AMD Phenom II X6 1090T Black Edition Performance April 27, 2010 Brandon Sandman Bell

Introduction

Page:: ( 1 / 12 )

[image]

Turbo Core/Specifications

Page:: ( 2 / 12 )

Like Intel, AMD obviously realizes that there are cases when all the processing cores inside their latest multi-core CPUs aren’t being used. Even today, many software apps are only single or dual-threaded, leaving two core idling in a quad-core CPU, and four cores idling in the case of a six-core Thuban CPU.

To address this issue, AMD offered application profiles in their latest builds of Overdrive that would automatically OC the active processing cores when certain programs were run.

This software-based solution required AMD to constantly update their profile database in Overdrive to properly work, and of course it requires the end user to actually run AMD’s Overdrive software at all times.

In other words, it’s an inelegant solution to the problem. A hardware-based solution that requires no effort from the end user, yet still delivers comparable results is more desirable. In addition, Turbo Core can help make up for some of the clock speed deficit that Thuban suffers from in comparison to today’s Deneb-based Phenom II X4 processors.

With Turbo Core, Thuban is capable of watching the utilization of the CPU’s six cores. If three or more cores aren’t currently being used by the app, the CPU can reduce the power state on those cores, and Turbo up the three cores that are being used. The unused cores can be idled down to as low as 800MHz, while the active cores can be OC’ed up to 500MHz on certain CPUs.

Notice that we said “active cores can be OC’ed up to 500MHz”, and “unused cores can be idled down to as low as 800MHz”. It’s important to note that these aren’t 100% guaranteed conditions. To keep the Thuban CPU within its 125W TDP, the CPU can’t always OC the active cores the full 500MHz. In some cases the active cores may just be clocked up 100MHz, or say 300MHz. We also noticed that the inactive cores don’t always go down to the full 800MHz clock speed. Sometimes they may be running at 1000MHz. This is a little different than Intel’s Turbo feature, which definitely does bump up the clocks the full 100% if certain conditions are met.

Also unlike Intel’s Turbo Mode, AMD Turbo Core kicks in only when three or more cores are being used. This means in cases where four cores are being taxed, AMD Turbo Core doesn’t do anything. It also doesn’t completely shut down the inactive cores like Intel Turbo Mode does. As a result, it isn’t as power efficient.

In other words, it’s definitely still much less advanced than what Intel’s offering with their latest CPUs, but it’s definitely an improvement over relying on AMD Overdrive software. It’s an incremental step that will hopefully be improved when AMD launches their next-generation Bulldozer CPUs next year.

All AMD CPUs with Turbo Core technology will be denoted with a “T” after the model number. So today we’re getting the 1090T and the 1055T CPUs, and in the future we’ll supposedly see 1075T and 1045T processors if you believe the rumors.

With that discussion out of the way, let’s go over the rest of the Thuban CPU’s specs:

AMD’s Phenom II CPU Lineup
	Phenom II X6 1090T BE	Phenom II X6 1055T	Phenom II X4 965 BE	Phenom II X4 955 BE	Phenom II X4 945	Phenom II X4 925
# of Cores	6	6	4	4	4	4
Clock Speed	3.2GHz	2.8GHz	3.4GHz	3.2GHz	3.0GHz	2.8GHz
Max Turbo Speed	3.6GHz	3.3GHz	N/A	N/A	N/A	N/A
L2 Cache Size	2MB	2MB	2MB	2MB	2MB	2MB
L3 Cache Size	6MB	6MB	6MB	6MB	6MB	6MB
TDP	125W	125W	125W/140W	125W	95W	95W
Price	$295	$199	$185	$165	$155	$145
Die Size	346mm2	346mm2	258mm2	258mm2	258mm2	258mm2
NOTE: Excludes Phenom II EE CPUs

Notes

If you recall our review of Intel’s Core i7-980X six-core CPU, you’ll remember that it ships with a larger 12MB L3 cache. Intel’s engineers increased the size of the L3 to maintain what they believe is the optimal ratio between L3 cache and cores which is 2MB of L3 cache per core (although keep in mind that the L3 cache is shared amongst all six cores, so one core can use as much cache as it needs, including the full 12MB L3 for single-threaded apps). Intel was able to increase the size of the CPU’s L3 cache without upping the chip’s die size thanks to its smaller 32-nm manufacturing process. Thanks to 32-nm, Intel could add cache without making the chip bigger, and thus it isn’t more expensive for Intel to produce (assuming of course Intel is getting equal yields which may or may not be the case).

