Introduction
The RIVA TNT is out for several months now and most people agree that it is currently the best 2D/3D solution available. Its only real competitor is the cheaper Voodoo Banshee, a 2D/3D chip that is performing worse than TNT, but which is taking advantage of the powerful brand name 3Dfx. The performance of TNT is only surpassed by 3Dfx’s Voodoo2 chip, but you need two add-on 3D-only accelerator cards with this chip to really achieve better performance than TNT. Two Voodoo2 cards are considerably more expensive though and you won’t necessarily experience a major difference in 3D gaming. The 3D image quality of TNT is definitely way ahead of products with 3Dfx chips and only the upcoming Rage 128 chip from ATI could give NVIDIA’s RIVA TNT a real run for its money.
TNT cards are usually running at 90 MHz chip clock and 110 MHz memory clock by default, but in many cases it’s possible to successfully overclock the TNT. The chart above shows that a RIVA TNT overclocked to 120/125 MHz can easily surpass ATI’s upcoming Rage 128 and it is getting pretty close to the 3D performance of two Voodoo2 cards in SLI mode. All in all the TNT is certainly the most attractive 2D/3D solution available today, so that you only need to know which particular TNT card to get.
There are currently seven card makers who supply a 2D/3D card with the RIVA TNT, six of them are represented in this test. The only missing card is from Canopus, it was unfortunately not available to me for this test.
Comparison
The cards will be compared by looking at the hardware as well as software features, of course the 3D performance at default clock settings, there will be a heat comparison between the cards, overclocking is another important issue and finally there will be a closer look at the video output ability. There won’t be a 2D benchmark comparison, because the 2D performance of all cards is identical, regardless if using Windows 9x or Windows NT.
Hardware Features
Feature | Memory Size | Memory Type | Cooling | Video Out | Video In |
Asus V3400TNT(/TV) | 16 MB | SDRAM | Active | Composite and S-VHS (optional) | Composite and S-VHS (optional) |
Creative Graphics Blaster RIVA TNT | 16 MB | SDRAM | Passive | None | None |
Diamond Viper V550 | 16 MB | SDRAM | Passive | S-VHS | None |
Elsa ERAZOR II | 16 MB | SGRAM | Passive | None | None |
Hercules Dynamite TNT | 16 MB | SDRAM | Active | None | None |
STB Velocity 4400 | 16 MB | SDRAM | Passive | Composite and S-VHS | None |
From the hardware-features point of view, the Asus V3400TNT/TV is the clear winner. It’s the only card that provides a video input and it comes with the best cooling device of all cards. Diamond’s and STB’s TNT card are both providing a video out feature, but it is less convenient as well as of lower quality compared to the Asus V3400TNT/TV. The only other card with an active heat sink next to the V3400TNT comes from Hercules, but so far there is no version of the Dynamite TNT available with video out. Elsa’s Erazor II used to have an active heat sink in the first shipments, but now it comes with a passive heat sink only. The heat report will show why this seems to have been a bad move by Elsa. The Erazor II is the only card that is using the more expensive SGRAM instead of SDRAM. Unfortunately that does not influence 3D performance in any positive way.
Software Bundles – Asus V3400TNT(/TV)
No software bundle
Software Bundles – Creative Graphics Blaster RIVA TNT
- MultiMedia MM200 SE from Scala
- Forsaken™ by Acclaim Entertainment
- Incoming from Rage software
Software bundle North America:
- Forsaken™ by Acclaim Entertainment
- Colorific® by Sonnetech, Ltd.
- 3DEEP by Sonnetech, Ltd.
Software bundle Europe:
Software Bundles – Diamond Viper V550
- Motorhead from Gremlin Software
- PhotoSuite SE from MGI Software
- FremontTM SE from Portola Dimensional Systems
- SoftDVD-Decoder (Region 2)
Optional software bundle North America:
- Motorhead (full version) by Gremlin Interactive Ltd.
