AMD Sempron 180 mini-review
First rumors about new Sempron model appeared in March of this year. No specifications for this CPU were known at the time, and the only confirmed information about this processor was its model number - 180. Some sites used the model number to extrapolate specifications from known Sempron models 140 and 145, and estimated 3.5 GHz core frequency. More than a month ago tray version of the Sempron 180 popped up in several online stores, and we used accompanied information to determine CPU characteristics. According to the information and published part number, we concluded that the CPU has two cores, and runs at significantly lower frequency than extrapolated 3.5 GHz. At the beginning of September we were able to acquire one of these Semprons, benchmark it, retrieve its CPUID information, and confirm previously posted specifications.
As we already mentioned, the main difference of the Sempron 180 from previous Semprons with model numbers 140 and 145 is the addition of the second core. Both CPU cores on the Sempron 180 operate at 2.4 Ghz. This is a few hundred MHz lower than 2.7 or 2.8 GHz frequency of Semprons 140 and 145, but, with two operational cores, the Sempron 180 should still be faster in multi-threaded applications than single-core models. The total size of L2 cache on the 180 is 1 MB, and the cache is split between two cores. Sempron 180 supports Streaming SIMD Extensions up to SSE3 and SSE4a, 64-bit instructions, Disable execute bit, and Virtualization technology. The processor is manufactured in 938-pin micro-PGA package, works in socket AM3 motherboards, and it is compatible with some socket AM2+ motherboards as well.
To estimate processor's performance we compared it with the following budget AMD and Intel microprocessors:
For quick reference, we combined specifications of all tested processors in the table below. For detailed characteristics on each CPU please click on the processor name in the table:
Tests of socket AM3 microprocessors were performed on ASRock M3A780GXH/128M motherboard equipped with 2 GB of dual-channel Corsair CM3X1024-1333C9 DDR3 memory. Socket 775 benchmark results were collected on ASRock 4Core1600Twins-P35 motherboard with 2 GB of dual-channel Corsair CM3X1024-1333C9 DDR3 memory (same as Socket AM3 motherboard). Both socket AM3 and 775 systems used 32-bit version of Windows XP, and had exactly the same type of hard disk drive and video card installed. The results in graphs in this mini-review do not show absolute numbers returned by benchmarks. Instead, the graphs show results as a percentage, representing relative performance of tested processor to performance of Sempron 180 CPU.
Sandra Dhrystone (MIPS)
Sandra Dhrystone test, part of SiSoftware Sandra 2004 Service Pack 2 (SP2b) suite, runs integer performance test on all logical processors. The test runs three times, and the final test result is an average of the three results.
The test depends on: CPU frequency, Integer performance, The number of cores / threads
Sandra Dhrystone measures integer performance, and it scales well with CPU frequency. The test also scales very well with the number of cores, therefore it's not surprising to see that all dual-core CPUs are about twice as fast as single-core Semprons 140 and 145. The Dhrystone benchmark is not dependent on the size of L2 cache, which is why larger size of L2 cache on the Athlon II 240 does not provide any benefit over the 220 model. Sempron 180 shows 12.7% and 16.8% lower performance than Athlon II X2 215 and 220 CPUs respectively, which very closely matches 12.5% and 16.7% lower clock frequency of the Sempron processor. Celeron E3300 and E3400 are in the lead in this benchmark.
Sandra Whetstone (MFLOPS)
Sandra Whetstone test, part of SiSoftware Sandra 2004 Service Pack 2 (SP2b) suite, runs floating-point performance test on all logical processors. The values are measured three times, and the final value is an average of the three results.
The test depends on: CPU frequency, Floating-Point performance, The number of cores / threads
Sandra Whetstone measures floating point performance of the CPU. Like the Sandra Dhrystone test, it scales extremely well with frequency and the number of cores, thus the Sempron and Athlon II results are not different from the Dhrystone test. Again, the Sempron 180 is 12.7% and 16.9% slower than Athlons 215 and 220, which can be attributed only to Sempron's lower core frequency. Celerons perform worse in this test, but they still about 10% faster than the 180.
Super PI (1M)
Super Pi benchmark measures time required to calculate pi up to 1 million digits after the decimal point. The test uses only one core.
The results depend on: CPU frequency, Floating-Point performance, Size of caches
Super Pi benchmark exercises only one core, as a result single core Semprons with their higher core frequency perform as fast or faster than the Athlon II X2 220. The benchmark favors large cache size on single-core Sempron and Celeron CPUs, therefore Semprons 140 and 145 outperform similarly clocked Athlon IIs 215 and 220 by 3%. The Sempron 180 with its low 2.4 GHz frequency is the last in this benchmark, and this is typical for all other single-threaded benchmarks too.
wPrime v1.55 (32M)
wPrime benchmark measures the time taken to calculate square roots of numbers from 1 to 33554431. The program uses Newton's method for estimating functions.
The results depend on: CPU frequency, Floating-Point performance, The number of cores / threads
wPrime test does not scale with the size of L2 cache. This results in lower Celeron performance, and almost identical performance of 220 and 240 models. As expected in multi-threaded benchmarks, the single-cores are far behind in this test, and the Sempron 180 is 12% - 16% slower than the Athlon IIs.
