First up some newer benchmarks tested on the Asus Crosshair V (Bios 9911) for the AMD CPUs and for the 2600K on the ASUS P67 Dlx and finally the i7 990X on the Asus Rampage 3 Extreme.

Common parts used where :

• Corsair Dominator PC12800C8 rams CL8-8-8-24 1T, 4gb in total for the Dual channel setups, 6Gb for X58
• MSI GTX 480 Lightning ( downclocked at stock GTX480 clocks )
• Nvidias 280.26 WHQL drivers and 285.35 for the Battlefield 3 game test
• Corsair AX 1200 PSU
• Windows 7 Pro SP1 with all the latest patches and updates

 

Cinebench Release 11 is up first. We've always used the reliable Release 10 version, so time for a new updated version.

 

 

No surprises here as the outcome is similar to the older variant. Bulldozer with it's "8" cores pulling away from the hexacore Thuban 1090T and Intels 2500K. The 2600K and gulftown 990X are out of reach.

We've read some complaints about not including enough rendering benchmarks, so POVRay and Specpreview were installed on our test hard drives. That's more than enough rendering action my heart can take.

 

 

The brand new FX again pulling away from it's predecessor, rendering 14 secs faster then the 1090T. The Sandybridge CPUs are going completely bezerk in the POVRay test. The same can be concluded from the SPECpreview benchmark suite. This time the FX 8150 is near Gulftown performance. Yet it's closest main rivals price wise the 2500K and 2600K are lightyears ahead in performance.

 

 

The Fritz 12 Deep Chess Benchmark clearly gives away the raw calculating power of a CPU architecture. Utilising all the cores to it's fullest potential. The Bulldozer FX-8150, ahead of the 1090T and 2500K. The latters bigger brothers are untouchable in this test.

 

 

For allround PC performance Futuremarks Pcmark test suite has always been a reliable tool. Stressing the CPU, GPU, HDD's as the entire complete subsytem. The Pcmark05 suite already takes a while to complete, the 07 version is an even longer stress test.

 

 

In the older PCMark 05 the FX 8150 has a hard time to unleash it's architectural potential, finishing dead last. In PCMark 07 it can just avoid that unhonourfull last place, sadly still not being able to keep up with the tested Intel offerings.

 

 

A lot of readers and forum members still have doubts about the SuperPi test. Since it's purely single threaded, they doubt the value of this benchmark. As it hardly benefits from a multi core environment. Time to pull out a multi threaded version of Superpi (on AMDs advice). We opted to use the 2.52Gb test nr 5, as it was a good balance between the available ram in Windows 7 and the installed 4Gb. ( i7 990X ofcourse had 6Gb installed in triple channel config ) While in SuperPi the FX had a hard time to come even close to the 1090T, it's the other way around now. The extra "cores" give it a slight 9 secs advantage. The Intel CPUs still outclass the AMD offerings by miles...

 

Latest test before we plunge ourselves in some added game titles is the latest winrar 64bit version. The included benchmark tool calculates the MB/s our beloved CPUs can archieve. The AMD FX-8150 taking an impressive victory here. Without a doubt the fastest Winrarring desktop CPU on the planet at the time of writing.

 

 

 

 

 

I always tend to include game tests at medium resolutions to show the potential of a new CPU and/or platform. Even if this are not the so called real world scenarios, I try to avoid as much as possible reaching the GPU bottleneck. Otherwise it wouldn't matter as the outcome would be similar amongst many CPUs. Let's restest the games at 1280 x1024 resolution with high detail settings ( iso medium in the previous review ) and 1920 x 1080 res with everything maxed out.

 

As there are a lot of theories of disabling clusters in the CPU (motherboard bios permiting) that should increase the Instructions per clock why not test them in a few games. So we throw the 4/4 CPU in the mix.