- FremontTM SE by Portola Dimensional Systems
- Microsoft’s Game Sampler for Windows® 98
- CrystalGraphics 3D SensationsTM
- Software DVD (Region 1)
- MGI’s PhotosuiteTM SE
- PLATINUM’s VRCreator/Learning Edition
- PLATINUM’s WIRL
- BackWeb Software
Optional software bundle Europe:
Software Bundles – Elsa ERAZOR II
- Need For Speed III
- Recoil
Software Bundles – Hercules Dynamite TNT
No software bundle
Software Bundles – STB Velocity 4400
- NFL Blitz
- 3Deep
- Web3D
- Soft-DVD-decoder
I don’t want to say much about the software bundles, since it’s pretty much a question of taste. I think that a software DVD player and the games Need For Speed III and NFL Blitz are the most attractive additions to a TNT card, but you may have different opinions. Hercules and Asus are not providing any software with their cards, which could be a negative thing for some of you. I personally prefer good hardware over an average hardware/nice software deal.
Test Setup
To make sure that only the hardware is influencing the benchmark, I was benchmarking all cards with NVIDIA’s reference driver 0048. Then I ran each card with its own driver. It was surprising to find out that there was no difference between the reference driver result and the driver provided by the card maker. Some card makers want us to believe they are tweaking their drivers, but my test showed that this is pure hype. The performance of the reference driver was identical to the generic driver in all cases.
The benchmarks were run on an Asus P2B-LS BX-motherboard with Intel Pentium II 400 CPU, 128 MB SDRAM and an IBM DGVS 09U UW-SCSI hard drive. Operating system was Windows 98, the screen resolution for both games was 1024×768, both at 16 bit color depth, the refresh rate was 85 Hz. Each card ran with 90 MHz chip clock and 110 MHz memory clock , the default clock recommended by NVIDIA. The Quake 2 version used was 3.20.
Driver | Version |
NVIDIA Reference Driver | 4.10.01.0048 |
Asus V3400TNT | 4.10.01.0102 |
Creative Graphics Blaster RIVA TNT | 4.10.01.2104-01.037 |
Diamond Viper V550 | 4.10.01.0239 |
Elsa ERAZOR II | 40.10.01.0200-0016 |
Hercules Dynamite TNT | 4.10.01.0048 |
STB Velocity 4400 | 4.10.01.0142 |
There wasn’t much of a performance difference expected between the six cards, only Elsa’s Erazor II should perform a bit better than the others, as it’s the only card using the more sophisticated SGRAM. The benchmarks showed a different result. When running each card at the default 90 MHz chip and 110 MHz memory clock, the only thing that made a difference was the memory timing. The SGRAM of the Erazor II seems to make it slower than almost all of the others.
3D Performance
You may be surprised that you find two results of the Asus and Creative cards. Those two cards are using more conservative memory timings, resulting in higher reliability, but lower performance. If you change the memory timings to the same as used by Diamond, Hercules and STB, e.g. by using Powerstrip, you get a significantly higher performance. I did not come across any instabilities after switching the two cards to the faster memory settings, they don’t become more or less stable than the Diamond, Hercules or STB cards. Elsa’s Erazor II can not be switched to the fast memory timings unless you want to run into a system crash. Thus the Erazor II is the real loser in the test, all the other cards are close to identical once fast memory timings are being used. You shouldn’t blame Asus or Creative for trying to ensure highest stability, both cards can be as fast as the other three if you wish.
Heat Report Part 1
Everybody who knows about NVIDIA’s latest high-end 3D chip is aware of it, the RIVA TNT is getting pretty hot. Thus the card manufacturer should provide some way of cooling, either a passive heat sink, or better a fan/heat sink combination. Testing the chip temperature with your finger may be pretty common, but is miles from accurate. I was using a thermal probe, placed at the PCB opposite the TNT chip. The chip is a BGA model, which ensures pretty good heat dissipation to the PCB over the whole mounting surface of the chip. This is the fairest place to put the probe, because this place is identical for all cards. Fixing it to the heat sink would not have been a smart idea.
The chips get really hot only when running a 3D application, the windows desktop keeps them at about 5-8єC less. I ran my 24-minute ‘sewer64’ Q2 demo ‘tbone.dm2’ at 1024×768 and waited until the chip reached its highest temperature. Please realize that the actual chip temperature is higher than what I measured at the other side of the PCB. Another thing to consider is that I tested with a completely open system, letting lots of air to the chip and heat sink. The temperatures measured inside a computer case will be at least another 5єC higher.
The difference in temperature is pretty significant as you can see:
At default clock, only the two cards with active heat sinks can ensure a temperature of less than 60єC. STB’s Velocity 4400 is already reaching more than 70єC at default chip clock, which I consider as pretty scary. In a closed case the Velocity can easily reach 80єC or even more.
