AMD’s FX-8150 vs. Core i7 & Phenom II – Bulldozer Arrives!

The new “Bulldozer” FX-series CPUs released today from AMD, and ABT has been evaluating the “Zambezi” 8-core flagship FX-8150 for the past 5 days. We naturally want to know if it is a good upgrade from AMD’s aging Phenom II architecture and we will be comparing its performance to our overclocked 980 Black Edition also on the AM3+ platform and also to our overclocked Intel Core i7-920 LGA-1366 platform.

We will briefly introduce the new FX Bulldozer architecture and then head right into our evaluation which will highlight performance in 20 modern games at 1920×1200 and 2560×1600. We will also run synthetic and real world benchmarks on our stock-clocked and overclocked FX-8150.  We will work to establish its performance in games when paired with a single HD 6970 and also with HD 6970 CrossFire-X2. In this way, we will see if the FX-8150 is worth the $245 that AMD is asking for it.

We notice that FX 8150 is positioned against Intel’s enthusiast CPU line-up and directly against the upper-midrange Core i5 2500K. Considering that AMD is introducing their new flagship processor at $245, we can expect that its performance should be between the 2500K and 2600K.

Here is how AMD visualizes their line-up.  Although they say it is not indicative of performance positioning, it generally is.  You may also notice that AMD does not include Intel’s highest end LGA 1366 platform and that Intel will again lock down the high end when Sandy Bridge-E launches later on this month.

AMD is releasing seven CPUs simultaneously and they are available globally later on today in quantity – from the lowest-priced 4-core FX 4100 for $115, to the the 6-core FX 6100 for $165; as well as the flagship 8-core FX 8150 for $245 and the FX 8120 which is also 8-core for $205 as shown in the table below:

Below is the Bulldozer family without the pricing but with more details of their differences. As you can see, the L3 cache remains the same as well as 1866MHz DDR support and all FX processors will use the AM3+ motherboard platform.  It is likely that older AM3 motherboards will support these new FX CPUs with a BIOS update, so it is an upgrade path for older Phenom IIs.

The Bulldozer Architecture

AMD has been working on new CPU architecture since long before Phenom II. Phenom II reached its performance peak in gaming with the 980 Black Edition (BE) which we reviewed here a few months ago.  It is AMD’s fastest 4-core processor and it sports a stock CPU speed of 3.7GHz. We were able to overclock ours to 4.3GHz and that is the speed that we shall test it at; very nearly at the current limit of the Phenom II architecture on air-cooling.

The stock speed of the FX-8150 is 3.6GHz; however, Turbo Core mode allows 4 cores to reach 3.9GHz on all 8-cores and 4.2GHz with 4-cores if there is sufficient thermal headroom. So basically we are putting up a 4.2GHz Bulldozer against a Phenom II clocked at 4.3GHz. And we also able to give our FX-8150 an overclocked boost of all 8 cores to 4.4GHz, to give you an idea of its scaling in games.

AMD’s Bulldozer versus Intel’s Sandy Bridge

AMD is positioning their top FX-8150 CPU against Intel’s upper midrange – specifically against the i5 2500K. Our test CPU is the older “Bloomfield” Core i7-920 which we can stably clock past 3.8GHz (with turbo on a single core to 4.0GHz). Overclocked, it should offer similar performance to the stock-clocked i5 2500K (3.4/3.8GHz) in gaming with a single powerful video card like HD 6970 or the GTX 580 at resolutions at or above 1080p.  Here is AMD’s chart which shows AMD’s FX on-paper advantages over Intel’s competing CPUs.

The FX-8150 looks great on paper.  The Phenom II’s tradition “weakness” against Intel’s Core2 processors have always been their single- and lightly-threaded performance.  And with the recent release of Intel’s Sandy Bridge i5 2500K and i5 2600K, the performance gap has widened in Intel’s favor.  We are going to be looking to see what AMD has done to close the gap with the new Bulldozer architecture.

Head over to the next page and we’ll find more out about Bulldozer architecture

–~~~~~~~~~~~~–

What is Bulldozer?

Phenom II has reached its end of its useful life.  There have been few changes to AMD’s basic architecture since about 2003.  Bulldozer is the next iteration of AMD CPUs that are particularly useful for servers and have also been adopted for the desktop as the basis for AMD’s next generation CPUs that will probably be in use for years to come.