Pull out the guns and blast them Mafiosi away with MAFIA II :

 

 

Mafia II slightly scales with raw processing power at the lower 1280 x 1024 resolution. At 1920 x 1080 high detail preset, there's hardly any difference between all the platforms. Logical as it's the GTX480 that can't keep up. Hardly any IPC gain, if any at all when running the FX-8150 at 4/4 cores. Ready for some zombie slashing ?

 

 

Frames per second are impressively different in this older Capcom game title. Benefitting heavily of the Intel architecture and it's raw processing power. The let's call them more refined AMD CPUs, still have more than enough power for a mega fluid game experience. At the maximum detail setting, the gaps get smaller, but it's no longer a secret that this game will render more FPS on an Intel based platform. As if your eyes could care less, as they can only spot 24FPS.

 

 

Capcoms Streetfighter IV, a Mortal Kombat I can push more buttons then you in a blink of an eye action game seems to use the new modules AMD CPU in a more efficient way. At the low resolution the FX CPU is ahead of the Gulftown and the 2500K. Weirdly enough the table turns when we up the resolution and detail level. Still a nice FPS advantage over the 1090T, yet the latter still manages to pump out more FPS then your brain can actually handle.

 

 

 

Lost Planet 2 was one of them you either love or hate it titles. Bringing nice DirectX11 to the gamers with incredible rendered monsters and landscapes. At the lowest tested resolution very small scaling, the Sandy Bridge based CPUs taking a nice lead. Similar outcome at the highest resolution, where the bottleneck is more graphics card orientated. Let's up the speed and go nuts in a Formula One car with Ubisofts latest sim title : Formula 1 2011.

 

 

 

Driving around the Monte Carlo track this game adores the current Intel CPU lineup. For the above test I used the timedemo, in game benchmark. Maybe not 100% accurate reflective for true performance, yet the raw cpu power is clearly visible. The FX CPU ain't bad at all, delivering plenty of FPS. Yet there are more affordable solutions out there, that do the same at a fraction of the cost.

Now two of AMD game titles :

Luckily Battlefield 3 arrived in due time for this follow up article. We ran the Swordbreaker mission and started Fraps once we stepped out of the armoured car, till the market place was completely cleaned of terrorists. The high detail preset was used for the single GPU tests at 1920 x 1080 resolution. We directly opted to go for possible daily overclocks to enhance the FX experience.

 

 

Battlefield 3 is on the AMDs preffered game title list. Meaning it's optimalised for it's new Bulldozer architecture. At 1920 x 1080 resolution all the platforms put out similar average FPS. Though with the minimum frames there's a nice boost over the previous Thuban CPUs.

 

 

Dirt 3 was tested with a ride at the Aspen track. A decent showing by the overclocked FX again, besting the more expensive Gulftown platform. The Sandybridge at a modest 4.5Ghz OC is just a tad better. All platforms scoring pretty close, lets increase the calculating power by overclocking the CPU and then adding a 2nd GPU...

 

 

In the first part of the FX-8150 review, we only pushed the FX 8150 CPU to new heights under the cooling authority of Corsairs Hydro 80. 4.8-4.9Ghz was doable during the entire test suite. Though for a daily overclock that all in one cooling solution will not be able to cut it at 4.9Ghz with my press CPU. Temperatures during Prime 95 8 thread test were rapidly breaking 92°C just after a few minutes. I hope for AMDs sake the retail silicon will be less of a hothead. And let's pray to the gods of hardware that they can bring both the heatouput and power consumption to more acceptable levels. But time for some daily OC settings in a quick comparative clash.

• 1090T overclocked at 4Ghz at 1.4Vcore, NB speed of 2600Mhz
• FX-8150 overclocked at 4.6Ghz at 1.42Vcore, NB speed 2400mhz
• 2600K overclocked at 4.5Ghz at 1.35Vcore
• i7 990X overclocked at 4.2Ghz at 1.35Vcore, 3200 uncore

Ram speeds were set at 1600mhz CL8-8-8-24 command rate 1T for all competitors. Only multiplier overclocking applied. We used the above speeds as they kept all of the CPUs in safe operating temperatures (sub 75°C) under Prime95 testing. Your mileage can vary ofcourse.