All cards are still running stable at 100 MHz chip clock, but Elsa’s Erazor II and STB’s Velocity 4400 are getting damn hot.
At 105 MHz chip clock the Elsa and STB cards crash. This is most likely due to their highly ineffective cooling. The cards from Diamond and Creative are also pretty hot already, but they still run fine.
Heat Report Part 2
At 110 MHz and 115 MHz the air is getting pretty thin. Only the two boards with active cooling are still in the race. The Asus card seems pretty unimpressed by the high clock, which shows the superiority of the Asus cooler.
Finally the V3400TNT(/TV) is all alone running at an incredible 120 MHz chip clock, 33% over the default clock.
The value of an active heat sink is pretty obvious now. I doubt that it was a wise decision when Elsa decided shipping the Erazor II without a fan anymore. Older Elsa Erazor II cards with the heat sink perform pretty much like the Hercules card from the temperature point of view. STB should wonder if they couldn’t use a heat sink that’s at least a bit better than what they use now. Diamond and Creative have decent heat sinks, but they certainly don’t deserve any kind of technical award for it. Hercules went into the right direction, Asus made it perfect.
Overclocking Part 1
Overclocking is a big issue with TNT, because almost each TNT card can run faster than the default speed of 90/100 MHz for core and memory clock. However, overclocked chips get hotter, so that a decent cooling should be provided. Let me first show you the performance impact of overclocking TNT:
Increasing the chip clock has definitely got a higher impact on 3D performance than increasing the memory clock. Running the TNT chip at 105 MHz instead of 90 MHz gives you at least a 10% improvement in Quake2’s Demo1 benchmark. A card that can be overclocked as high as my Asus V3400TNT to incredible 120/125 MHz reaches a 26% 3D performance improvement, something that’s certainly noticeable.
Please consider that overclocking a 3D chip is following the same rules as overclocking a CPU. Not every chip has the same ability to run at the same high speeds. Even in case of an excellent cooling as provided by the Asus TNT boards, you may not be able to reach the 120/125 MHz setting. However, a chip that is cooled as badly as in case of the STB Velocity 4400, will hardly ever be able to reach high clock speeds. The same is valid for the memory. Some memory chips can cope with up to 125 MHz, others can’t. Memory chips are a bit more consistent though.
Overclocking Part 2
The only card in the test that already ships in overclocked mode is the Hercules Dynamite TNT. The chip runs at 98/125 MHz by default and Hercules is the only manufacturer that provides a full overclocking utility (Hercumeter) with its drivers. Hercules guaranties that the card will run at the 98/125 MHz, which already gives you a 12% performance edge over any other card in this test. Creative’s Graphics Blaster RIVA TNT gives you the chance to at least overclock the memory via its drivers, for all the other cards you’ll need a utility.
The ‘Hercumeter’ from Hercules.
Creative Lab’s tool for overclocking the video memory.
Overclocking Part 3
The best and pretty well known tool for overclocking graphic cards is Powerstrip from EnTech, Taiwan. In the ‘Advanced Options, About the Powerstrip’-menu there is a submenu ‘Performance’, which lets you switch to fast memory settings and which provides sliders for adjusting the memory and chip clock.
To make sure that a card can really cope with the new speed, I was running my 24 minute Quake2 ‘tbone.dm2’ demo, which is based on my favorite map ‘sewer64’. This demo is giving the whole system a really hard time and will make any unstable card crash. Sometimes you only see triangle drawing errors, which is a safe sign that you overclocked too much. Here are the results that I found:
Product | Highest Chip Clock Setting | Highest Memory Clock Setting |
Asus V3400TNT | 120 MHz | 125 MHz |
Creative Graphics Blaster RIVA TNT | 105 MHz | 115 MHz |
Diamond Viper V550 | 105 MHz | 115 MHz |
Elsa ERAZOR II | 100 MHz | 120 MHz |
Hercules Dynamite TNT | 115 MHz | 125 MHz |
STB Velocity 4400 | 100 MHz | 120 MHz |
The best memory seems to be used by Asus and Hercules and those two would also allow the highest chip clocks due to their active cooling devices. Diamond and Creative let the chip clock go up to 105 MHz, but the memory can’t cope with more than 115 MHz.
