AMD A6-3650 vs Intel Pentium G870
Mini Review
AMD A6-3650 advantages
 |
|
Intel Pentium G870 advantages
 |
|
A6-3650 vs G870 performance comparison
The charts demonstrate performance of A6-3650 and Pentium G870 processors in a number of different task types. The numbers for 'Single-threaded performance', 'Multi-threaded performance' and 'Integrated Graphics performance' charts are calculated as an average of several related benchmarks, that you can find in the "Benchmarks" tab. Consequentally, the shown results may differ from results, that you may see in individual tests. The 'Overall performance' shows an average of all-around performance benchmarks, and benchmarks, specific to each type of application. The 'Price-performance ratio' chart displays how much more performance you will get per dollar spent. This ratio does not include the cost of other system components. | |||||||||||||||||||||||||||||
Higher is better |
Higher is better |
Higher is better |
|||||||||||||||||||||||||||
Higher is better |
Higher is better |
Higher is better |
|||||||||||||||||||||||||||
Full list of benchmarks, that show difference in performance between A6-3650 and G870 in various types of programs, can be seen in the "Benchmarks" tab.
A6-3650 vs G870 power consumption and price comparison
Lower is better |
Lower is better |
Detailed side by side comparison of AMD A6-3650 vs Pentium G870 specs can be found in the "Specifications" tab. Even more detailed comparison of low level features of both microprocessors is located in the "CPUIDs" tab.
Pros and Cons summary
| A6-3650 | Pentium G870 |
|
General recommendations: Faster overall,
Performs much better in graphics programs, Better price / performance ratio |
General recommendations: Much faster in single-threaded tasks,
Supports SSE4.1 / SSE4.2 instructions, Somewhat more energy efficient, A bit cheaper |
|
Drawbacks: Slow in single-threaded applications,
Lacks some instructions, Less energy efficient, A bit more expensive |
Drawbacks: Not as fast in general,
Performs significantly worse in graphics programs, Somewhat worse price / performance ratio |
Similar processors
| A6-3650 | Pentium G870 |
Similar CPUs use the same socket and architecture as AMD A6-3650 and G870.
Specifications
| AMD A6-3650 | Intel Pentium G870 | |
| Basic details | ||
| Market segment | Desktop | |
|---|---|---|
| Manufacturer | AMD | Intel |
| Family | AMD A6-Series | Intel Pentium Dual-Core |
| Model number | A6-3650 | G870 |
| CPU part number | AD3650WNZ43GX | CM8062307260115 |
| Box part number | AD3650WNGXBOX | BX80623G870 |
| Introduction date | July 3, 2011 | June 3, 2012 |
| Current official price | $77 (as of Jun 2013) | $75 (as of Jan 2013) |
| CPU features | ||
| Core name | Llano | Sandy Bridge |
| Platform name | Lynx | |
| Microarchitecture | K10 | Sandy Bridge |
| Technology (micron) | 0.032 | |
| Data width (bits) | 64 | |
| Socket | Socket FM1 | Socket 1155 |
| Frequency (MHz) | 2600 | 3100 |
| Clock Multiplier | 31 | |
| L1 cache | 256 KB (code) / 256 KB (data) | 64 KB (code) / 64 KB (data) |
| L2 cache (KB) | 4096 | 512 |
| L3 cache (KB) | 3072 | |
| Max temperature (°C) | 72.7 | 69.1 |
| TDP (Watt) | 100 | 65 |
| Core voltage (V) | 0.45 - 1.4125 | |
| Cores | 4 | 2 |
| Threads | 4 | 2 |
| Multiprocessing | 1 | |
| Instruction set extensions | ||
| 3DNow! | + | - |
| AMD64 / EM64T 64-bit technology | + | |
| MMX | + | |
| SSE | + | |
| SSE2 | + | |
| SSE3 | + | |
| SSE4.1 | - | + |
| SSE4.2 | - | + |
| SSSE3 / Supplemental SSE3 | - | + |
| Supported technologies | ||
| PowerNow! / Enhanced SpeedStep | + | |
| Virtualization | + | |
| Virus Protection / Execute Disable bit | + | |
| Integrated Graphics | ||
| GPU Type | Radeon HD 6530D | HD (Sandy Bridge) |
| Base frequency (MHz) | 443 | 850 |
| Integrated Memory Controller(s) | ||
| The number of controllers | 1 | |
| Memory channels | 2 | |
| Supported memory | DDR3-1866 | DDR3-1066 DDR3-1333 |
Rows with different specifications or features are highlighted.
For detailed specifications of "AMD A6-3650" or "Intel Pentium G870" parts please click on the links in the table header.