 

 

Once the CPUs are overclocked, the Cinebench scores scale accordingly. The gap between the the 6 core and FX "8" core CPU remains tiny. The 2600K@4.5Ghz enlarges the gap significantly. Chess action now. Deep Fritz 12 tells a similar tale then the Cinebench results. the FX CPU pulling away from the predecessor, even if not by a big amount.

 

 

Linear scaling with the added clockspeeds in POVRay. The gap between the AMD CPUs remains pretty constant, even at a 600Mhz CPU advantage for the FX CPU. The Gulftown CPU get's a decent boost over the stock setup, where the result was closer to the FX CPU.

 

 

Encoding a video into an HD version via the X264 encoder still shows strong pass 2 performance of the FX 8150. Intel being too strong in pass 1 of the 2.0 version.

 

 

Winrar is FX territory. Not even the 1100mhz higher clocks of the 2600K can bring it closer to AMDs latest offering. Weirdly enough the Gulftown hardly benefits from the extra speed.

 

 

Multithreaded Superpi at 2.42Gb test setting, shows the FX CPU gaining serious ground on the Intels. While at stock there was over half a minute gap, this has been reduced to half the time.

 

 

SLI gaming tests up next...

 

 

 

 

In the game tests on page 2 we saw that we were closing in on the GPU wall. Raw CPU speed didn't help much more as the GPU was fully stuffed with polygons. Time to lift that bottleneck and include a 2nd GTX 480 GPU in the tests. Let's see if the AMD platform is still able to match the competitors and it's previous processors. The CPUs remain overclocked to deliver enough speed to the dual GPUs. Sadly I don't have access to a bigger than 24" monitor. So no extreme high resolution gaming, 1920 x 1080 was the native gaming resolution tested. This will probably will be another discussion point.  Yet it can clearly show the strenghts and/or weaknesses of a platform.

 

 

Futuremarks 3DMark 06 hardly scales by adding a 2nd GPU with the AMD lineup. The Intels score an impressive 9-10K extra, while the Thuban scales a bit better then the FX CPU. Both AMD CPUs were tested on the Crosshair V with 9911 bios. Take note that the 1090T will be even more efficient on the older ASUS 890FX boards. Not a good showing of the FX in 3DMark 06. Something newer now with a Vantage run.

 

 

The Vantage score was a real surprise to me. The Thuban still outperforms the FX "8" core CPU , the latter running at 600Mhz higher clock speeds. Interesting enough is the good CPU score set by the FX 8150, yet the GPU score is way below the 1090T. At first I thought I had a few bugged runs (results are the average out of three). So I reinstalled the 1090T and redid the Nvidia drivers. Sadly for the FX the same outcome. Is this an indication for games to stay way from the Bulldozers when using multiple GPU setups ? Let's run a few more tests...

 

 

Mafia II still runs better on the older overclocked technology. At stock clocks this game was pretty evenly matched accross the different stock CPUs. Though with the added CPU clockspeed and the extra GPU we see a healthy increase in FPS ( from 61 to +100 ), but the 1090T and Intels are pulling away from the FX.

 

 

Lost Planet II shares the same experience. Close to doubling the framerates from stock setups with a single graphics card. The SandyBridge 2600K is in a class of it's own here. Crysis II tested on hardcore mode next...

 

 

Crysis II benefitting from the extra GPU big time. All platforms giving tremendous power to the GPUs. Yet again bulldozer trailing the rest....

 

 

 

Far Cry 2 known to be a great scaling engine with extra CPU and GPU speeds. Close game again between the two AMD CPUs, the Intels remain untouchable. Racing F1 cars on Monte Carlo shedding a different light on the FX ?