With more L3 cache on board, the Core i7-980X was faster than the Core i7-975 even in cases where only four cores were being used. This is because the extra cache gave the CPU a performance boost in apps that weren’t threaded for six cores.

Since Thuban is built at 45-nm, AMD doesn’t have this luxury. The CPU’s die is already 88 square millimeters larger than Deneb. This comes from the two extra cores. If they would’ve bumped up the CPU’s L3 cache, the chip’s die size would be even larger and they wouldn’t be able to price it aggressively and still make money. As you can see in the chart above, it therefore ships with the same L2 and L3 cache configuration as quad-core Deneb Phenom II CPUs.

The future

While AMD won’t confirm it, the rumor mill says that AMD will eventually roll out quad-core Thuban CPUs later this year. These parts will merely have two of their processing cores disabled. Presumably the first CPU will be the Phenom II X4 960T. Rumors have also suggested we’ll see a 940T part at some time as well.

890FX Chipset and the ASUS Crosshair IV Formula

Page:: ( 3 / 12 )

While the new Thuban CPUs will run fine in existing socket AM3 790FX motherboards with a BIOS update supporting the new processors, today AMD is also rolling out a new chipset, 890FX.

Again, like the 890GX launch a month ago, the new 890FX chipset isn’t a next-generation built from the ground up. It’s an evolutionary rather than revolutionary upgrade over its direct predecessor, with the most significant changes found in the South Bridge of the chipset.

Tweaks have been made to the North Bridge, but many of the key specs remain largely unchanged. PCI Express lanes for instance remains the same at 42, which is enough for 2x16 or 4x8 CrossFire (in comparison 890GX is limited to 2x8).

Like the 890GX chipset, 6x1 PCI Express 2.0 lanes hang off the North Bridge, which can be used by motherboard manufacturers to implement full bandwidth USB 3.0 support. While this isn’t as elegant as a native USB 3.0 interface that’s built-in to the chipset, it’s better than nothing.

890FX is paired with AMD’s SB850 South Bridge. As we mentioned in the 890GX article, it offers more USB support (14 USB 2.0 ports in SB850 versus SB750’s 12), a native Gigabit Ethernet MAC, and two additional x1 PCIe 2.0 lanes. Its most notable feature is its 6Gbps SATA support. SB850 supports up to six 6Gbps SATA drives.

The ASUS Crosshair IV Formula Motherboard

To test the capabilities of the new 890FX chipset, AMD sent us a Crosshair IV Formula motherboard from ASUS. The Crosshair IV Formula is ASUS’ newest Republic of Gamers motherboards, and is loaded with features that enthusiasts will appreciate.

[image]

System Setup

Page:: ( 4 / 12 )

Intel Core 2 Duo E8600
Intel Core 2 Quad Q6700
ASUS P5E3 Premium

AMD Athlon II X4 635
AMD Phenom II X4 965 Black Edition
Gigabyte GA-790FXTA-UD5

AMD Athlon II X6 1090T Black Edition
ASUS Crosshair IV Formula

Intel Core i7-870
Intel Core i5-750
ASUS P7P55D Deluxe

4GB (2x2GB) OCZ Platinum @ DDR3-1333 Speeds

Intel Core i7-980X Extreme Edition
Intel Core i7-975 Extreme Edition
Intel Core i7-920
Gigabyte X58A-UD7

6GB (2x3GB) OCZ Reaper HPC @DDR3-1066 Speeds

ATI Radeon 5870
Catalyst 10.2

2TB Seagate Barracuda XT

Windows 7 Ultimate 64-bit

Benchmarks

Resident Evil 5
Crysis
Far Cry 2
HAWX
Lost Planet
Cinebench 10
WinRAR 3.9
LAME MT
DivX 7
VirtualDub 1.9.8

Media Encoding/Rendering Benchmarks

Page:: ( 5 / 12 )

Valve Particle Simulation Benchmark

Far Cry 2

Page:: ( 6 / 12 )

Far Cry 2 – Direct3D

Crysis

Page:: ( 7 / 12 )

