Now that the GF110 is out in the form of the GTX580 (what Fermi was meant to be), we are beginning to wonder what nVidia’s strategy will be.
Will they start the long slow process of replacing the 4 series? Will they do something like AMD and allow them to coexist? Well, the wait for that answer is over. Now we have the latest of the GTX 5xx series GPUs out in the wild and we know what nVidia is looking to do.
After the semi-botched job of the GTX480 nVidia was able to pull things together with the GTX470. Believe it or not, this GPU was a much better-selling than the GTX480 was. This was especially true of the OEM non-reference designs which in many cases were able to outperform the GTX480.
So the GTX580 stepped in to fully replace the GTX480. This gives nVidia a real flagship for the lineup (the GTX flavor of GPU), but that leaves a rather big performance gap between the GTX580 and the 470 (even the super clocked ones). At this point nVidia needs something in the middle, not only for performance but also for price. Their answer to this is the GTX570. This $349.99 card should fit nicely between the GTX580 and the 470. So let’s take a look and see if nVidia’s strategy is going to work out.
The GTX570 uses the same GF110 that we find in its bigger brother with a reduced shader count and clock speed. The same hardware power and thermal monitoring system found on the GTX580 is reproduced on the GTX570. We also find the same vapor chamber cooling system. So there is not really much that is new here.
The final specs of the GTX570 come out something like this:
- Core Clock Speed 732MHz
- 480 CUDA Cores at 1464MHz
- 15 PolyMorph Raster Engines
- 60 Texture Units
- 40 ROP units
- 3 x 64 Bit memory controllers [nVidia lists this as 384 Bit]
- 1.28GB GDDR5 memory at 3800MHz
- 219 Watt TDP
- 97c Thermal Threshold
A closer look at the GTX 570 Like we said above there is not a vast difference between the GTX580 and 570 in terms of hardware and even less of one in terms of physical appearance, as witnessed in the below image.
The test system build and comments
- GTX570 and GTX580 (both supplied by nVidia)
- Asus EAH5870 V2 S.T.A.L.K.E.R. Call of Pripyat Edition (supplied by Asus)
MSi N465GTX Twin Frozr II (supplied by MSi)
eVGA GTX480 Hydro Copper (supplied by eVGA)
- ASUS EAH5850TOP DirectCU (supplied by ASUS)
- AMD Radeon HD 5870 (supplied by AMD)
- AMD Radeon HD 5970 (supplied by AMD)
- Zotac GTX 285 AMP! Edition 1GB, flashed to stock BIOS (supplied by Zotac)
Before I begin, let me just note that overclocking is always going to be a hit and miss thing. The speeds that we are able to reach do not indicate what all cards are capable of. We can show you potential only, your speeds will vary based on the systems you use and the individual cards purchased.
With that in mind, overclocking the GTX570 was not much fun. I was not able to get much more than 815 on the core and 1950 on the memory clock. Added to this was my frustration with the nVidia system tool continually dropping the overclock back to stock speeds (sometimes in the middle of the app).
After a few annoying hours I had to just deal with the minimal overclock run through the apps. The good thing here is that companies like Asus, eVGA, GIGABYTE, and MSI will be producing non-reference cards based on this GPU that will have voltage tweaks to help stabilize the GPU at higher clocks. I am pretty sure that once these are out we will see the true overclocking potential of the GTX570 come to light.
Synthetic tests are nice for many reasons; they offer an easy to use repeatable standard for testing. Most are readily available for free on the internet and can be used by anyone. But unfortunately they rarely tell the full story. Still they are an important part of the story and need to be included in any testing.
Our typical synthetic tests are Futuremark’ s 3DMark Vantage, Furmark, Stalker Call of Pripyat, Direct Compute, and Unigine’s Heaven benchmark for both DX10 and DX11. With these three tests we can give you a good idea of base performance of the GPU in question.
3DMark Vantage is one of the benchmarks that everyone knows. It is used commonly for bragging rights. If you have the highest score, you win. The suite of tests covers the gamut of DX9, DX10, AI and Physics processing that you would normally see in gaming. We will run both the performance test as well as pushing the cards to complete the Extreme test run as well.