CINEBENCH R10 CPU score
CINEBENCH R10 multi-threaded benchmark is based on 3D software CINEMA 4D. The benchmark renders photo-realistic image using all available CPU resources, and reports the render time.
The results depend on: CPU frequency, SIMD performance, The number of cores / threads, Size of caches, Memory bandwidth
CINEBENCH R10 measures SIMD performance, and it heavily depends on the frequency, the number of cores and the size of caches. As you can see from the graph, the results for Athlon II and Sempron CPUs are similar to previous multi-threaded benchmarks. Extra 1 MB L2 cache on the Athlon II 240 results in 4.5% performance boost over the 220. Celeron E3400 takes the top spot here, but it's only a few percent faster than the X2 240.
WinRAR compressing/decompressing speed
WinRAR compressing / decompressing benchmark reports compression / decompression speed, in KB per second, of random data by WinRAR archiver. The test is a part of WinRAR software.
The results depend on: CPU frequency, Integer performance, The number of cores / threads, Size of caches, Memory bandwidth.
WinRAR benchmark utilizes all available logical processors, therefore CPUs with larger number of cores have an advantage here. The test requires a lot of memory bandwidth, so having integrated memory controller and faster DDR3-1333 memory helps all Athlon IIs and Sempron 180 beat Celeron E3300 and E3400 in this test. This, perhaps, is the only type of test, where Sempron 180 outruns Celerons by a large margin.
H.264 video encoding - multiple processes
"H.264 encoding using multiple processes" benchmark measures the number of frames, converted by all logical processors, per second. The test runs multiple instances of MEncoder program (part of MPlayer software), one instance per each logical processor. Each instance converts the same movie snippet in MPEG2 format to a file in H.264 format using single-pass encoding. To avoid extra disk I/O the encoded file is copied to a NUL: device.
The results depend on: CPU frequency, SIMD performance, The number of cores / threads, Size of caches, Memory bandwidth.
"H.264 video encoding using multiple processes" is yet another multi-threaded benchmark that shows the same pattern, i.e. 15% - 16% higher score of the Athlon II 220 CPU, compared to Sempron 180, and 40% - 45% worse performance of single-core Semprons. Celeron E3400 in this test is only 8% faster than the 180. To be fair, the Celeron performs better in other media tests, and it's not uncommon for the E3400 to beat the Sempron 180 by 15% - 20%.
H.264 video encoding - single process
"H.264 encoding using single process" benchmark measures the number of frames, converted by one logical processor, per second. The test runs one instance of MEncoder program (part of MPlayer software), which converts the same movie snippet in MPEG2 format to a file in H.264 format using single-pass encoding. To avoid extra disk I/O the encoded file is copied to a NUL: device. The results depend on: CPU frequency, SIMD performance, Size of caches, Memory bandwidth.
"H.264 video encoding using single process" is the last benchmark that we included in our mini-review. Benchmark results are close to those of Super PI single-threaded test. Sempron 140 overtakes the 180 by 12%. Sempron 145 is slightly ahead of the dual-core Athlon II 220, and both are 15% - 16% faster than the Sempron 180. The only difference from the Super Pi benchmark is the performance of Celeron processor. In Super Pi test Celeron had 42% lead over the Sempron, which is unusual. In this test the Celeron has 11% higher performance, which is more common.
As expected, Sempron 180 offers better performance in multi-threaded applications than Sempron 140 and 145 CPUs, up to 35% - 45% in our benchmarks, and up to 16% lower performance in single-threaded apps. The 180 performed 12% - 17% worse in comparison to Athlon II X2 215 and 220 models, and it is up to 20% slower than the Athlon II 240. Compared to Celeron CPUs, the 180 may match their performance in one or two benchmarks, but on average it's 10% - 15% slower than Celerons, and in some tests it falls behind the E3300 and E3400 models by as much as 25% or 30% respectively.
At present, Sempron 180 is offered in various OEM systems - Compaq Presario CQ3211CX, Presario CQ3219CN and Presario CQ3230CX, just to name a few. Buying such system would make sense if the system is sold at very reasonable (read: low) price, and the system is used only for basic tasks, such as e-mail, browsing, word processing, etc. The Sempron 180 could be a better fit than single-core Semprons in situations when user runs many tasks at once, and frequently switches between them, or when increased system responsiveness is a must. Getting one of these systems is not recommended if you plan to run one or more CPU intensive applications on it.
OEM version of the Sempron 180 CPU is available in stock in numerous online stores, such as Provantage.com and Buy.com. Currently, the processor is priced from $50 and up before shipping is added. The best deal for this CPU is $50 and some cents with free shipping from Buy.com. In our opinion, even at the best price the OEM processor is overpriced. Much better deal would be getting boxed Athlon II X2 245 CPU from Newegg.com. For mere $8 extra you'll get up to 20% better performance, three years warranty from AMD and new fan/heatsink.
For detailed specification of the Sempron 180 CPU, it's CPUID information and more benchmarks, please visit AMD Sempron 180 specifications page.