Introduction and Package
We have been waiting for months, but Bulldozer is finally here. The weeks leading up to the launch of the new FX line of processors hasn't been the best for AMD with a large amount of leaks that haven't put the new FX-8150 in the best light.
Today, though, leaks don't matter; unofficial results don't matter. We've got all the latest drivers, the latest BIOS updates and it's time to see what Bulldozer can do. There's no denying that AMD are going to have a tough fight on their hands, though. AMD have stated that the new platform is only designed to compete against the current crop of Sandy Bridge CPUs.
The problem for AMD is that the 2600k has just done such a good job cementing itself as the overclockers CPU to get since it launched. Not only that, Intel's next generation high end platform to replace X58 is due out in literally 4 - 5 weeks, it seems. Sandy-Bridge E and X79, while will no doubt come with a higher price tag, is going to shine it seems when it comes to overall performance.
AMD over recent years, though, have done a good job of standing out in the key area that is pricing. At launch we're going to see three FX series processors hit and the most expensive one, the one we're looking at today, comes in at $245 US. Before we get into the performance of the platform, we're going to see some of those other models and have a look at what price point they hit us at.
From there we'll take a closer look at our processor to see what's going on before we talk about our Test System setup and overclocking. From there we'll get into the performance side of things to see just what the new Bulldozer CPU can do.
FX Models and Specifications
The new Bulldozer platform helps round out the new line-up of processors for the main retail market. You can see below we started with the E-Series APU which brings with it a focus mainly on office work. You can see we then have the A-Series APU and that brought with it the A55 and A75 chipset. At this point we start to get into a platform that's able to do editing and transcoding HD media while also offering us CrossFireX support, which is best seen when a HD 6600 series GPU is combined with the onboard GPU built into the APU.
Today, we round it all off with the AMD FX line of CPUs. This brings with it a focus on overclocking, high resolution gaming and AMD Eyefinity support. Having the focus on it is one thing, though. But can AMD deliver in the area? Well, that's exactly what we intend to find out. Let's first dive a little deeper into what's going on with the new line-up here today.
At the moment we know that the new FX line-up will consist of seven parts ranging from four core options all the way to eight core ones which is what AMD have mainly been promoting prior to the launch. Above we're able to get a good idea on the clocks these processors will come in at. The few things that we want to pay attention to here are the fact that we've got two versions of the CPU, a 95w and 125w TDP option depending on what model you go for.
You can also see that the L2 Cache scales with the amount of cores that are present on the CPU. The only other thing we see that differs between some models is the NB speed. You can see some come in at 2.2GHz while others come in at a slightly lower 2GHz. The numbers that are identical between all seven processors included the 8MB of L3 Cache, support for 1866MHz DDR memory and the fact they're all on the AM3+ packaging.
Out of the seven processors, though, only three of them will be available at launch. They include the FX-8150, FX-8120 and FX-6100. As you can see from the below, that's two eight core CPUs and a six core one. At launch there's not going to be any of the four core models.
We won't go into all the details of the three models here, as you can see a larger break down of what's on offer two images up. Instead we'll extract the information here that's really important. The pricing; first we start with the FX-8150. At $245 US this is priced between the i5 2500k and i7 2600k. AMD have always been about the value of their products, so you'd mainly be looking at the FX-8150 against the 2600k. These are the two flagship products from both companies when it comes to this mainstream market. We say mainstream market because we understand the 990X / X58 configuration and the soon to be SB-E platform.
The FX-8150 pricing isn't what impresses me, though; instead it's more the FX-8120 and FX-6100 that comes in at $205 US and $175 US* respectively. While Intel have of course got models that compete with these CPUs in terms of price, something that we always try to remind people of is that the only Sandy Bridge based CPUs that can offer any real overclock include the two K series CPUs; the 2500k and 2600k. Those two CPUs come in at $219.99 US and $314.99 US respectively.
At the $205 US price point we've got the i5 2400S at $200.99 and the i5 2500 at $209.99 and you can see above the advantages of the new eight core AMD processors against the i5 line-up. The stand outs would have to be more cores, x16 / x16 CrossFireX support, and the fact it's fully unlocked.