As we said above, the GTX570 fits in perfectly right behind the GTX580. It is even ahead of the HD5970 in this run of the benchmark.
In our Extreme run the GTX570 also falls into line in the upper range of grouping. Not too shabby, however let’s not forget that this is only a synthetic and not the whole story. We have more than a few items to get through before we have a clear picture.
The new benchmark from Futuremark, 3DMark11 (check out our review) takes off where Vantage stops. Here you get Tessellation, Direct Compute Physics and much more. We will be transitioning to this new benchmark now that it is officially out and on the market.
Again the scores look very good from the small sampling we have here. We will be adding more detail to this graph as we now have access to the scoring site and get the full details with each test run. However we should most definitely note that the HD5970 is out in front as you would expect a dual GPU card to be.
Unigine Heaven (DX10 and DX11)
Unigine’s Heaven bench was one of the first tests that offered a good measure of DX11 performance. Although we now have DX11 cards from both AMD and nVidia we are still going to maintain the DX10 runs just to see how these GPUs handle the more bulky DX10 Code.
Under Unigine’s DX11 Render engine the GTX570 does well but the eVGA GTX480 manages to run a little faster than the 570 by a handful of points. We do see some decent results from the very small OC we managed but as your mileage may vary it is really only there to represent potential. Of course it is important to remember that the GTX480 is now EOL and no more will be made moving forward.
Under DX10 we see the HD5870 squeak by with about 13 points and .2 FPS. Not anything that you would ever see if you were sitting in front of the monitor though.
Stalker Call Of Pripyat Bench (DX10 and DX11)
The Stalker Call of Pripyat test uses the stalker rendering engine and images. It is capable of emulating the most common effects from DX11 and DX10 in a run of different scenes. We ran our GPUs through both and have recorded the average frame rate below. We sorted this bench by the ‘SunShafts’ test as it appeared to be the toughest on the GPUs in question.
With DX11 the GTX570 performs as we would expect. Right behind the GTX580; unfortunately the AMD cards outperform it. But then again this is an AMD based benchmark.
For DX10, well things are not good for now. On the bright side we spoke with nVidia and found out that this is more than likely an issue with AA on this particular test. We are waiting for them to do some additional testing to confirm this. After that we should see a fix for the problem and the scores even out.
Well, so far the GTX570 is performing as we expected. It is certainly not the fastest card out but it does fill in the gap between the GTX580 and the GTX470. This is an important spot to take as is helps to close up the upper end of the market. Of course we are not done yet. We still have the real world stuff to cover.
As we mentioned above we use both synthetic and real-world testing to give you the full picture of any product. For GPUs the best real world test is in-game testing. Time Demos and benchmarking tools with render engines cannot come close to showing what a graphics card can really do once it is in you system. As such we have chosen five games two DX9, two DX10 and two DX11.
These games are Metro 2033, Just Cause 2, Bioshock 2, Battlefield Bad Company 2, FarCry2, and Modern Warfare 2. We played a specific level of each of these and used FRAPS 3.0.3 to measure the ability of our test subjects to render each game. All of the charts below are sorted based on the minimum frames per second each card was able to achieve. We have also pulled out the DX10 based GTX 285 from all of the DX11 games.
Modern Warfare 2 (DX9)
The final sequel is Call of Duty Modern Warfare 2. This game brings the stress and chaos of modern combat to life. It is a very short but very intense game, in it you take on multiple roles as you move through different locations around the globe and engage in all forms of combat. For testing I ran from after you are selected until the liberation of the school. Settings are shown below.
Modern Warfare 2 is an excellent game. It is one that still gets play time even with the issue over dedicated servers. We keep it around for now just for this reason. In terms of performance the GTX570 shows up a little behind the EAH 5870 V2 from Asus (by two FPS). In other words you would not be disappointed if you bought one of these and Modern Warfare was your thing.