Here is AMD’s now-public roadmap:

Now that we know that Bulldozer is expected to scale beyond 2014, lets look at the architecture.

Bulldozer’s design allows two cores to each share a larger, higher-performance function unit – as a floating point unit (FPU), for example – with less total die area than having separate, smaller function units for each core.

The floating point unit in “Bulldozer” has been re-designed from Phenom II. It has been improved to support new instructions and to allow resource sharing between its cores. There are two 128-bit FMACs shared per module which allows for two 128-bit instructions per core or one 256-bit instruction per dual-core module.

“Bulldozer” Front End

Bulldozer’s front-end unit drives the processing pipeline and is designed to make sure that all of the cores are kept busy with information. The front end works with each dual core unit to allocate threads to individual cores. AMD made significant changes that include decoupled predict and fetch pipelines, as well as prediction-directed instruction prefetchers.  It also means that Bulldozer’s pipeline is longer.

A Prediction Queue manages direct and indirect branches that are fed with a L1 and L2 Branch Target Buffer which stores destination addresses. Bulldozer’s modules can each decode up to 4 instructions per cycle versus 3 on AMD Phenom II processors.


The prediction pipeline produces a sequence of fetch addresses. The Fetch pipeline does a look up in the instruction cache, and pulls 32 bytes per cycle into the fetch queue which feeds the decoders.

Bulldozer uses a physical register file (PRF) which is a single location that holds the register results of executed instructions. This reduces power by eliminating unnecessary data movement and data replication by keeping one copy instead of broadcasting the data.

Caches

Each Core is equipped with a 16 KB Level 1 Data cache, a 32-entry fully associative DATA TLB, and a fully out of order load/store – capable of two 128-bit loads per cycle or one 128-bit store per cycle. Each dual-core module includes a 2 MB 16-way unified L2 cache with an L2 TLB capable of 124 entry, 8 way that services both instruction and data requests.  Bulldozer supports up to 23 outstanding L2 cache misses for memory system concurrency.


Finally AMD has designed a shared 8 MB L3 cache with 64 way associativity for both cores in a each module.  And Bulldozer is planned for use in both the desktop and server areas.

AMD also redesigned the Northbridge.

Turbo Core

One impressive new feature that appears to work well in practice is AMD’s Turbo Core technology.  It is far more sophisticated than similar technology used in their 6-core Phenom IIs.  It allows for the FX-8150 to increase from 3.6GHz to 3.9GHz across all eight cores, and a further double-boost on 4 cores to 4.2GHz if TDP headroom allows it.  Generally the boost effect can range from instantaneous to over a longer period, depending on the thermal situation.

Of course, this could give some significant performance improvements as AMD’s internal testing shows.

And of course we need to look at the AM3+ chipset.

It all comes together looking like this at launch:

The Results of the changes

According to AMD, FX processors offer:

  •   the World’s First 8-Core Desktop Processor designed for highly threaded applications,
  •   new instruction support for FMA4, XOP, AES, AVX, and SSE 4.2 for next generation PC applications
  •   unlocked to allow customer customization and higher system performance

Specifications:

FX Processors:

We just had a very brief overview of Bulldozer’s architecture.  As with all CPU designs, there are compromises that make Bulldozer particularly attractive for AMD’s large server market but not so perfect for desktops, perhaps.  We we need to look at is how the Bulldozer Flagship FX-8150 performs in applications and especially in games.  Please read on.

–~~~~~~~~~~~~–

Bulldozer FX vs. Phenom II vs. Intel Core i7 in gaming

This evaluation is focused on the FX-8150 in gaming. And we are going to set up a “David vs. Goliath” scenario using the more expensive Core i7 LGA 1366 platform. You can expect to pay more for a tri-channel Intel system that we are comparing as the current i7-960 runs at 3.2GHz and sells for about $300.  Of course, a savvy gamer will more likely choose the higher-performing-in-gaming Sandy Beach i5 2500K or 2600K and overclock them over 4.0GHz.  In this case, our 1366 LGA i7-920 overclocked to 3.8GHz with Turbo Boost will perform similarly to a stock i5-2500 in gaming.

The review that we are presenting today evaluates the Bulldozer FX-8150 versus the overclocked Phenom II 980 Black Edition versus the overclocked Core i7 920. We will use the HD 6970 and HD 6970 CrossFire as our tests card because we are looking at a relatively high-end gaming system – something an enthusiast would likely pair with a Core i7 or i5 processor..