 

 

+

 

F1 2011, does not benefit in our test version from the 2nd GPU. (EDIT : This is apparently is driver issue related as testing with a newer driver gives better results on all platforms ) with the newer driver So results remain pretty evenly matched.Time to give Dices latest First Person shooter some extra GPU power by adding a 2nd GTX480. While with a single GPU the average FPS were rather close, this time it shows a totally different outcome. We crancked up the game detail to the preset Ultra setting, yet due to monitor limitations, the resolution stays put at 1920 x 1080. Even though the AMDS manage fluid framerates, the raw power of the Intel design gives far more minimum FPS. Naturally if a bigger or dual monitor setup is used things will get more GPU bound again.

 

 

In Dirt 3 we drove around the Aspen track several times to see how the different platfroms reacted to the added GPU power. Very high FPS all round and finally a good FPS increase over the 1090T.

 

 

 

 

 

 

The AMD press kits came without a cooler, so the reviewers had to use whatever was available in the lab. A week later most got the rebranded Asetek AMD FX all in one liquid cooling solution on the doorstep. Let's give it a spin.

 

 

 

As said before it's a rebranded Asetek/Antec 920 solution. Look at it as a pimped Corsair H70 with more flexible tubing and finally software fan control out of the box.

 

 

 

The thick 49mm 120mm fan radiator is aided by two fans to disperse the heat dumped by the FX-8150 CPU.

 

 

 

The dual fan solution, so push and pull can spin at a very low 700 RPM till 2400 RPM. They are controlled by the included software. Either opt for the silent approach ( easily doable up to 4.4-4.5Ghz ), Extreme preset or adjust it yourself with your own custom profile.

 

 

 

Quite normal to only find the AMD mounting kit in the box. I never reviewed the Antec H²0 920 cooling kit, but the mounting has seriously improved over the first Asetek/Corsair solutions.

 

 

The included Chillcontrol software leaves hardly any option open to adjust the fan speeds. Via either one of the two preset profiles ( silent and extreme ) or via a custom user defined setting. RGB colours can be adjusted, temperatures monitored, etc...

Some shots of the features of the Chill Control software :

 

 

 

 

 

We throw into the bundle Corsairs Hydro 100, 240mm rad solution and Alpenföhns K2 cooler. For the 3.6 and 4.5Ghz tests we set the Corsair to Balanced mode. Allowing the unit itself to choose the fan rpm itself. For the 4.7Ghz test the Hydro 100 was set to performance mode.

 

 

Running this CPU out of the box is no problem for any solution, even the FX cooling kit at silent mode ( really inaudible ) keeps it well under 60°C. The temps are read out by Coretemp, calibrated with a 14°C offset. Calibration was done by placing a sensor near the IHS. At 4.5Ghz and 1.39 Vcore the Silent mode starts to get a hard time to keep things cool. The FX cooling solution at Extreme preset keeps in touch with the Hydro 100 and cools adequately.

 

 

At 4.7Ghz at 1.47Vcore we ditched the silent mode and went either extreme or peformance mode for the all in one liquid cooling solutions. The monstrous K2 air cooler remains totally inaudible. The Hydro 100 takes first spot with a 5°C lead over the Asetek/AMD solution.

 

 

We could conduct the entire benchsuite at 4.9Ghz with the FX liquid cooling solution. Though prime95 was another cup of tea. Take note that your results can vary, our press CPU was an ES ( Early Sample ) CPU with a low VID of 1.2625. For your info low VID usually means a hot CPU :p

 

 

I think the former pages filled with new and maybe for some critical readers improved tests, still shed the same light on AMD's latest offering. My thoughts on the Bulldozer CPU have not really changed during the last two weeks. Not even the newer tests persuade me to reconsider changing my initial feeling about this product.

