A month ago we saw the launch of the Phenom II X4 940 & 920 processors. Today AMD is launching FIVE new chips to add to their Phenom II lineup. We are finally seeing the socket AM3 (DDR3) design that has been talked about for many months. Unfortunately, while the chips are ready for DDR3, motherboard manufacturers are still working diligently to get AM3 boards in the retail channels as soon as possible. However, these processors are designed to work in existing AM2+ motherboards requiring nothing more than a simple BIOS update. So whether building a new system or upgrading an existing one, these processors offer an inexpensive solution for the enthusiast that wants to get most performance out of their AM2+ system.
The two chips we will be taking a look at specifically are the Phenom II X4 810 and X3 720 Black Edition. Being a Black Edition processor means that the CPU multiplier is unlocked, making it much easier to overclock. The Northbridge Frequency (HT/MC) is back up and running at 2GHz for all these AM3 processors (The X4 940/920 are running at 1.8GHz). While the X3 7xx series have 6MB of L3 cache (just like the X4 9xx series), the X4 8xx processors have only 4MB of L3 cache.
In the picture below, the CPU on the left is the typical AM2+ processor, the one on the right is an AM3. Notice the omission of a couple pins? The corresponding AM3 sockets will have those holes filled to ensure a person doesn't try to insert an AM2+ processor into an AM3 board. Meanwhile, the AM3 chip will fit perfectly into an AM2+ board.
AMD Phenom II Common AM3 Processor Technical Specifications:
* Processor Frequency & L3 Cache:
o X4 910 = 2.6 GHz / 6 MB L3
o X4 810 = 2.6 GHz / 4 MB L3
o X4 805 = 2.5 GHz / 4 MB L3
o X3 720 = 2.8 GHz / 6 MB L3
o X3 710 = 2.6 GHz / 6 MB L3
* L1 Cache Sizes:zes: 64K of L1 instruction and 64K of L1 data cache per core (512KB total L1 per processor)
* L2 Cache Sizes: 512KB of L2 data cache per core (2MB total L2 per processor)
* L3 Cache Size: 4MB or 6MB (shared between cores)
* Memory Controller Type:ype: Integrated 128-bit wide memory controller*
* Memory Controller Speed: Up to 2.0GHz with Dual Dynamic Power Management
* Types of Memory Supported:
o PC2 8500 / DDR2-1066MHz
o PC3 10666 / DDR3-1333MHz**
* HyperTransport 3.0 Link: One 16-bit/16-bit link @ up to 4.0GHz full duplex (2.0GHz x2)
* Total Processor Bandwidth: Up to 33.1 GB/s total bandwidth
* Packaging: Socket AM3 938-pin organic micro pin grid array (micro-PGA)
* Fab location: Fab 36 wafer fabrication facilities in Dresden, Germany
* Process Technology: 45-nanometer DSL SOI (silicon-on-insulator) technology
* Approximate Transistor count: ~ 758 million (45nm)
* Approximate Die Size: 258 mm² (45nm)
* Max Ambient Case Temp: 71° Celsius for X4 / 73° Celsius for X3
* Nominal Voltage: 0.875 - 1.425 Volts
* Max TDP: 95 Watts
* Code Name: Deneb
*Note: Configurable for dual 64-bit channels for simultaneous read/writes
**Note: For DDR3-1333, AM3 boards will support 1-DIMM-per-channel @ 1333MHz
Target Market & Competition:
AMD is marketing these new Phenom II AM3 processors against Intel's mainstream / budget quad & dual core chips. Typically gamers like to buy the dual-core processors because there isn't much of an advantage using a quad core in games and that extra cash can be put towards a better video card. AMD has capitalized on the tri-core design, giving a little extra boost in performance over dual-core chips, but at a substantial cost savings over quad-core chips.
In the table below are the closest processors Intel has in price. You can certainly spend more for a faster processor, but here we are trying to show what kind of bang you get for your buck. For the Intel chips I just took the prices from Newegg (my favorite place to buy computer parts). Sometimes it's hard to make good comparison with so many different mixes of CPU speed, CPU cores, cache size, bus speed, etc... It can really start to make your head hurt if you think about it too much. Anyhow, you can see that the next jump in Intel chips is about $20-$25 more, which the price for a Q8300 would be better compared to a X4 920 (currently $189 at Newegg).
Test Setup, Methodology, and Benchmarks Used:
Too keep things consistent and try to maintain and "apples to apples" comparison, we try to use as much common hardware across the board. Since the Phenom II is still running only DDR2, we went with that for the comparative Intel systems. Each system's BIOS was set to "Optimized Defaults", with a verification of CPU & Memory speeds and timings.