As in previous similar testing we will prove a point in gaming – that the video card makes the most difference as long as it is paired with a capable CPU. If you look at just about every other tech site, they compare the Intel CPUs to the Phenom IIs using low to medium details and at low to medium resolutions with 3 or 4 games – something no gamer can relate to. Instead, let’s give you practical gaming results at resolutions you play at – beginning with 1920×1200 resolution which 1080p gamers can also completely relate to. And let’s use twenty (20) modern games to give you a good feel for what to expect from our 3 test CPU platforms.

We need to know if the new Bulldozer can give similar performance to a high-end CPU platform with fast graphics and at high resolutions. Using our HD 6970 and HD 6970 CrossFire should be similar performance-wise to using either a GTX 570 or GTX 570 SLI; something an enthusiast might well pair with either Core i7 or a Phenom II processor. We want to see if the AMD FX platform suffers practically in gaming when compared to Intel’s solution and how well it does compared to the overclocked Phenom II.

We had very little time to spend testing our CPUs and HD 6970s.  In fact, we had such poor scaling with CrossFire on the AM3+ platform, we suspect driver issues.  In our follow-up Bulldozer article, we will also set up Nvidia video cards and test our FX-8150 with even more powerful graphics – GTX 580, GTX 580 SLI and HD 6970-X3 TriFire.

Overclocking

Our FX-8150 overclocked easily to 4.4Ghz by upping the stock multiplier of 18 to 22 (calculate using the multiplier times the bus frequency of 200MHz to get the CPU core’s clockspeed). We added a little voltage to the core to make it completely stable at 1.4V and the  temperatures remained nearly as cool as at the stock settings.

Beyond 4.4GHz we were stopped.  We could not get 4.5GHz stable no matter how much voltage we dared to use (up to 1.55V).  It may be an issue of using our old Thermalright UltraExtreme120, a good CPU cooler in its day.  It still manages to get our Phenom II 980 BE from 3.7GHz to 4.3GHz, so it is at least still a decent air-cooler.  However, for our follow-up article, we will use watercooling and a full tower case instead of the mid tower we currently use.

We used the BIOS to adjust our CPU overclock.  The ASUS AM3+ CrossHair V motherboard in incredibly full-featured and an absolute joy for the overclocker to use – it is that intuitive, easy-to-use and functional!

[insert MB image]

However, there is another overclocking option within Windows that is available in the (formerly-named) Catalyst Control Center  (now called AMD VISION Engine Control Center) that allows you to adjust the CPU core speed also:

Naturally we settled on the “free” +600MHz over stock (+200MHz over the extended 4-core Turbo Core mode) and we ran our benches at 4.4GHz as well as at the stock speeds to give you an idea of the framerate increase from increasing the core speed. Just remember that most games will not use more than 4 cores and you will only get +200 more MHz over the automatic turbo mode – even if the boost is for an instant.

To achieve 4400MHz, we also disabled Turbo Core and power management since it would not clock higher manually.
Here is CPU-Z information about the FX-8150 at  idle and we can see power savings at work.

AMD sent us a very nice watercooling unit that will be available with the FX-8150 as a bundle that may cost around $100 more. Here is a picture of it.

[insert pic]

Since we received our watercooling unit from AMD yesterday and since we have not yet unboxed our new Thermaltake Tower case, we will naturally do a follow-up article on the FX-8150 that will feature further overclocking.  There is a lot more to be explored with the new processor that we have only touched on in this introduction.

In fact, this week we are planning to build our new AMD platform in a brand new Thermaltake Chaser MK II full-sized Tower case that is set up for watercooling – expect an evaluation of this beautiful case this month (plus a Thermaltake mechanical gaming keyboard evaluation!) as well as Part Two of the Bulldozer FX-8150 evaluation which will include Part 3 of SLI vs. CrossFire.

Power Draw

We kept an eye on power and the FX-8150 kept within its specified TDP of 125W – that is, until we overclocked it and then readings went way up.  We will cover this thoroughly in Part Two when we get watercooling set up.

Please continue on to the next page for the complete hardware and software setup of our platforms – AMD’s FZ-8150 versus Phenom II 980 BE versus Intel’s Core i7-920. We shall see what happens with high performance again in PC gaming with fast graphics in the Autumn of 2011.