Move to the $175 US price point seen on the FX-6100 which also is fully unlocked and the situation begins to look really good for AMD. At this price point the closest Intel CPU is the i5-2300 Quad Core processor at 3.1GHz after turbo boost. This comes in at $179.99 US, so just ever so slightly more expensive.
The FX-6100 has on it, though, two more cores +800MHz after Turbo, 8MB of L3 Cache verses 6MB on the i5-2300 and is of course fully unlocked which means that our speed isn't limited to maybe one or two MHz more when overclocking comes into play.
* After this was written, the price for the FX-6100 was updated to $165 US by AMD. It just makes our point even stronger.
FX Models and Specifications Continued
AMD are for the most part comparing the new AMD FX-8150 to the current crop of i5 / i7 Sandy Bridge based processors and you can see on paper that the comparison of them does tend to indeed favor the FX-8150. No surprises here since the slides are from AMD, but let's take a look at exactly what's going on.
Specifically below we can see that the main comparison is the i7 2600k as the Base Frequency and Turbo Frequency of 3.4GHz and 3.8GHz coincide with the i7 2600k processor from Intel. Before we look at this, though, let's just move to the top to see exactly what's going on.
First we've got the memory compare. A good one, but not really a fair one to be honest. While native support on the CPU may only be 1333MHz DDR, with the help of XMP profiles, speeds of 2133MHz DDR are had with absolute ease. What I do like about this, though, is that we might finally see AMD begin to improve in the memory department; an area that they've been quite weak in for years.
As we move through the list, we can see that the FX-8150 has the leg up in the core department and the L2 Cache one, while L3 Cache is the same between both. On the clock front the FX-8150 has an extra 200MHz at its base frequency, but 400MHz at the turbo frequency. The problem is that we all know that raw MHz isn't the be all and end all of a good CPU.
On the instruction front we can see that AMD offer us FMA4 and XOP which Intel don't. The question is, are we really going to see software that can make use of this? CrossFireX support is a real stand out with x16 / x16 support offered on the 990FX series chipsets. This compares to the Z68 which offers only x8 / x8 when it comes to multi GPU setups. While x16 / x16 boards are on offer for the Z68 platform, they bring with it the expensive NF200 based chip which does add quite a chunk of change to your motherboard.
Finally, the crème da la crème; the fact that the AMD FX line-up is unlocked means a lot for the simple reason Intel only offer two unlocked processors, both of which carry quite a price increase which we covered in a large amount of detail on the previous page. The problem here, though, is that the 2600k which is ultimately being compared here is unlocked and those unlocked speeds venture into the 5GHz+ territory. What matters, though, is when it all comes down to it, what's going to come out ahead?
Well, before we move onto the performance side of things, we can just have a quick look at what's going on with the package with the 8-Core models offering us a "Black" setup while the 6-Core model offers us a "White" setup. If you look closely, at the bottom we can see that both have got "Black Edition" labeling which is something the whole FX line-up should offer.
What's really funny, though, is that the box in the official PDF press kit shows an Intel CPU on the front with the "FX" labeling over the top. Fortunately, the package on the actual 8-Core CPU has an AMD CPU present.
You can tell it's an Intel CPU because of the layered IHS that's present and the notches at the bottom for the 1155 platform. I have no idea why AMD would've chosen to do it when there seems to be no way to distinguish the FX CPUs apart when there's no numbering present. Whatever, though, it gave me some "LOLs".
Test System Setup and Overclocking
I found myself wondering what path I was going to take when it came down to comparing the new FX-8150 today. I haven't been in charge of the CPU category for long, so I don't have piles of CPUs like I do video cards. The biggest thing I was trying to think about was whether or not we add the Intel Core i7 990X into the mix.
In the end I decided to not worry about it. The simple fact is that Bulldozer doesn't go against the X58 platform, let alone the highest end CPU in it. Instead, we opted for the two most important models to the FX-8150.
The first is the Phenom II X6 1100T. If you're the user of an AMD based system the chances are you're using something like an 1100T, or maybe the model below it. Either way, if you want to move to the FX-8150 on hopefully your already equipped AM3+ motherboard, then it's going to be one of the most important comparisons.