We started out with a clean install of Vista and performed all the Windows updates to bring it up to current levels as of when this review was written. Then we installed all out benchmark programs and ran a disk defrag. Afterwards we disabled some un-needed Vista features that could affect results such as SuperFetch, System Restore, Defrag, Indexing, and Remote Assistance. Furthermore we ran the Bapco Auto-Configuration Tool and used its "Level 3" settings which disables a bunch more stuff:
* Disables screen saver
* Disables Windows Update
* Disables Desktop Cleanup Wizard
* Disables User Account Control (UAC)
* Disables the Sidebar
* Disables Windows Security Center warning messages
* Disables Windows Firewall
* Disables incoming Remote Desktop connections
* Disables Windows Error Reporting to Microsoft
* Prevents the Welcome Center from running at startup
* Disables Windows Defender
Each of the benchmarks were run at least five times, with the average being reported.
The following tests were performed:
* Synthetic Benchmarks:
o SiSoftware Sandra - 2009 SP1 (15.60)
o Lavalys Everest Ultimate Edition - v4.60
o WinRAR - 3.80
* 3D Rendering Performance:
o POV-Ray - 3.7 beta 29
o MAXON CINEBENCH - R10
* Video Encoding Performance:
o Tech ARP's x264 Benchmark - v2.0
o Windows Media Encoder 9 (x64 Edition)
o Divx Encoder - 6.6.1 (Codec 6.8.5)
* Math Calculation Performance:
o Wolfram Research Mathematica - v7
o wPrime - v2.00
o PiFast - 4.3
o HyperPi - 0.99b (Contains Super-Pi Mod v1.5)
* Gaming Performance:
o Lost Planet: Colonies Edition
o Far Cry 2
o Crysis Warhead
o Enemy Territory: Quake Wars
o Valve Particle Simulation Benchmark
o Valve VRAD map compilation tool
* Miscellaneous:
o Notfred Folding @ Home Benchmark
Everest CPU Queen's Problem:
This simple integer benchmark focuses on the branch prediction capabilities and the misprediction penalties of the CPU. It finds the solutions for the classic "Queens problem" on a 10 by 10 sized chessboard.
At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. For example -- with HyperThreading disabled -- the Intel Northwood core processors get higher scores than the Intel Prescott core based ones due to the 20-step vs 31-step long pipeline. However, with enabled HyperThreading the picture is controversial, because due to architectural bottlenecks the Northwood core runs out of internal resources and slows down. Similarly, at the same clock speed AMD K8 class processors will be faster than AMD K7 ones due to the improved branch prediction capabilities of the K8 architecture.
CPU Queen test uses integer MMX, SSE2 and SSSE3 optimizations. It consumes less than 1 MB system memory and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.
Everest CPU PhotoWorxx Processing:
This integer benchmark performs different common tasks used during digital photo processing.
It performs the following tasks on a very large RGB image:
* Fill
* Flip
* Rotate90R (rotate 90 degrees CW)
* Rotate90L (rotate 90 degrees CCW)
* Random (fill the image with random coloured pixels)
* RGB2BW (color to black & white conversion)
* Difference
* Crop
This benchmark stresses the integer arithmetic and multiplication execution units of the CPU and also the memory subsystem. Due to the fact that this test performs high memory read/write traffic, it cannot effectively scale in situations where more than 2 processing threads used. For example, on a 8-way Pentium III Xeon system the 8 processing threads will be "fighting" over the memory, creating a serious bottleneck that would lead to as low scores as a 2-way or 4-way similar processor based system could achieve.
CPU PhotoWorxx test uses only the basic x86 instructions, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.
Overclocking:
Since the Phenom II X3 720 chip is a Black Edition (unlocked multiplier), I simply used the AMD Overdrive tool to increase the multiplier and adjust the CPU voltage. The X4 810 has a locked multiplier, thus the only way to overclock it is to increase the HT reference clock. For air cooling I tried a Zalman CNPS-9700 and Scythe Infinity. Unfortunately because of some other unforeseen issues (not related to this review) I was not able to do as thorough of overclocking as I would have liked. Overall I'm quite pleased with the results, and they seem to be in the same ballpark as others.