–~~~~~~~~~~~~–

Test Configuration

Test Configuration – Hardware

  • Intel Core i7 920-reference 2.66GHz and overclocked to 3.8GHz; 21x multiplier for 3.97 GHz, Turbo Boost is on.
  • AMD FX-8150 stock clocked at 3.6GHz with Turbo Core enabled and also overclocked to 4.4GHz with Turbo Core disabled.
  • Gigabyte EX58-UD3R (Intel X58 chipset, latest BIOS, PCIe 2.0 specification; CrossFire/SLI 16x+16x).
  • ASUS CrossHair V AM3+ RoG motherboard (PCIe 3.0 Specifications; CrossFire/SLI 16x+16x)
  • 6 GB DDR3 PC1800 Kingston RAM (3×2 GB, at PC1600 speeds for Intel’s tri-channel; 2×2 GB supplied by Kingston)
  • 4 GB DDR3 PC1600 Kingston RAM (2×2 GB, at PC1600 speeds; RAM supplied by Kingston)
  • ATI Radeon HD 6970 (2GB, reference clocks, 880/1375MHz) supplied by AMD
  • ATI Radeon HD 6950 (2GB, flashed to stock HD 6970; 880/1375MHz) supplied by AMD
  • Two identical 500 GB Seagate Barracuda 7200.12 hard drives configured and set up identically from drive image; one partition for Intel drivers and one for AMD drivers
  • 2 x Thermaltake ToughPower 775 W power supply units supplied by Thermaltake
  • Thermaltake Element G Case supplied by Thermaltake for the Intel platform
  • Coolermaster Elite 430 Case, supplied by Coolermaster for the AMD platform
  • Noctua NH-U12P SE2 CPU cooler (for Intel CPU), supplied by Noctua
  • Thermalright UltraExtreme120 CPU cooler (for AMD platform)
  • Philips DVD SATA writer
  • HP LP3065 2560×1600 thirty inch LCD.

Test Configuration – Software

  • ATi Catalyst 11.9 WHQL driver for all Radeons for games and Preview 11.10 for all other tests; highest quality mip-mapping set in the driver; surface performance optimizations are off; “use applications settings” are checked
  • Windows 7 64-bit; very latest updates
  • DirectX July/November 2010
  • All games are patched to their latest versions.
  • vsync is forced off in the control panel.
  • Varying AA enabled as noted in games; all in-game settings are specified with 16xAF always applied if possible; 16xAF forced in control panel for Crysis.
  • All results show average rates as noted.
  • Highest quality sound (stereo) used in all games.
  • Windows 7 64, all DX9 titles were run under DX9 render paths, DX10 titles were run under DX10 render paths and DX11 titles under DX11 render paths.

The Benchmarks

The Synthetic and Real World Tests

The Game benchmarks

  • Vantage
  • 3DMark11
  • Batman: Arkham Asylum
  • Serious Sam, Second Encounter HD (2010)
  • Wolfenstein
  • Left 4 Dead
  • Mafia II
  • Crysis
  • World in Conflict
  • Far Cry 2
  • Just Cause 2
  • Resident Evil 5
  • Alien vs. Predator
  • Battleforge
  • STALKER, Call of Pripyat
  • F1 2010
  • Metro 2033
  • Lost Planet 2
  • H.A.W.X. 2
  • Civilization 5
  • Dirt 3
  • Crysis 2
  • Heaven 2

OtherBenchmarks

We have got an interesting project going. First we look at synthetic and real world benches before we head for gaming results.

–~~~~~~~~~~~~–

Synthetic Tests

HD Tune 2.55

HD Tune is a hard disk utility but it may give us some differences between the AMD AM3+ platform and Intel’s X58 motherboards. We are using the default setting of 64KB blocks for testing. First we run the standard benchmark with the Core i7-920 and the x58 MB.

Now we run the benchmark for the Phenom II and AM3+ motherboard; the minimum speeds are way down; something choked during the test.Now we look at the FX-8150 in the same AM3+ motherboard as the Phenom II.

There is nothing conclusive here.  It will be interesting to rerun the HDD tests with a SSD.

ATTO

The ATTO Disk Benchmark is an aging performance measurement tool which measures storage systems performance with various transfer sizes and test lengths for reads and writes. Several options are available to customize the performance measurement including queue depth, overlapped I/O and even a comparison mode with the option to run continuously. Here are the results of the Core i7-920/X58 platform benchmark.

