of that EVGA RTX 2080 XC Ultra in a couple of weeks). The difference between brands coming from the heatsinks used, power delivery, and clocks out of the box.
About the only complaints with this card simply come from the sheer size of it. Measuring in over a foot long and 5.5″ high, potential buyers will need to make sure whatever case they are plan to use will fit the beast. Outside of that, the card was silent on idle/desktop and when gaming, the default fan curve proved to work quite well keeping the card cool and temperatures down. In fact, we saw a 10 °C difference between the FE temperatures and this card with a presumably higher TDP. It is clear the TRI-FROZR cooler does the job well.
With the performance of the card so close in our testing, the value proposition for this card, and what will end up being most cards, comes down to its ability to keep the card cool and quiet during normal operations and in particular overclocked scenarios. In this case, the MSI RTX 2080 Gaming X Trio proved to be a solid performer on that front running notably cooler and quieter, even with a higher base TDP and during overclocking. If a user is planning on pushing things and enjoys quiet operations, partner cards like the MSI Gaming Trio X look to fit the bill and at a reasonable increase in price over the FE models.
Joe Shields (Earthdog)