The other CPU we've got today is of course the 2600k.While more expensive than the FX-8150, the simple fact is that it's Intel's mainstream flagship processor which is what the FX-8150 is for AMD. We could've thrown in the 2500k, but it's cheaper than the FX-8150 and we already know how the 2500k and 2600k compare. Going through previous coverage, you will have an excellent idea of what the 2600k offers over it, and for the most part it's very little.
What we really wanted to make sure we offered, though, was overclocking of the CPUs we've tested, and that's the main reason we've cut back on the CPUs we've tested. While we only have three CPUs here today, we've got results for six setups and with overclocking being something that AMD are pushing, that was really important to us.
On the i7 2600k front we've got that benched at stock and overclocked to 5.2GHz. This isn't the highest we've had this CPU at; it's topped out at about 5.38GHz. 5.2GHz is what it runs at on our VGA testbed whenever we do any testing, though, and we know it's 100% stable. For that reason it's the number we decided to concentrate on; that and the fact that 5.2GHz for end users is more realistic than 5.38GHz.
Moving onto our next CPU, the Phenom II X6 1100T we tested at stock and overclocked to 4.06GHz. This is the same as what our Crosshair V Formula achieved when we first tested the board. We had hoped that the most recent BIOS would push that clock up a little higher, but all the BIOS updates that have come for the board over the last week or two have concentrated only on Bulldozer.
Finally, we have our FX-8150. First things first, what's overclocking like? Really easy actually, because we're dealing with the AM3+ platform that we've used for a while and we know where everything is when it comes down to the BIOS - overclocking the product is actually pretty easy.
Overclocking is something that AMD are promoting and with our Corsair H100 water cooler we managed to achieve an overclock of 4.765GHz as you can see above. This is a really nice overclock, especially compared to the 1100T which sees just over 4GHz.
Before we get into the performance side of things, I just want to touch on what AMD say about overclocking in the reviewers guide. While they've got stupid irrelevant numbers like 8GHz and 8.4GHz for Liquid Nitrogen and Liquid Helium cooling, when it comes down to normal cooling they say we should see 4.6GHz to 5GHz depending on the amount of cores for AIR and 4.9GHz to 5.2GHz for water.
We could get into Windows around the 4.9GHz to 5GHz mark with our H100. The one program that would crash out a lot was actually Cinebench. We could get Hyper PI to finish a run here and there; we also had some joys in a couple of other benchmarks. On a whole, though, we couldn't run everything solid, and for that reason we ended up with the 4.765GHz clock we mentioned above.
Let's get started!
PCMark 7 and HyperPi
Version and / or Patch Used: 1.04
Developer Homepage: http://www.pcmark.com
Product Homepage: http://www.pcmark.com
PCMark 7 includes a range of tests that give different views of your system's performance. In the Advanced Edition you can choose which tests to run. The common use and hardware component tests are unavailable in the Basic Edition.
Overall system performance is measured by the PCMark test. This is the only test that returns an official PCMark score. The Lightweight test measures the system capabilities of entry-level systems and mobility platforms unable to run the PCMark test, but it does not generate a PCMark score. Common use performance is measured by the scenario tests - Entertainment, Creativity and Production - each of which results in a scenario score. Hardware component performance is measured by the hardware tests - Computation and Storage - each of which results in a hardware score.
We can see under our first test, PCMark 7 that the performance of both the FX-8150 and 1100T are very close to each other with little separating them. When we overclock you can see we don't see much difference in the "Lightweight" score. Looking at the overall PCMark score, though, we do see a nice boost in performance with the 1100T coming in at 4424 and the FX-8150 coming in at 4708.
Unfortunately for AMD, though, while we see some light at the end of the tunnel when compared to the 1100T in this instance, you can see that the Intel i7 2600k sits well and truly ahead.
Version and / or Patch Used: 0.99
Developer Homepage: www.virgilioborges.com.br
Product Homepage: www.virgilioborges.com.br
HyperPi is a front end for SuperPi that allows for multiple concurrent instances of SuperPi to be run on each core recognized by the system. It is very dependent on CPU to memory to HDD speed. The faster these components, the faster it is able to figure out the number Pi to the selected length.