Benchmarks
System setup
Below is a partial set of AMD A6-3650 and Intel Pentium G870 benchmarks from our CPU benchmark database. The former processor was tested on ASUS F1A75-V Pro motherboard with 4GB dual-channel Crucial BallistixPC3-14900 (DDR3-1866) memory, and Integrated on-chip ATI Radeon 6000-series graphics. The latter processor was tested on ASUS P8H61-M motherboard, that had 2 GB dual-channel Corsair CM3X1024-1333C9 DDR3 memory, and Integrated on-chip Intel Sandy Bridge graphics. For different system configurations please see tests of A6-3650 and Intel Pentium G870 CPUs in our benchmark database.
All tests were performed at default frequency and voltage, using manufacturer's stock fan/heatsink. None of the components were overclocked. Motherboard BIOS options were left at default settings.
The results of all benchmarks are broken into four categories: multi-threaded, single-threaded, memory intensive and graphics / gaming.
A6-3650 vs Intel Pentium G870 single-threaded benchmarks
Single-threaded benchmarks run on a single CPU core, and do not depend on such features as the number of cores, or Hyper-Threading technology. Additionally, they do not utilize on-chip caches, dedicated to other cores.
AMD A6-3650 Intel Pentium G870 | ||
3DMark03 CPU score benchmark | ||
Longer is better  | ||
|---|---|---|
| 100% | ||
| 59.3% | ||
CINEBENCH R10 CPU score (1 core) benchmark | ||
| Longer is better | ||
| 100% | ||
| 59.1% | ||
FLAC encoding - single process benchmark | ||
| Longer is better | ||
| 100% | ||
| 69.3% | ||
H.264 video encoding - single process benchmark | ||
| Longer is better | ||
| 100% | ||
| 64.3% | ||
LAME MP3 encoding - single process benchmark | ||
| Longer is better | ||
| 100% | ||
| 60.9% | ||
AMD A6-3650 Intel Pentium G870 | ||
PCMark2002 CPU score benchmark | ||
| Longer is better | ||
| 100% | ||
| 71.8% | ||
Super PI (1M) benchmark | ||
 Shorter is better | ||
| 100% | ||
| 46.3% | ||
XVid video encoding - single process benchmark | ||
| Longer is better | ||
| 100% | ||
| 68.1% | ||
A6-3650 vs Intel Pentium G870 multi-threaded benchmarks
Multi-threaded benchmarks utilize all CPU cores and other on-chip resources (on-chip caches, internal buffers, etc). Intel's Hyper-Threading feature also helps to improve multi-threading performance.
AMD A6-3650 Intel Pentium G870 | ||
3DMark06 CPU score benchmark | ||
| Longer is better | ||
|---|---|---|
| 100% | ||
| 75.4% | ||
Apache web server - dynamic pages benchmark | ||
| Longer is better | ||
| 100% | ||
| 71.5% | ||
Apache web server - static pages benchmark | ||
| Longer is better | ||
| 100% | ||
| 95.3% | ||
Blowfish encryption benchmark | ||
| Longer is better | ||
| 100% | ||
| 76.6% | ||
CINEBENCH R10 CPU score benchmark | ||
| Longer is better | ||
| 100% | ||
| 92.5% | ||
AMD A6-3650 Intel Pentium G870 | ||
CINEBENCH R10 render time (seconds) benchmark | ||
| Shorter is better | ||
| 100% | ||
| 92.4% | ||
CrystalMark ALU benchmark | ||
| Longer is better | ||
| 100% | ||
| 81.2% | ||
CrystalMark FPU benchmark | ||
| Longer is better | ||
| 100% | ||
| 65.1% | ||
Euler3D benchmark | ||
| Longer is better | ||
| 100% | ||
| 98.7% | ||
FLAC encoding - multiple processes benchmark | ||
| Longer is better | ||
| 100% | ||
| 72.3% | ||
AMD A6-3650 Intel Pentium G870 | ||
Fritz Chess Benchmark benchmark | ||
| Longer is better | ||
| 100% | ||
| 69.9% | ||
H.264 video encoding - multiple processes benchmark | ||
| Longer is better | ||
| 100% | ||
| 78.1% | ||
H.264 video encoding - multithreaded benchmark | ||
| Longer is better | ||
| 100% | ||
| 94.3% | ||
LAME MP3 encoding - multiple processes benchmark | ||
| Longer is better | ||
| 100% | ||
| 82.3% | ||
PCMark05 CPU score benchmark | ||
| Longer is better | ||
| 100% | ||
| 92.5% | ||
AMD A6-3650 Intel Pentium G870 | ||
PCMark05 score benchmark | ||
| Longer is better | ||
| 100% | ||
| 88.2% | ||
Sandra Dhrystone (MIPS) benchmark | ||
| Longer is better | ||
| 100% | ||
| 75.6% | ||
Sandra MultiMedia Floating Point (it/s) benchmark | ||
| Longer is better | ||
| 100% | ||
| 73.6% | ||
Sandra MultiMedia Integer (it/s) benchmark | ||
| Longer is better | ||
| 100% | ||
| 48.3% | ||
Sandra Whetstone (MFLOPS) benchmark | ||
| Longer is better | ||
| 100% | ||
| 56.3% | ||
AMD A6-3650 Intel Pentium G870 | ||
Sandra Whetstone SSE2 (MFLOPS) benchmark | ||
| Longer is better | ||
| 100% | ||
| 72.9% | ||
XVid video encoding - multiple processes benchmark | ||
| Longer is better | ||
| 100% | ||
| 72.4% | ||
wPrime v1.55 (32M) benchmark | ||
| Shorter is better | ||
| 100% | ||
| 70.7% | ||
A6-3650 vs Intel Pentium G870 graphics benchmarks
Graphics benchmarks depend on the type of integrated or discrete graphics adapter, and to less extent on the processor performance. Because these benchmarks are synthetic, they may not truly represent gaming performance. However, they still can used to estimate whether one processor will perform faster or slower in games than another CPU.