Here's a breakdown of the numbers, all tests were done on air (as mentioned above):
* X4 810
o 2.6 GHz / 1.325V / 200 HT = Stock
o 3.06 GHz / 1.325V / 203 HT = Max Stable OC @ Default Voltage
o 3.4 GHz / 1.525V / 262 HT = Max Stable OC
* X3 720 BE
o 2.8 GHz / 1.325V = Stock
o 3.3 GHz / 1.325V = Max Stable OC @ Default Voltage
o 3.7 GHz / 1.50V = Max Stable OC
The two chips we will be taking a look at specifically are the Phenom II X4 810 and X3 720 Black Edition. Being a Black Edition processor means that the CPU multiplier is unlocked, making it much easier to overclock. The Northbridge Frequency (HT/MC) is back up and running at 2GHz for all these AM3 processors (The X4 940/920 are running at 1.8GHz). While the X3 7xx series have 6MB of L3 cache (just like the X4 9xx series), the X4 8xx processors have only 4MB of L3 cache.
In the picture below, the CPU on the left is the typical AM2+ processor, the one on the right is an AM3. Notice the omission of a couple pins? The corresponding AM3 sockets will have those holes filled to ensure a person doesn't try to insert an AM2+ processor into an AM3 board. Meanwhile, the AM3 chip will fit perfectly into an AM2+ board.
AMD Phenom II Common AM3 Processor Technical Specifications:
* Processor Frequency & L3 Cache:
o X4 910 = 2.6 GHz / 6 MB L3
o X4 810 = 2.6 GHz / 4 MB L3
o X4 805 = 2.5 GHz / 4 MB L3
o X3 720 = 2.8 GHz / 6 MB L3
o X3 710 = 2.6 GHz / 6 MB L3
* L1 Cache Sizes:zes: 64K of L1 instruction and 64K of L1 data cache per core (512KB total L1 per processor)
* L2 Cache Sizes: 512KB of L2 data cache per core (2MB total L2 per processor)
* L3 Cache Size: 4MB or 6MB (shared between cores)
* Memory Controller Type:ype: Integrated 128-bit wide memory controller*
* Memory Controller Speed: Up to 2.0GHz with Dual Dynamic Power Management
* Types of Memory Supported:
o PC2 8500 / DDR2-1066MHz
o PC3 10666 / DDR3-1333MHz**
* HyperTransport 3.0 Link: One 16-bit/16-bit link @ up to 4.0GHz full duplex (2.0GHz x2)
* Total Processor Bandwidth: Up to 33.1 GB/s total bandwidth
* Packaging: Socket AM3 938-pin organic micro pin grid array (micro-PGA)
* Fab location: Fab 36 wafer fabrication facilities in Dresden, Germany
* Process Technology: 45-nanometer DSL SOI (silicon-on-insulator) technology
* Approximate Transistor count: ~ 758 million (45nm)
* Approximate Die Size: 258 mm² (45nm)
* Max Ambient Case Temp: 71° Celsius for X4 / 73° Celsius for X3
* Nominal Voltage: 0.875 - 1.425 Volts
* Max TDP: 95 Watts
* Code Name: Deneb
*Note: Configurable for dual 64-bit channels for simultaneous read/writes
**Note: For DDR3-1333, AM3 boards will support 1-DIMM-per-channel @ 1333MHz
Target Market & Competition:
AMD is marketing these new Phenom II AM3 processors against Intel's mainstream / budget quad & dual core chips. Typically gamers like to buy the dual-core processors because there isn't much of an advantage using a quad core in games and that extra cash can be put towards a better video card. AMD has capitalized on the tri-core design, giving a little extra boost in performance over dual-core chips, but at a substantial cost savings over quad-core chips.
In the table below are the closest processors Intel has in price. You can certainly spend more for a faster processor, but here we are trying to show what kind of bang you get for your buck. For the Intel chips I just took the prices from Newegg (my favorite place to buy computer parts). Sometimes it's hard to make good comparison with so many different mixes of CPU speed, CPU cores, cache size, bus speed, etc... It can really start to make your head hurt if you think about it too much. Anyhow, you can see that the next jump in Intel chips is about $20-$25 more, which the price for a Q8300 would be better compared to a X4 920 (currently $189 at Newegg).
Test Setup, Methodology, and Benchmarks Used:
Too keep things consistent and try to maintain and "apples to apples" comparison, we try to use as much common hardware across the board. Since the Phenom II is still running only DDR2, we went with that for the comparative Intel systems. Each system's BIOS was set to "Optimized Defaults", with a verification of CPU & Memory speeds and timings.