For our testing we use the 32M run. This means that each of the four physical and four logical cores for the i7 and the four physical cores of the i5 is trying to calculate the number Pi out to 32 million decimal places. Each "run" is a comparative to ensure accuracy and any stability or performance issues in the loop mentioned above will cause errors in calculation.
HyperPi performance doesn't look good. Compared to the 1100T it lags behind and overclocked, even with 700MHz on offer the performance is still slightly lower. Considering we're supposed to have improved memory performance and HyperPi can benefit a fair bit from that extra performance, it's disappointing to see these results.
No doubt we're off to a bit of a rocky start with Bulldozer; let's see how we go as we move into some other benchmarks to see what's going on.
Version and / or Patch Used: 1.00.1035BETA
Developer Homepage: http://www.aida64.com
Product Homepage: http://www.aida64.com
Replacing Everest in our labs is AIDA64. This new testing suite is from the core development team from Lavalys and continues that tradition. The guys have thrown in better support for multithreaded CPUs as well as full 64 bit support. We use this to test memory and HDDs for now, but may find ourselves opening this up to other areas of the motherboard.
Getting into AIDA64, we can start to look at some numbers that concentrate on the CPU. You can see when it comes to Read performance, the L1 Cache performance is very strong when compared to even the 2600k. You can also see it benefits a lot from the overclock. L2 Cache performance is strong against the 1100T, but when overclocked it went a bit weird. L3 Cache on the other hand while looking significantly stronger than the 1100T, does fall behind the i7 2600k by a fair bit.
Write performance is another story; we see strong gains in the L3 department and L2 looks ok against the 1100T, but L1 performance is horrible. There seems to be some form of optimization issue here that is bringing the score down. It shouldn't be 1/2 of what the 1100T is offering.
Looking at Memory performance, we can see that we've got some serious improvements over the 1100T even though our memory is running at 1600MHz on all platforms. We can see while it's not quite as strong as the 2600k still, there's been some major improvements made when compared to the last generation AMD CPU. Unfortunately, when overclocked, we don't see the same kind of boost in performance on the FX-8150 like we do on the i7 2600k.
Version and / or Patch Used: 7.0.1021
Developer Homepage: http://www.passmark.com/products/pt.htm
Product Homepage: http://www.passmark.com/products/pt.htm
Fast, easy to use, PC speed testing and benchmarking. PassMark PerformanceTest allows you to objectively benchmark a PC using a variety of different speed tests and compare the results to other computers. Twenty-eight standard benchmark tests are available in six test suites plus there are seven advanced testing windows for custom benchmarking.
We can see again that memory performance is strong compared to the 1100T with some good gains being seen. The FX-8150 continues to sit behind the i7 2600k. More importantly, though, it's probably disappointing to see that it's not scaling as strong. We saw the same kind of thing under AIDA64.
Looking at the CPU performance, we can see a good boost in certain areas with the best gains being seen in the Integer and Floating Point tests. We also see some strong scaling in these areas when compared to the 1100T. Compared to the 2600k, we can see that the Floating Point performance is exceptional. Even in the Integer we see the performance is very strong for the FX-8150. Overall in our CPU score, though, we see that the 2600k comes out ahead and scales extremely well.
CINEBENCH, Adobe Lightroom and MediaEspresso
Version and / or Patch Used: R11.5
Developer Homepage: http://www.maxon.net/
Product Homepage: http://www.maxon.net/downloads/cinebench/cinebench-115.html
CINEBENCH is a real-world cross platform test suite that evaluates your computer's performance capabilities. CINEBENCH is based on MAXON's award-winning animation software CINEMA 4D, which is used extensively by studios and production houses worldwide for 3D content creation. MAXON software has been used in blockbuster movies such as Spider-Man, Star Wars, The Chronicles of Narnia and many more.
CINEBENCH is the perfect tool to compare CPU and graphics performance across various systems and platforms (Windows and Mac OS X). And best of all: It's completely free.