Overclocking Part 4
This chart shows you how much 3d performance could be achieved by each card when running it at the fastest settings:
Logically the V3400TNT is the fastest, because it was overclocked the most, Elsa’s Erazor II was the slowest in the test with the default settings and due to its bad heat sink it cannot go up any further than 100 MHz so that it only reaches the last place. The Creative Graphics Blaster RIVA TNT was also using the ‘fast memory timings’ from Powerstrip.
Summarizing the overclocking comparison shows that the best boards for overclocking are clearly the Asus and the Hercules board. The Dynamite TNT is more convenient, because you don’t need Powerstrip for overclocking, but the V3400TNT is cooled better. The STB and Elsa board should rather be avoided by overclockers.
Video Out
Three of the test-cards offer a video out option. Now this feature is mainly interesting for two things. The most important one is probably the ability to play games on your big screen television. In this case please consider that a resolution of more than 800×600 is neither possible nor sensible. For US and Canadian citizens using TVs with the NTSC-standard, it’s even questionable if a resolution of more than 640×480 is making sense. NTSC is only displaying 480 lines, so that the 600 lines of the 800×600 resolution cannot get displayed properly anyway. The European PAL is providing 600 lines, which makes a 800×600 screen look a lot better than on NTSC. Still, the 480 or 600 lines are both only a theoretical maximum, in reality the TV output resolution is about 5-10% less.
The second use of video out is using the computer as your DVD-player. In this case you don’t have to care about the above said.
The most interesting thing about video out is its quality. Does it flicker? Can you recognize everything? Does it fill the TV-screen, or is it too large or too small? Can you use your computer monitor at the same time and if yes, do you have to use the low TV refresh rate?
To give you some kind of idea what it looks like, I’ve captured the video output of each card with the Matrox Marvel G200. The Marvel is excellent at capturing snap shots and highly impressive for video editing. You can see a good difference between the three cards, but please note that the Asus card is PAL, the other two are NTSC.
Asus V3400TNT/TV
Diamond Viper V550
STB Velocity 4400
The quality of the Asus card is clearly best, but it has got an advantage due to the PAL format. Diamonds card needs to be adjusted to maximum flicker reduction, otherwise the picture flickers terribly. The Velocity 4400 flickers permanently, regardless if you switch the flicker filter to ‘for games’ or ‘for text’. The V3400TNT/TV is also the only card that can display on the TV as well as on the computer monitor at the same time. You can even adjust a high refresh rate for the monitor, leaving the TV picture untouched.
The Asus V3400TNT/TV is clearly the most convenient card for video out, but I cannot say if the quality of it is indeed better than of the Diamond card, because I would need to test the NTSC-version of the V3400TNT/TV for that. The Diamond Viper V550 is clearly superior to the STB card when it comes to video out.
Price
Product | Lowest Price as of January 5, 1998 |
Asus V3400TNT | $125 |
Creative Graphics Blaster RIVA TNT | $125 |
Diamond Viper V550 | $127 |
Elsa ERAZOR II | $129 |
Hercules Dynamite TNT | $133 |
STB Velocity 4400 | $116 |
There isn’t really that much of a price difference between the cards. STB’s Velocity is the cheapest, which is fair enough, considering the fact that it’s good for frying eggs. Elsa’s Erazor II is definitely too expensive. It’s pretty much the worst card in the test and one of the most expensive ones as well. The price for the Hercules Dynamite TNT seems a bit on the high side too, considering that there’s no software bundle.
Recommendations
It has hardly ever been as easy to give my recommendations as in this graphics card comparison. The only TNT cards that I would buy are the Asus V3400TNT or even better the Asus V3400TNT/TV as well as the Hercules Dynamite TNT. Although those cards are the only two that come without a software bundle, the hardware and usability features speak for themselves. The Asus cards have the best cooling, good drivers, excellent ‘overclockability’ and the V3400TNT/TV has the best video-out feature and is the only one with video-in feature as well. The Hercules card also comes with very good cooling, the best drivers and a guaranteed ‘overclockability’ to 98/125 MHz.