AMD A6-3650 Intel Pentium G870 | ||
3DMark03 game score benchmark | ||
| Longer is better | ||
|---|---|---|
| 100% | ||
| 39.9% | ||
3DMark06 3DMark score benchmark | ||
| Longer is better | ||
| 100% | ||
| 48.0% | ||
3DMark2001 score benchmark | ||
| Longer is better | ||
| 100% | ||
| 59.5% | ||
CINEBENCH R10 OpenGL score benchmark | ||
| Longer is better | ||
| 100% | ||
| 37.3% | ||
PCMark05 Graphics score benchmark | ||
| Longer is better | ||
| 100% | ||
| 44.3% | ||
A6-3650 vs Intel Pentium G870 memory performance
Memory-intensive tests or programs move large amounts of data to/from memory, and they depend more on memory throughput and the size of on-chip caches, rather than on CPU integer/FP/SIMD performance.
AMD A6-3650 Intel Pentium G870 | ||
7Zip compressing/decompressing speed benchmark | ||
| Longer is better | ||
|---|---|---|
| 100% | ||
| 82.0% | ||
CrystalMark Memory benchmark | ||
| Longer is better | ||
| 100% | ||
| 97.0% | ||
PCMark05 Memory score benchmark | ||
| Longer is better | ||
| 100% | ||
| 78.0% | ||
PCMark2002 Memory score benchmark | ||
| Longer is better | ||
| 100% | ||
| 53.6% | ||
WinRAR compressing/decompressing speed benchmark | ||
| Longer is better | ||
| 100% | ||
| 75.6% | ||
CPUIDs
The table below compares two random CPUID records for AMD A6-3650 and Intel Pentium G870 microprocessors, that were submitted to our CPUID database. Please note that different steppings of Intel processors, identified by S-spec numbers or CPUID, may have slightly different features. For that reason the comparison table below only applies to CPUs with specific S-Spec number / CPUID.
| CPUID 1 | CPUID 2 | |
| Manufacturer | AMD | Intel |
|---|---|---|
| CPU Family | A6-Series | Pentium Dual-Core |
| Model / Processor Number | A6-3650 | G870 |
| Frequency | 1945 MHz | 3110 MHz |
| CWID version | 0.4 | 0.5 |
| Part number | AD3650WNZ43GX | CM8062307260115 |
| S-Spec / Stepping Code | DA 1124APM | SR057 |
| General information | ||
| Vendor | AuthenticAMD | GenuineIntel |
| Processor name (BIOS) | AMD A6-3650 APU with Radeon(tm) HD Graphics | Intel(R) Pentium(R) CPU G870 @ 3.10GHz |
| Cores | 4 | 2 |
| Logical processors | 4 | 2 |
| Processor type | Original OEM Processor | |
| CPUID signature | LN-B0 | 206A7 |
| Family | 300F10 | 6 (06h) |
| Model | 18 (012h) | 42 (02Ah) |
| Stepping | 1 (01h) | 7 (07h) |
| TLB/Cache details | 0 (00h) | 64-byte Prefetching Data TLB0: 2-MB or 4-MB pages, 4-way associative, 32 entries Data TLB: 4-KB Pages, 4-way set associative, 64 entries Instruction TLB: 4-KB Pages, 4-way set associative, 128 entries L2 TLB: 1-MB, 4-way set associative, 64-byte line size Shared 2nd-level TLB: 4 KB pages, 4-way set associative, 512 entries |
| Cache | ||
| L1 data: Associativity | 2-way set associative | 8-way set associative |
| L1 data: Comments | Direct-mapped | |
| L1 data: Line size | 64 bytes | |
| L1 data: Lines per tag | 1 | |
| L1 data: Size | 4 x 64 KB | 2 x 32 KB |
| L1 instruction: Associativity | 2-way set associative | 8-way set associative |
| L1 instruction: Comments | Direct-mapped | |
| L1 instruction: Line size | 64 bytes | |
| L1 instruction: Lines per tag | 1 | |
| L1 instruction: Size | 4 x 64 KB | 2 x 32 KB |
| L2: Associativity | 16-way set associative | 8-way set associative |