Getting into CINEBENCH, we can see that against the older 1100T we do see a nice performance increase. Overclocked, though, we see a really strong boost in performance as we break that 8 point mark. The problem is, we continue to see the 2600k, which isn't much more expensive, dominate.
Version and / or Patch Used: v3.4
Developer Homepage: http://www.adobe.com/
Product Homepage: http://www.adobe.com/products/photoshoplightroom/
Adobe Photoshop Lightroom 3 software helps you bring out the best in your photographs, whether you're perfecting one image, searching for ten, processing hundreds, or organizing thousands.
Create incredible images that move your audience. Experiment fearlessly with state-of-the-art nondestructive editing tools. Easily manage all your images. And showcase your work in elegant print layouts, slide shows, and web galleries, as well as on popular photo-sharing sites. All from within one fast, intuitive application.
We can see some strong performance again. We've got an improved time on the 1100T and the overclock helps push the FX-8150 to a point where it's a good few seconds faster. We continue to see the same issue, though. The i7 2600k at stock is not only strong, but the biggest problem is the scaling for the FX-8150 just isn't as present as it is for the i7 2600k.
Version and / or Patch Used: v6.5
Developer Homepage: http://www.cyberlink.com/
MediaEspresso is a blazingly fast media universal converter that can transcode your videos, photos and music files and out put them to a huge range of portable devices including mobile phones, portable media players and even game consoles. With technologies like Smart Detect, Direct Sync and CyberLink's TrueTheater™ video enhancements, you can not only forget about complicated format, resolution and output settings, but your converted file will come out the other side looking better than when it went in!
MediaEspresso we start to really see some strong performance; we get to shave a good couple of minutes off the 1100T and we start to come with in distance of the 2600k. The same problem creeps up again and again, though. We see that scaling performance is just so strong on the 2600k and while overclocked the FX-8150 performs extremely well when overclocked here, it's still struggling against what is ultimately its main competition.
3DMark 11 and Aliens vs. Predator
Version and / or Patch Used: 1.0
Developer Homepage: http://www.futuremark.com
Product Homepage: http://www.3dmark.com/3dmark11/
3DMark 11 is the latest version of the world's most popular benchmark. Designed to measure your PC's gaming performance 3DMark 11 makes extensive use of all the new features in DirectX 11 including tessellation, compute shaders and multi-threading. Trusted by gamers worldwide to give accurate and unbiased results, 3DMark 11 is the best way to consistently and reliably test DirectX 11 under game-like loads.
Compared to the 1100T, we can again see a nice little performance boost and we start to see the gap between the 2600k and FX-8150 close. Overclocking helps push that Performance score up even higher, but again, it just lags behind that 2600k, and while it's more expensive, it's still the biggest competitor for the FX-8150.
Aliens vs. Predator
Version and / or Patch Used: Standalone Benchmark
Timedemo or Level Used: Built in Benchmark
Developer Homepage: http://www.rebellion.co.uk/
Product Homepage: http://www.sega.com/games/aliens-vs-predator/
Aliens vs. Predator is a science fiction first-person shooter video game, developed by Rebellion Developments, the team behind the 1999 original PC game, and published by Sega for Microsoft Windows, the PlayStation 3 and the Xbox 360. The game is based on the Alien vs. Predator franchise, a combination of the characters and creatures of the Alien franchise and the Predator franchise. There are three campaigns in the game, one for each race/faction (the Predators, the Aliens and the Colonial Marines), that, while separate in terms of individual plot and gameplay, form one overarching storyline.
Following the storyline of the campaign modes comes the multiplayer aspect of the game. In this Multiplayer section of the game, players face off in various different gametypes in various different ways.
The thing is, we move away from a synthetic test like 3DMark 11 and get into an actual gaming one and across the board we can see we've got nothing more than a bit of fluctuation that translates to a single FPS. Gaming performance is something we'll be examining in more detail.
Power & Temperature Tests
Power Draw Tests
Looking at power draw, we can see the new FX-8150 does a good job at idle coming in lower than both the 1100T and the i7 2600k. You can see when we look at load, though, it sits higher than our other processors. Even though it's not overclocked as high either, you can see that load at 4.76GHz is quite high.