Asus V3400TNT
There are two different V3400TNT cards available from Asus, the AGP-V3400 and the AGP-V3400TNT/TV. The latter comes with video output as well as video input, both either as composite or as S-VHS signal. The video input is unique amongst the tested cards and it can prove pretty useful for things like video conferencing or taking snapshots, of course only in case you have got a camera that you can connect to it. This is the little program provided for the video input:
The video output is also a pretty nice feature, particularly for the ones of you that like playing games on their TV screen. Of the three cards with TV-out in this test, the V3400TNT/TV is the only one that can run both, the computer monitor and the TV screen at the same time, you can even choose any refresh rate for your computer monitor, it won’t affect the TV output. From the card-feature point of view, the V3400TNT/TV is definitely the most attractive card in the test.
The V3400TNT is one of the two cards in this test that come with an active heat sink. The fan and heat sink combo is the largest and most effective in this test, although it is not too large to inhibit you in any way. You can still plug in a PCI card right next to the AGP slot. The temperature test showed that the V3400TNT is the by far coolest card in the test, which is most likely the reason why it is also the one card that could be overclocked higher than any other.
The benchmark comparison shows you that at default settings, the V3400TNT is the slowest card in the test. However, the picture changes completely when you tweak the V3400TNT a little. Then it becomes one of the fastest. The reason for those strange results is Asus’ conservative policy. HC Hung, the man in charge of the graphics cards division at Asus, told me that he found some minor instability with the fast memory settings suggested by NVIDIA. He told me that Asus’ high reputation for excellent quality may not be jeopardized by a graphics card, which is why Asus is using the conservative slow memory settings. If the customer prefers having the same (in)stability as provided by Diamond, STB or Hercules, all of them using the fast memory settings, he can change the memory settings to achieve the same performance as the just mentioned cards. You will require Powerstrip to change the memory timing.
Asus’ drivers are offering a wide spectrum of tweaks, but they would not let you change the chip or memory clock or the memory timing. The driver lets you change the refresh rate of your monitor, which is a very practical feature in my eyes. For the proper screen adjustment Asus is providing test patterns as well.
Summary
All in all, the AGP-V3400TNT(/TV) is well worthy to win this competition. The only thing I would ask for is a nice overclocking tool within the drivers, just as shown by Hercules.
Creative Labs Graphics Blaster RIVA TNT
Creative’s version of a RIVA TNT card looks a bit similar to the Asus card, but the lack of an active cooling device as well as an optional video out feature already show, that Creative didn’t have the high-end customer in mind when designing the GBRT. As in case of the Diamond card, the memory is only good for a maximum of 115 MHz, although Creative supplies this card with a little tool to overclock the memory. The passive heat sink is not as bad as in case of the Elsa or STB board, but Hercules and Asus showed how to do it better.
The driver of the GBRT leave little to be desired, you can tweak a huge amount of settings, alas there’s no way of changing the refresh rate without using the normal Windows98 ‘Adapter’-settings.
Summary
The Graphics Blaster RIVA TNT is an average board for an average price with average features and average performance. I guess it’s the right TNT card for anyone who feels average enough for it.
Diamond Viper V550
Diamond’s viper V550 is one of the smallest TNT cards, similar to the Hercules Dynamite TNT. It comes with video out and it’s supposed to ship with a Soft-DVD-decoder, which is a convenient thing if you haven’t got a Home-DVD-player yet. The memory of the Viper is only running at up to 115 MHz, which is not a lot and the passive heat sink is also making sure that the chip will never freeze.
The drivers are a bit poor in comparison with the others, there is not much that you can change really. Diamond provides a set of the well-known InControl-Tools as well.
Summary
The Viper V550 is another example for an average TNT card, but it comes with the largest software bundle after all. So if you are interested in the software that’s provided with this Diamond product and if you like the video out, you may want to consider the Viper V550. It’s my 3rd choice in this comparison.
Elsa Erazor II
Elsa is the only German company in this test and that’s why I would love to show some patriotism. Unfortunately Elsa didn’t leave me much positive to go on about.
The Erazor II is the only board in this test that comes with SGRAM instead of SDRAM. SGRAM is more expensive, because the architecture is more sophisticated. The performance should be at least a little bit better than with SDRAM, because SGRAM offers several different block read commands that should improve the memory performance. Unfortunately it turns out the opposite way, the Erazor II is the slowest TNT card in case you tweak the memory timings of the Asus card, something you cannot do with the Erazor II. In the past Elsa shipped the Erazor II with a little fan and heat sink on top of the TNT chip, keeping the temperature of the RIVA TNT at the same levels as the Hercules card. Now Elsa has decided to save the costs for the fan and the new passive heat sink makes it one of the literally hottest boards in this comparison. The memory timings cannot be tuned with Powerstrip, but you can run it at up to 120 MHz. The TNT chip however doesn’t cope with more than 100 MHz, which puts the Erazor II into last place for overclocking as well.