| L2: Comments | Non-inclusive|Direct-mapped | |
| L2: Line size | 64 bytes | |
| L2: Lines per tag | 1 | |
| L2: Size | 4 x 1 MB | 2 x 256 KB |
| L3: Associativity | 12-way set associative | |
| L3: Comments | Inclusive|Shared between all cores | |
| L3: Line size | 64 bytes | |
| L3: Size | 3 MB | |
| Instruction set extensions | ||
| 3DNow! | + | - |
| AMD extensions to MMX | + | - |
| Extensions to 3DNow! | + | - |
| MMX | + | |
| SSE | + | |
| SSE2 | + | |
| SSE3 | + | |
| SSE4.1 | - | + |
| SSE4.2 | - | + |
| SSE4A | + | - |
| SSSE3 | - | + |
| Additional instructions | ||
| Advanced Bit manipulation | + | - |
| CLFLUSH | + | |
| CMOV | + | |
| CMPXCHG16B | + | |
| CMPXCHG8B | + | |
| FXSAVE/FXRSTORE | + | |
| MONITOR/MWAIT | + | |
| PCLMULDQ | - | + |
| POPCNT | + | |
| PREFETCH/PREFETCHW | + | - |
| RDTSCP | + | |
| SKINIT / STGI support | + | - |
| SYSCALL/SYSRET | + | - |
| SYSENTER/SYSEXIT | + | |
| XSAVE / XRESTORE states | - | + |
| Major features | ||
| 64-bit / Intel 64 | + | |
| Enhanced SpeedStep | - | + |
| Intel Virtualization | - | + |
| NX bit/XD-bit | + | |
| On-chip Floating Point Unit | + | |
| PowerNow! / Cool'n'Quiet | + | - |
| Secure Virtual Machine (Virtualization) | + | - |
| Other features | ||
| 1 GB large page support | + | - |
| 100MHz multiplier control | + | - |
| 128-bit SSE instructions | + | - |
| 36-bit page-size extensions | + | |
| 64-bit debug store | - | + |
| Advanced programmable interrupt controller | + | |
| CPL qualified debug store | - | + |
| Core multi-processing legacy mode | + | - |
| Debug store | - | + |
| Debugging extensions | + | |
| Digital Thermal Sensor capability | - | + |
| Extended APIC space | + | - |
| Hardware P-state control | + | - |
| Hardware thermal control | + | - |
| Instruction based sampling | + | - |
| LAHF / SAHF support in 64-bit mode | + | |
| LBR virtualization | + | - |
| LOCK MOV CR0 means MOV CR8 | + | - |
| Machine check architecture | + | |
| Machine check exception | + | |
| Memory-type range registers | + | |
| Misaligned SSE mode | + | - |
| Model-specific registers | + | |
| Nested page tables | + | - |
| OS visible workaround | + | - |
| Page attribute table | + | |
| Page global extension | + | |
| Page-size extensions (4MB pages) | + | |
| Pending break enable | - | + |
| Perfmon and Debug capability | - | + |
| Physical address extensions | + | |
| Power Limit Notification capability | - | + |
| Process context identifiers | - | + |
| SVM lock | + | - |
| Self-snoop | - | + |
| Software thermal control | + | - |
| Support for NRIP save | + | - |
| THERMTRIP | + | - |
| TSC rate is ensured to be invariant across all states | + | |
| Temperature sensor | + | - |
| Thermal monitor | - | + |
| Thermal monitor 2 | - | + |
| Thermal monitor and software controlled clock facilities | - | + |
| Time stamp counter | + | |
| Timestamp counter deadline | - | + |
| Virtual 8086-mode enhancements | + | |
| Watchdog timer support | + | - |
| xTPR Update Control | - | + |
- - feature is not supported
Features, not supported by all processors in the table, are not displayed
Compare with ...
i5-3570K
Intel Celeron 2 GHz
Ivy Bridge
Pentium 4 800FSB
AMD 3.3GHz 8MB
| Terms and Conditions · Privacy Policy · Contact Us | (c) Copyright 2003 - 2010 Gennadiy Shvets |