Battlefield Bad Company 2 (DX11)
Another game in the BF franchise this one takes a stab at a more modern form of warfare. You do start off toward the end of WWII but soon are in the future (present?) for the single player campaign. The game features some decent AI (more mob style than anything) and has some great weapons. Our test run is the entire first mission. Settings are shown below.
With Bad Company 2 the GTX570 does pull out a little in front of the EAH5870. Granted it is only by 6 FPS. This is right on the threshold of what you would be able to notice at these speeds.
Farcry 2 is the third of the sequels we used for testing. It is a large sandbox game that does not truly have levels like most first person shooters. For testing I played from after you wake up until you free the hostage. Settings are shown below.
Oddly enough we did not see the performance we would have thought with FarCry2. In fact we see performance that is behind the GTX480. We will have to check into this one but we have a feeling it is a driver issue as the ones we used are fairly new.
Bioshock 2 (DX10)
Another sequel shows up in our testing, this one pits you in an undersea utopia gone horribly wrong. The engine is based off of the Unreal engine and offers us some DX10 surfaces to help increase the rendering load. For testing I ran the level from the beginning of the game until you find Dr Tenenbaum. Settings are shown below.
Well what did you expect from a TWIMTBP game? The GTX570 is still in its place behind the 580 though, so things are certainly firming up with the 570 and the marketing position that nVidia wanted to put it in.
Metro 2033 (DX11)
Metro 2033 is another in the post-apocalyptic genre. This one takes place in the year 2033 20 years after a terrible event took place that almost wiped out mankind. The survivors took refuge in metro stations in Moscow. Now you must venture out to prevent the death of the few remaining people left.
The game itself plays a lot like Stalker; it is slow and filled with mutated horrors (4A games, was founded by developers that worked on the X-Ray Engine used in STALKER). For our testing we played for 30 minutes starting with your exit from the station. Settings are shown below.
Ok, I have to say this. It was nice to finally play this game with something resembling a good frame rate at minimum. I felt like I was back playing Crysis. However both the nVidia GPUs do a great job here with the GTX570 coming close to the 32FPS make for its minimum performance.
Just Cause 2 (DX11)
This one is plain cheesy, but is a decent one for third person action. It is also a DX9 console port. The game put you as a CIA operative looking to regain some lost data cards and destroying everything in your path. The graphics are ok, but the AI and general story are terrible.
For Just Cause (with the DX11 extensions) we find that the GTX570 does well again. It is right behind our overclocked EAH5870. This is not a bad showing at all.
We have to say that the GTX570 has shaped up to be what we were expecting it to be. It is an upper-end GPU that is designed to fill the performance gap between the GTX580 and the GTX470. We see from the gaming performance here that it can certainly handle the demands of most of today’s popular games without too much of a strain.
We also have noticed an increase in image quality. nVidia has returned to what really matters to most gamers. Sure amazing frame rates a are nice, but once you hit a certain level it is overkill. They have shifted focus and are working on improving the look and feel of the rendered images as well as the frame rate. This is nothing but good news for gamers.
According to everything we have heard the GTX570 will hit the streets around $349.99. This puts it right in line with the market it is aimed at. The performance level is on par with the cost as well. Of course this is not a “set in stone” price as some OEMs will add their own tweaks and swag with their versions that will cause the price to vary somewhat. However that does not kill the value of the GPU. I think that nVidia has overcome a serious hurdle with the new GF110 GPU and have decided to keep the price in line with the market (which is a smart move).
The GTX570 is aimed at a very interesting segment of the market. This segment is populated by people that want good solid performance but do not have a ton of income (or the sense not to spend it all on a GPU) as such the people looking to buy this level of GPU are much pickier than your “I have to have the best” buyer or even the value minded buyers. To court this buyer you have to show good value for the product along with some head room to handle things ahead in the road map. I think nVidia has done that with the GTX570.
It is a card that can handle current games and looks to have the head room to handle the ones coming out later this year and early next as well. Although I have not always been impressed with nVidia and feel that they could have executed a little better on the GF100 (and other things) I am impressed with the recovery they have pulled off with the GF104 and the GF110.