The drivers of the Erazor II are amongst the best available, it offers a very convenient way to adjust your monitor settings. You can’t really blame Elsa for not including an overclocking tool into the drivers, the Erazor II isn’t really designed for that.
Summary
Elsa decided against a video-out option, so that in summary the Erazor II is slower than the others, less overclockable than most, one of the worst cooled TNT card without any additional hardware features, but it comes with good drivers. I wish I could find one reason to recommend this card, but it is even one of the most expensive ones. Too bad for Elsa, I hope the next graphics product will be better again.
Hercules Dynamite TNT
Hercules is having a rough time lately. The Savage3D card was not exactly a success so far and the previous products did also have a hard time against competitors as Diamond or the likes. Hercules got the reputation of being a provider of rather cheap and unexciting products. Then in the second half of 1998 we learned that Elsa was planning to buy Hercules for a rather small sum, which had an impact on Hercules’ reputation again. The worst thing for Hercules was when Elsa announced that the financial situation of Hercules was so bad that they would decline from the deal.
Finally Hercules sat down and thought things over. What was needed was a really good product that would get them back into the positive news. The Dynamite TNT was pretty much the last product with NVIDIA’s TNT chip of all the major card manufacturers, but it turns out to be one of the best as well. It seems as if Hercules did indeed put a huge effort into this card and that seems to pay off now.
The Dynamite TNT is next to the V3400TNT the second TNT card with an active cooler, that does a pretty good job as you can see in the heat comparison. Hercules is courageous enough to team this small and well designed card up with a set of drivers that overclock the Dynamite to 98/125 MHz by default. This makes the Dynamite faster than all the others right after you’ve installed it. The memory and the card layout seem to be very good, because only the Asus card is also able to run the memory as fast as the Hercules card, all others fail at 125 MHz. The Dynamite does not include a video out, but if you look at the board you will find that the layout is ready for it, so that I expect a Dynamite TNT with video out sometimes soon as well. The drivers set is one of the best, and it’s the only one that offers you a full overclocking tool by default. I have got three Dynamite TNT boards to make sure that the quality and stability of this product is not only found in one specially prepared ‘reviewers board’. All boards are running rock stable at the 98/125 MHz, one of them even at 115/125 MHz, the other two go up to 110/125 MHz.
The driver offers a scaringly large amount of features, but there’s no way of changing the refresh rate of the monitor, which I consider as a bit of sad. However, except of that little glitch, the drivers are excellent, particularly for speed freaks.
The Hercules Dynamite TNT is my other winner of this comparison. This board is the number one choice for the 3D gaming speed freak.
STB Velocity 4400
STB was the first company to release a graphics card with the RIVA TNT, but nothing much has changed since then. Today everyone knows that the TNT gets very hot and you would expect that a card maker is aware of that. The heat sink of the Velocity 4400 is the worst of them all, the chip gets up to 71 degrees Celsius already at 90 MHz chip clock only. This is probably also the reason why I was completely unable to clock the card any higher than 100 MHz chip clock, any higher speed would most likely lead to melt down conditions. It wouldn’t surprise me if you come across instabilities in small computer housings at the default 90 MHz already.
The Velocity is equipped with a video out, which is nice, but unfortunately the quality is not up to satisfying any higher expectations, the flicker is visible almost permanently.
The drivers are ok, not particularly impressive but certainly better than the Diamond drivers. Again, the STB driver doesn’t give you the ability to adjust the monitor refresh rate, you have to use the Windows 98 ‘Adapter’ setting.
Summary
The Velocity 4400 is the cheapest TNT-board available, but its heat sink is just about living up to the price of the card as well. The performance of the card is good, but you can hardly overclock it. All in all the STB card gives me the impression as if the designers were rather interested in rushing out a product than spending any time on detail. Since the launch nobody was bothered to improve the board anymore, and that is why the Velocity 4400 is only reaching 5th place in this TNT card comparison of six cards.