Crysis – Direct3D

HAWX	Page:: ( 8 / 12 )

Tom Clancy’s HAWX – Direct3D

Lost Planet

Page:: ( 9 / 12 )

Lost Planet – Direct3D

GTA: Episodes from Liberty City

Page:: ( 10 / 12 )

Overclocking/Power Consumption

Page:: ( 11 / 12 )

Power Consumption

Overclocking

With its extra cores, you’d naturally assume that the Phenom II X6 1090T wouldn’t scale as well as Phenom II X4 CPUs – especially given that it’s based on the same 45-nm manufacturing process. However, we hit a new high with our 1090T sample. Check it:

[image]

Conclusion

Page:: ( 12 / 12 )

Anyone who plans to use the processor for heavy-duty rendering work will really appreciate the new Thuban processors. Our results in Cinebench 11.5 in particular were the most impressive. Scoring 5.69 points, the Phenom II X6 CPU was third only to the Core i7-980X and Core i7-975 EE CPUs in performance. Keep in mind that it costs hundreds of dollars less than the Intel CPUs; you could actually buy three Phenom II X6 1090T CPUs and still have over $100 in your pocket. Or you could build yourself an entire Phenom II X6 1090T system with Radeon 5850 graphics for the price of one Core i7-980X.

This processor is simply a tremendous value if you happen to fall in this category.

Most of our readers are die-hard gamers though and here the results are mixed. In apps that are heavily threaded, the Phenom II X6 1090T obviously shines the most. For an example of this, there’s no better example than Capcom’s MT Framework engine used in games like Lost Planet, Devil May Cry, and Resident Evil 5. In Lost Planet the Phenom II X6 1090T ran 9% faster than the Phenom II 965. At the same time however this still wasn’t enough performance for the chip to outrun Intel’s Core i7-920.

In the other games, the Phenom II X6 1090T was outrun by the Phenom II X4 965. Here its 3.2GHz clock speed was an obvious liability.

AMD has tried to address this issue with Turbo Core, but we’ve found that it doesn’t consistently hit its peak Turbo Core clock speeds. Using Overdrive to monitor clocks, we noticed that speeds were all over the map, even in the same app. Sometimes the cores on the 1090T would go up to 3.3GHz or even 3.4GHz for some time, then they’d go back down as the other cores speeds came back up.

We never saw Turbo Core hit 3.6GHz in games, and even in the apps where it would get that high, it was never there for more than a few seconds. Also keep in mind that in apps that have four threads, Turbo Core is inactive.

As always with game tests though, the results at 800x600 with 0xAA aren’t very relevant for gamers, as no one plays games that way. Instead most gamers crank up the screen resolution and graphics settings, including turning on AA. In these instances, your performance is entirely dependent on your GPU, not the CPU – here the Phenom II X6 1090T runs just as fast as the Core i7-980X.

We didn’t note anything game-changing in MP3 encoding, compression, or testing with Valve’s particle simulation benchmark (which is multithreaded) for the Phenom II X6 1090T. In these cases it took a backseat in performance to comparably priced Core i CPUs from Intel.

Based on this information, it’s harder to recommend the Phenom II X6 1090T to gamers than say the more affordably priced Phenom II X4 955 BE and 965 BE. We have a feeling that the 960T could be more attractive as well depending on pricing.

It’s a shame that Turbo Core’s clocks are so variable. On paper at least it looks extremely attractive. In practice though, the advertised Turbo Core speeds aren’t hit consistently enough, and even when they are, it can be for short spurts. Not long enough to dramatically impact performance unfortunately. Hopefully this is something AMD will address with their next-generation CPUs due next year.

The beauty of AMD’s Phenom II X6 CPUs are their power consumption. Despite integrating two more cores on the same manufacturing process, we observed lower power consumption for the 1090T than our 125W Phenom II X4 965. Keep in mind that we use Cinebench to test CPU power consumption, so all six cores were being taxed for this test. Overclocking looks extremely solid as well.

This is simply a remarkable achievement that’s a testament to Globalfoundries 45-nm process. Clearly they’ve now got it very well dialed in.

Now we await the arrival of cheaper six-core Phenom II CPUs, as well as the first quad-core Thubans. All of these chips should sport the sweet power consumption of the 1090T, but with lower prices. This could be a killer combo that will be hard for many enthusiasts to pass up.