Next the same benchmark in the same PC and identical conditions with the Phenom II/AM3+ motherboard.

It doesn’t really prove anything, but synthetic benchmarks are a little faster on the AM3+ platform with Phenom II. Let’s see if it holds for the FX 8150:

Well, we do see a higher transfer rate with the AM3+ motherboard even though the hard disks are identical 500GB Seagate drives.  Again, it will be interesting to see results between X58 and AM3+ using a Solid State Drive (SSD).

Super Pi

We didn’t run Super Pi on Core i7 so here it is on the Phenom II.

And now the FX-8150

Phenom II is stronger in this bench.  We will see this over and over.  Unless an application is heavily-threaded or optimized, it will run faster on the Phenom II (nevermind the i5 or i7 which is faster than the Phenom II).  We notice a trend of lower IPC with FX that AMD is trying to make up with more cores and increased clockspeed.  Did they succeed?  Continue on to Fritz Chess Bench.

Fritz Chess Bench

Fritz Chess Benchmark is found within the game’s program files and basically it crunches numbers to test your processor’s speed.  Deep Fritz takes advantage of massive calculations and multi-threaded performance to work any CPU fully. It loads all threads 100% and will drive your CPU temperatures way up.

Besides showing relative speed when compared to a P3 1.0GHz CPU, it also shows the nodes completed.  The faster your CPU, the more nodes completed.  First up is the i7

Next the Phenom II

Finally the FX-8150

The FX-8150 is faster than the Phenom II but still lags behind the Core i7.

CustomPC Benchmark

CustomPC benchmark use widely available open-source applications to carry out the tasks that most of us perform on a regular basis. There are three tests, each of which measure different aspects of a PC’s performance. These tests themselves are not synthetic benchmarks but instead they use real world image, video and multi-tasking tasks to test the performance of your computer.

The tests are:

  • GIMP Image Editing
  • H.264 Video Encoding
  • Multi-tasking

As usual we test with the Core i7-920 first.

Now the same test is repeated on the Phenom II 980 BE and it is much slower.

 

Finally we test the FX-8150.

The Phenom II partly by virtue of being +100MHz faster, scores slightly better than the FX-8150. It cannot be denied that the new Bulldozer architecture is a bit weaker for single-threaded applications than even the Phenom II architecture is.  Clearly the Core i7-920 at 3.8GHz is easily much faster in image editing and video encoding where the applications are not heavily multi-threaded. Even the multi-tasking test is solidly in favor of the Intel architecture.

CINEBENCH

CINEBENCH is based on MAXON’s professional 3D content creation suite, CINEMA 4D. This latest 11.5 version of CINEBENCH can test up to 64 processor threads accurately and automatically.

First up, the Core i7 with a rather shortened CPU test.  (We will update this)

Next up the Phenom II which got a lot more detailed testing. The i7 is significantly faster in the CPU test than the Phenom II.

And now the FX-8150

In some ways, FX architecture has improved over the Phenom II architecture except for single core IPC, but it still lags behind Intel’s older Core i7/X58 platform.  Again we see FX improvements for heavily-threaded applications but a decline in single-threaded performance from Phenom II.

X264

We ran this bench quite late in out testing and only managed to see if the FX 8150 had bettered the overclocked Phenom II 980 BE.  Basically this test encodes a HD video clip into a x264 video file. The first pass is very quick and the second one is much slower and much more demanding of a task as it does the actual encoding.  This benchmark is heavily mult-threaded and should favor the new architecture.

Faster is better and first up is the Phenom II 980 BE.

Next up,  the FX-8150.

As we can see, although the first pass is quicker on the Phenom II, it is far more important to have the second pass encode quickly as it is far more time-consuming.  The first pass simply calculates and doesn’t even use 4 cores whereas the second pass will use all eight for the actual encoding.  Finally a solid win for FX over Phenom II in a heavily-multi-threaded task.

–~~~~~~~~~~~~–

The Benchmark Suites – PC Mark and SANDRA

PCMark Vantage
PCMark Vantage is a PC benchmark suite designed for Windows Vista and 7 that offers one-click simplicity for casual users and detailed, professional grade testing for industry, press and enthusiasts. A PCMark Vantage score is a rough measure of a computer’s performance across a variety of common tasks. We have used this bench for 3 three years and will soon be upgrading to PC Mark’s latest suite, PCMark 7.  As an interesting note, AMD expects Windows 8 to be far more optimized in scheduling for Bulldozer architecture than Windows 7 currently is.