We started out with a clean install of Vista and performed all the Windows updates to bring it up to current levels as of when this review was written. Then we installed all out benchmark programs and ran a disk defrag. Afterwards we disabled some un-needed Vista features that could affect results such as SuperFetch, System Restore, Defrag, Indexing, and Remote Assistance. Furthermore we ran the Bapco Auto-Configuration Tool and used its "Level 3" settings which disables a bunch more stuff:
* Disables screen saver
* Disables Windows Update
* Disables Desktop Cleanup Wizard
* Disables User Account Control (UAC)
* Disables the Sidebar
* Disables Windows Security Center warning messages
* Disables Windows Firewall
* Disables incoming Remote Desktop connections
* Disables Windows Error Reporting to Microsoft
* Prevents the Welcome Center from running at startup
* Disables Windows Defender
Each of the benchmarks were run at least five times, with the average being reported.
The following tests were performed:
* Synthetic Benchmarks:
o SiSoftware Sandra - 2009 SP1 (15.60)
o Lavalys Everest Ultimate Edition - v4.60
o WinRAR - 3.80
* 3D Rendering Performance:
o POV-Ray - 3.7 beta 29
o MAXON CINEBENCH - R10
* Video Encoding Performance:
o Tech ARP's x264 Benchmark - v2.0
o Windows Media Encoder 9 (x64 Edition)
o Divx Encoder - 6.6.1 (Codec 6.8.5)
* Math Calculation Performance:
o Wolfram Research Mathematica - v7
o wPrime - v2.00
o PiFast - 4.3
o HyperPi - 0.99b (Contains Super-Pi Mod v1.5)
* Gaming Performance:
o Lost Planet: Colonies Edition
o Far Cry 2
o Crysis Warhead
o Enemy Territory: Quake Wars
o Valve Particle Simulation Benchmark
o Valve VRAD map compilation tool
* Miscellaneous:
o Notfred Folding @ Home Benchmark
Everest CPU Queen's Problem:
This simple integer benchmark focuses on the branch prediction capabilities and the misprediction penalties of the CPU. It finds the solutions for the classic "Queens problem" on a 10 by 10 sized chessboard.
At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. For example -- with HyperThreading disabled -- the Intel Northwood core processors get higher scores than the Intel Prescott core based ones due to the 20-step vs 31-step long pipeline. However, with enabled HyperThreading the picture is controversial, because due to architectural bottlenecks the Northwood core runs out of internal resources and slows down. Similarly, at the same clock speed AMD K8 class processors will be faster than AMD K7 ones due to the improved branch prediction capabilities of the K8 architecture.
CPU Queen test uses integer MMX, SSE2 and SSSE3 optimizations. It consumes less than 1 MB system memory and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.
Everest CPU PhotoWorxx Processing:
This integer benchmark performs different common tasks used during digital photo processing.
It performs the following tasks on a very large RGB image:
* Fill
* Flip
* Rotate90R (rotate 90 degrees CW)
* Rotate90L (rotate 90 degrees CCW)
* Random (fill the image with random coloured pixels)
* RGB2BW (color to black & white conversion)
* Difference
* Crop
This benchmark stresses the integer arithmetic and multiplication execution units of the CPU and also the memory subsystem. Due to the fact that this test performs high memory read/write traffic, it cannot effectively scale in situations where more than 2 processing threads used. For example, on a 8-way Pentium III Xeon system the 8 processing threads will be "fighting" over the memory, creating a serious bottleneck that would lead to as low scores as a 2-way or 4-way similar processor based system could achieve.
CPU PhotoWorxx test uses only the basic x86 instructions, and it is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.
Overclocking:
Since the Phenom II X3 720 chip is a Black Edition (unlocked multiplier), I simply used the AMD Overdrive tool to increase the multiplier and adjust the CPU voltage. The X4 810 has a locked multiplier, thus the only way to overclock it is to increase the HT reference clock. For air cooling I tried a Zalman CNPS-9700 and Scythe Infinity. Unfortunately because of some other unforeseen issues (not related to this review) I was not able to do as thorough of overclocking as I would have liked. Overall I'm quite pleased with the results, and they seem to be in the same ballpark as others.
Here's a breakdown of the numbers, all tests were done on air (as mentioned above):
* X4 810
o 2.6 GHz / 1.325V / 200 HT = Stock
o 3.06 GHz / 1.325V / 203 HT = Max Stable OC @ Default Voltage
o 3.4 GHz / 1.525V / 262 HT = Max Stable OC
* X3 720 BE
o 2.8 GHz / 1.325V = Stock
o 3.3 GHz / 1.325V = Max Stable OC @ Default Voltage
o 3.7 GHz / 1.50V = Max Stable OC
No comments:
Post a Comment