Looking at the idle temps on the FX-8150, we'll choose to ignore them at the moment. Instead we'll just concentrate on the load which is pretty good. Overclocked numbers are also very good. To be honest, though, we don't want to put too much emphasis on the AMD temperature numbers for the simple fact the idle numbers are just crazy low. Load numbers look to be correct, but we'll just leave it at that.
Damn! You can't help but almost feel disappointed with what's going on here today and while I can completely understand that the 2600k is more expensive, like we've already said here a number of times, you've ultimately got to compare the flagship product from Intel against the flagship product from AMD, and that's ultimately what we've done.
I think the problem isn't so much just the performance against the Intel i7 2600k, but also against the Phenom II X6 1100T. Sure, it's faster, for the most part; at times we see it dip a little below, but for the most part it's faster. The problem is, it should be faster every time!
The worst thing is, I don't think it's that the FX-8150 is a poor performing CPU, it's just that nothing is really able to make use of it and that's a problem for AMD. We do see in certain situations it shines, especially against the 1100T, but we never find ourselves going "WOW".
What does look positive for AMD is the gaming performance at the moment. We'll have to dive into that more soon, though. If performance in games is similar between the 2600k and FX-8150, then we could be onto something.
It's all going to come down to price, that's what's going to separate what people buy. For starters, if you're on an X4 or an X6 and you have an AM3+ compatible board, then you can get an upgrade really cheap. You can get into a top of the line FX processor for $245 US. That's so much cheaper than what a move to a Z68 platform will cost you for the simple fact all you need to do is buy a new CPU.
Then there's someone on a P55 board, or an older AMD system that doesn't support the new processor. In the CPU department you're looking at $245 US for the FX-8150 and $314.99 US for the i7 2600k. That's a good chunk of change, but the chances are if you're looking at an FX-8150, the other option you're looking at is the 2600k because they're both the Flagship product for both companies.
Then you move onto motherboard pricing. We could use the Crosshair V Formula as an example, but the problem is that there's no Maximus IV Formula-Z and the Extreme-Z is significantly more expensive. Instead, let's use GIGABYTE as an example. GIGABYTEs GA-Z68X-UD7-B3 is going to cost you $349.99 US. A lot of that is due to the fact the board supports x16 / x16 via the NF200 chip.
The GIGABYTE GA-990FXA-UD7 is going to set you back $249.99 US. With $100 US saved on the motherboard and $70 US saved on the CPU, you've got yourself a saving of $170 US. The difference between a GTX 570 and GTX 580 is $115 US. The difference between a GTX 560 TI and GTX 580 is $200 US and this gives us a really good idea of what we're able to do.
A FX-8150, GA-990FXA-UD7 and GTX 580 will set you back $894.98 US. A 2600k, GA-Z68X-UD7-B3 and GTX 560 Ti will set you back $864.97 US. Both are the flagship CPU from AMD and Intel, both are the flagship motherboard from GIGABYTE and while that 2600k system might win out in AIDA64 and HyperPi, when it comes down to gaming, it's not going to take an expert to figure out which one is going to come out ahead.
If you went with the GTX 580 in our 2600k system, the price would come in at $1064.97 US. Are you going to get an extra $169 US worth of performance out of the Intel one? Well, at the moment we can't fully confirm because we really need to give the 2600k and FX-8150 a head to head. It's for that reason we'll wrap up our final thoughts on not just the FX-8150, but the platform on a whole. So keep an eye out for that tomorrow.
One side of me is let down with a lot of the numbers we see here today; the other side of me sees the potential of what a system is able to be for $900 with a FX-8150 when compared to one with a 2600k. Let's give a full run down of what you're able to see when both systems go through our VGA testing gauntlet and really wrap everything up.
PRICING: You can find products similar to this one for sale below.
United States: Find other tech and computer products like this over at Amazon.com
United Kingdom: Find other tech and computer products like this over at Amazon.co.uk
Australia: Find other tech and computer products like this over at Amazon.com.au
Canada: Find other tech and computer products like this over at Amazon.ca
Deutschland: Finde andere Technik- und Computerprodukte wie dieses auf Amazon.de