The HDD Test suite in PCMark Vantage consists of the following 8 tests:

  • HDD 1- Windows Defender
  • HDD 2- Game HDD
  • HDD 3- Importing pictures
  • HDD 4- Windows Vista start-up
  • HDD 5- Video editing
  • HDD 6- Media Center
  • HDD 7- Adding music to Windows Media Player
  • HDD 8- Application loading

First we run the standard PCMark Vantage benchmark suite using our Core i7-920 and note that it scores the lowest overall.

Now we run the same benchmark suite with our Phenom II 980 Black edition and we see right away that it scores 500 points higher.

Finally we run the same benchmark suite with our FX-8150 and see that it is rated higher overall than either of its two competing platforms.

In all cases, the FX 8150 is faster with PCMark Vantage than with the other platforms.  You may look over the individual scores to see the strengths and weaknesses of each processor and platform.

[insert summary chart]

And now on to SANDRA, another comprehensive benching suite.

SANDRA 2011 SP5

SANDRA, short for the System ANalyser, Diagnostic and Reporting Assistant is an information and diagnostic utility. It provides a tremendous amount of information about every hardware and software component in your PC. It also provides various benchmarks including for measuring and reporting disk performance. SANDRA 2011 Service Pack 5 is already out and it has a lot of new functionality for the new APUs and the FX processors.

Aggregate Arithmetic Performance (GOPS)

First up is the Core i7 920 and we see the GOPS is made up of several scores followed by a very detailed analysis.

And the next charts give the detailed analysis of Core i7′s GOPs with a final score of 21.47kPT

 And now let’s’ check out the FX-8150

[insert summary chart]

Let’s head for synthetic game testing.

–~~~~~~~~~~~~–

Gaming Synthetic testing

Vantage

Let’s take a look at FX-8150 running Vantage at stock with a single stock HD 6970.  Since the Phenom II 980 BE runs on the same AM3+ motherboard that FX-8150 does, it makes for a much closer comparison where only the CPU is the variable.  We did not include detailed Core i7 results as we would be comparing platforms, not CPUs.

Now let’s rerun Vantage with our FX-8150 overclocked to 4.4GHz.

We next explore CrossFire-X2 performance with dual HD 6970s.  Please note that we encountered CrossFire scaling issues that we are inclined to believe is software or possibly driver-related.  AMD’s internal testing could not replicate our issues, so we will return to future testing with a clean install of Windows to explore CrossFire scaling with FX-8150 in our follow-up article which will include further overclocking of both the CPU and GPU.  In effect, our next article will also be part three of our SLI versus CrossFire series as we also want to see how Nvidia video cards perform with the new FX 8150!

And just for comparison, here is the older Phenom II architecture’s 980BE at 4.3GHz as scored by Vantage.  You see right away that the FX CPU score is higher in the first test and lower in the Physics test while the two graphics tests are faster with the FX.

Let’s look at Futuremarks’ latest graphics benching suite, 3DMark11.

3DMark 11

This is Futuremarks’ latest test and it makes more use of multi-core than Vantage does.

First up is the FX.

Next up is the same test with the 4.3GHz Phenom II 980 BE.

As you can see, there isn’t a lot of difference in the final “score”.  Let’s look at 20 modern game benchmarks to see if we can draw any conclusions.  Because we only had 4 days to evaluate our FX-8150, our usual detailed discussion of each game has been dropped in favor of performance summary charts covering multiple games.

–~~~~~~~~~~~~–

FX-8150 Performance in 20 games

Here are the summary charts of our 20 games and 3 synthetic tests.  In the first three columns – FX-8150, Core i7-920 and Phenom II 980 BE – the performance winner is in Bold; if there is a tie, all are in bold. CPU core speeds are shown with stock/turbo (if any) in GHz.

All results except for Vantage and 3DMark11 show average framerates and higher is always better.  In-game settings are fully maxed out and they are identical across all platforms except for Left 4 Dead where the Intel CPU uses a slightly different benchmark from the AMD CPUs (Steam updated the game during the testing).