The biggest issue is that Bulldozer was too late on the market. This in reference to the performance it currently offers in most applications. Delay after delay is never a good indication of how things will fianlly turn out. Bundle the late arrival together with the unofficial leaks, the hype created around the architecture and it's performance. The overconfident PR campaign labelling this product as the first true eight core CPU. And then it finally arrives after a 5 year wait and most users notice it has a pretty rough time distinguishing itself from the previous AMD generation. What a dissapointment it must be for them that awaited a new competitive CPU. Only the real dedicated AMD fan boys still try to find a way to make AMDs latest offering shine. It must be applauded, nice dedication and devotion lads, but we will have to await at least newer and hopefully better steppings. In it's current state it's partly doomed. Piledriver might be what Bulldozer should have been right from the start.

There are two things that can improve the performance of this bulldozer CPU.

• First is to overclock it. Preferably at speeds over 4.7Ghz for daily usage. Partypooper will be the current Global Foundries limped manufacturing, which limits the performance per watt drastically. Go over 4.9Ghz and pull easily close to 400Watts out of the wall socket. Which is insane for a brand new 2011 CPU, especially with this mainstream performance. This bonkers power consumption is also one of the reasons I have a hard time with the value awards given on some websites. No idea what the value part refers too, as running a BD overclocked will be noticeable in your electricty bill. Secondly the street prices are still too high. On top of that the current low availability will not improve the price level at all. Last but not least the cost of a decent 990FX board that can cope with the power draw of these CPUs is also not that cheap. Cooling it by a basic air cooler, let alone the boxed cooler, will not handle the heat generated by a decent overclock. And honestly running this CPU at out of the box stock speeds, doesn't seem to make it worthwhile to be even considered as an upgrade over the previous AMD CPU lineup.

• A very honest comment I spotted on the web was that Bulldozer might be slightly ahead regarding application support. If the used software commands the cores/clusters correctly, then indeed the performance is quite decent. Will Windows 8 come to Bulldozers aid ? Only time will tell, but there's more required then just an improved task scheduler. The latter could boost performance with 5-10%. The trend for disabling clusters, modules is interesting to watch, but why on earth buy a so called 8 core and then just run it as 4/4. Setting different affinities per program, ... Really this should not be required to be done by the end user. Hopefully a lot of programs/games will receive an appropriate coding overhaul to be Bulldozer compatible. Yet again time will tell.

Completely off topic now. It's so weird wading through some reviews and to see the different opinions, remarks, points of criticism (if any) and the given awards. So funny to see a pretty spot on conclusion, then still award the FX 8150 with a 9.5 out of 10. Recommend it as value, high tech, must buy, and so on... No idea where the reviewer honesty level is at lately. Do they even dare to look at themselves in the mirror ? The readers that dig this kind of behaviour must love to be fooled or partly lied too. Not many grasp the whole concept behind the award rating. How it can be used as an act to get more goodies. The manufacturers praise them sites as their samples will get awarded over and over again. Good for all of us ofcourse, that there aren't any bad/flawed products around lately.

With this second addition to the FX 8150 review I  hope the Madshrimps test suite is now a bit more up to date and the Bulldozer CPU had a more fairer run.

Major point to stress, is that price wise AMD needs to react to the Intel price cuts. Intels 2500K is retailing already lower than 190 euros (FX-8120 price level)  The 2600K is plus minus retailing at  260 euros. You still have to pay about 220 euros for an FX-8150. Plus if you opt to overclock it, please consider a high end cooling solution.

For AMD's sake let's pray that Global Foundries can deliver the goods with the next batches. Taming the dreadfull power consumption to a more respectable 2011 level. From the software side it will be hard to get the required support onboard from game and other software developpers, etc...

But in BD's defence, if you look at the newer game results, you can get pretty decent frames per second. For multiple GPU solutions I honestly think the AMD FX series lack the raw power. In the synthetic SLI runs it's quite clear that this CPU will hardly be used by benchers. A suicide run is the only thing I see it usefull for. Quite harsh to outsiders, but hard reality to the overclocking community. For daily users, gamers, depending on the application/game run the Bulldozer can shine, but it's only a little comfort if it doesn't cut it with older software.

Let's all hope we will get a nice high end performing AMD surprise in 2012...

 

 

 

Logging off for the 2nd time...