DISCLAIMER: CrossFire Scaling is so bad on the AM3+ platform that we are inclined to suspect driver issues instead of the CPU hitting any “wall” as some other sites have suggested.  Our follow-up Part Two, will start again with a clean install of Windows and we will use even more powerful graphics – HD 6970-X3 TriFire (880/1375MHz) and GTX 580 SLI.

Summary Charts

Please note that we use HD 6970 and HD 6970 CrossFire (not HD 6870).  In the first 3 columns,”wins” are Bolded and if there is a tie, all are bolded.

As we can see, it requires overclocking the Phenom II 980 BE to approximately 4.3GHz to match the Intel i7-920 at 3.8GHz with turbo boost on.  And the FX-8150 at stock 3.6GHz (with turbo on to 4.2GHz on 4-cores) generally does well in most of the benches, winning the most of any platform for the single-GPU HD 6970.  And overclocking a further +200MHz, to 4.4GHz makes it the clear winner in most cases.  However, the picture changes dramatically with CrossFire scaling and the AM3+ platform.

Unfortunately, we do see some issues with drivers and CrossFire scaling with our FX-8150 (and to a lesser extent, out Phenom II) that originally had us looking to see if we had inserted one of the CrossFired HD 6970s into a 4x PCIe slot instead of into a 16x slot. However upon confirming that the slots are definitely 16x+16x PCIe and using both Catalyst 11-9 and 11-10 Preview drivers, we experienced similar issues with poor CrossFire scaling.  Since we had less than 5 days total to test our platform and write our article, we did not update the motherboard’s BIOS.

We also noted the exact same configuration when used with a Phenom II in place of the FX-8150 CPU gives slightly better CrossFire scaling in more than a few games and we have decided to reinstall Windows for our follow up article as AMD was unable to reproduce our CrossFire scaling issues.  Civilization 5, Call of Pripyat, Lost Planet 2, F1 2010, Far Cry 2, Left 4 Dead and Batman are the worst CrossFire scaling offenders; generally CrossFire scaling is excellent as evidenced by the Intel platform.

We also want to note that the Catalyst 11.10 preview drivers which we used for our non-gaming benchmarks has a new look and more functionality.

Well, that finishes up our testing of the FX-8150 in the less than a week that we were allowed to preview it.  Although we are going to follow up with a much more thorough testing of the new FX processor, we feel we can give you some conclusions.

–~~~~~~~~~~~~–

Conclusion

This conclusion is only a partial one as much more testing needs to be done – especially with CrossFire scaling in games.  It is clear to us that the Phenom II solution was nearing the end of it’s useful life as their quad-core flagship 980 BE overclocked to 4.3GHz was barely able to hold its own in high-resolution gaming against the three-year old Core i7-920 at 3.8GHz.

Even with a lower individual IPC than with Phenom II, it now appears that AMD is again competitive in the midrange as the new bulldozer architecture is forward-looking and something for them to build upon.  They really need to get their clockspeed much higher and their roadmap indicates that they need 10 to 15% more each year just to remain competitive with Intel.

AMD’s price of $245 seems a bit optimistic as an i5 2500K can be got for less than that, but market pricing will come into play shortly after this processor is fully stocked and in quantity for sale.  However, we expect that there are thousands of these processors available now for eager AMD fans to buy right away, and if you have a AM3+ motherboard and a slower Phenom II, you should consider the FX-8150.

We look forward to AMD’s new FX line and expect them to build upon it as the process matures and for them to reach far higher clockspeeds than they do today.  We also will bring you much more about the FX-8150 in gaming and give you a real conclusion within the next two weeks as we test with Nvidia’s GTX 580 SLI and AMD’s HD 6970-X3 Tri-Fire.  And of course, we will set up watercooling for our FX-8150 to see what the future Bulldozer performance may bring.

Make sure you follow us on our ABT forum.  You will get the very latest news about FX-8150 benching and you can help guide us in future testing.

Mark Poppin

ABT Senior Editor

Please join us in our Forums

Become a Fan on Facebook

Follow us on Twitter

For the latest updates from ABT, please join our RSS News Feed

Join our Distributed Computing teams

«»

Pages: 1 2 3 4 5 6 7 8 9

apoppin

Founder and Senior Editor of ABT.

  • Ocre

    Your review is a lot different than most of them out there. And that is a good thing. Its a different view comparing BD with the phenom 980 and the i7 920. Most other sites used the phenom x6 and intel’s Sandy Bridge. With the latter configurations its harder to see BD in a good light. Especially considering AMDs own 6 core phenoms which in, my opinion, currently a much better value. As a slightly revised model, I believe many ppl have overlooked the 1100t. it pulls away from then phenom x4 by a decent amount. Its a bar that bulldozer is struggling to surpass.

    But your review focus is interesting because it does show BD isnt that bad compared to the phenom2s (4 core versions). Its not so bad at all in this comparison. AMD can only improve on this design from here. It will get better. Its just a terrible way to start it off. only if that 1100t (and 1090) wasnt in the picture, bulldozer would be a much better looking AMD option!

    Thanks for your angle in reviewing the BD, it is something unique and useful. Its a review that i enjoyed mostly because it helps see a more complete picture.

  • http://None nomail

    The benchmark charts was a clusterfuck that I couldn’t understand. The gaming chart was much better. Need to create readable charts and organize the scores accordingly instead of copy and paste results at tiny pics which are hard to see, this review is a fail.

  • hansmuff

    I don’t understand why this review was posted in the shape it’s in.

    First off, apoppin, this is way below your standard just in terms of presentation. The pasted screenshots make comparisons awkward.

    SATA scores with some 500GB seagate drive.. wtf? Use some high-speed SSD, maybe some fast USB flash to evaluate the south bridge performance.

    Sometimes you have Intel scores, sometimes you don’t. How much work is it really to run SuperPi on the Intel setup?

    And lastly, why the 920? The AMD chip just came out, I think it makes more sense to compare it to current Intel chips. The Sandy Bridge comparison is fair in terms of price and availability. Who buys a 920 today?

    Overall this isn’t up to ABT’s standards, I feel.

  • http://alienbabeltech.com/main/?cat=3 apoppin

    First of all, I have to agree. And I am not going to make excuses for it. There is not even a real conclusion in my article.

    Let me start with the last issue first – the i7-920. Don’t forget that it is benched at 3.80GHz with turbo on to 4.0GHz. That is faster that the current stock i7-960, Intel’s fastest LGA 1366 quad-core on their lead platform – the X58 motherboard (until it is replaced this month). As noted in the article, it would be about as fast as a stock i5-2500K. So it is a very valid comparison. That said, we are also again evaluating new CPUs from Intel and I just received a Core i3-2105 for comparison so we can see how the dual cores perform, including the Phenom II X2 in gaming.

    The main issue with the FX-8150 evaluation was one of lack of time. Reviewers mostly had 7 days from the receipt of the FX-8150 until publication which is time pressure enough. However, the ABT review kit’s ASUS AM3+ motherboard was DoA; a hardware issue and it would not even post.

    It took over two days to get a replacement AM3+ MB and that left only about 3 days for the entire evaluation (considering that it takes a day to set up Windows and 20 games and all of the benchmarks, updates and patches – and I did it twice).

    Worst of all, CrossFire didn’t scale well with the stock-clocked FX-8150 on the replacement motherboard and it also had issues which meant a lot of retesting. While overclocking the FX-8150, the MB’s LAN became defective and a PCIe riser pulled loose from the MB with the video card, and I am now waiting for a third motherboard as replacement so that I can finish Part two of this FX-8150 Evaluation.

    In the meantime, I will clean up my charts and put up easier to see images as needed. And I will add in the few comparable Intel benches which I had no time to run. An overall summary chart would also be nice. But I am saving that for Part two which will also explore CrossFire scaling with the overclocked FX-8150 (at 4.4GHz on air and perhaps further under watercooling).

    All of ABT’s evaluations are always done from a gamer’s point of view. That is why you see 20 games benched as a minimum with far less emphasis placed on other CPU functions. Also, normally I run Total War: Shogun II and Deus Ex: Human Revolution, but they BSoD the PC when I tried to launch them with the FX-8150. AMD acknowledged it and believes that it may be due to an issue with Steam.

    Over the next couple of days, I will clean up this article. But watch for Part II – that is the important part with a real conclusion about FX-8150 as a gaming CPU. Having two weeks with it (now) makes a lot of difference than taking a superficial 3-day look at it. I am not under time pressure now.

    If I may make a suggestion to everyone reading this, follow us on ABT forum. The members find out exactly what is going on – first, before it is polished up for an article and you may even have input into the way the testing is conducted before we do it. How and what we test is driven by our members.

  • hansmuff

    Fair enough — waiting for Part II. Thanks for the explanations!