AMD Sempron for Socket AM2
Tests
We ran the tests in several stages. Since the processor is
aimed at value systems, the motherboards that we used were chosen those
offering onboard video - let's see what acceleration of the memory
subsystem and video memory will give. Now move on to the external video.
Test configuration:
|
Motherboards
|
ASUS A8N-VM CSM on nForce 430 + GeForce
6150
ASUS M2NPV-VX on nForce 430 + GeForce 6150
|
|
Memory
|
2õ DDR-400 512 MB Hynix PC3200
2x DDR2-667 512 MB Hynix PC5300
|
|
HDD
|
Samsung HD080HJ (80 GB, 7200 rpm, 8 MB, SATA-300)
|
|
Power supply unit
|
SuperPower 480X, 480 W
|
|
Operating system
|
Microsoft Windows XP SP2
|
|
Drivers
|
NVIDIA ForceWare 91.31 WHQL
|
As can be seen from the first synthetic tests, the faster RAM
and increased size of the L2 cache don't make up for the reduced clock
speed. Now move on to the practical application of computing
capabilities - the faster, the better.
Again we can see a lack of the CPU clock speed - it is evident
that overclocking starts improving the situation. Moving on to gaming
synthetics and games.
There aren't any serious changes - the loss is in some ways
greater, somewhere it is less, but it is anyway a loss. AMD's rating
system glitches – reduction of the clock speed can hardly be
made up for without architectural changes. Anyway, since the new
processors were compared first by ratings versus the single-channel
Sempron for Socket 754 which are a bit slower than for Socket 939, we
can somehow justify the rating system.
Let's see how the result will change if we disable the
integrated video and enable external video – we install
Inno3D GeForce 6600GT 128 MB GDDR3.
At first glance, replacement of the video has not brought any
changes, but the system on the base of Sempron AM2 ran more confidently
in video tests.
Accelerated operating of graphic applications due to the use
of faster RAM is seen already in SPECViewperf 9.0 suite, but we need a
powerful video card.
Futuremark's 3DMark suites also demonstrate accelerated
handling of graphics, but they try to create a minimum load on the CPU,
unlike that in popular games.
Once the load on the CPU starts going up, the results quickly
tend to those previously produced.
But before we make final conclusions, let's investigate one
more aspect – the need of Sempron AM2 processors for the RAM
bandwidth. To this end, using the BIOS settings we turned DDR2-667 into
DDR2-533 and DDR2-400, with the latency timings set as per SPD.
DDR2-400
DDR2-533
DDR2-667
The most noticeable performance leap occurred due to
transition from DDR2-400 to DDR2-533. The former is already rarely
available in the retail, and the latter is not much cheaper than
DDR2-667. It turns out that if you are into saving without much loss,
the most optimum will be use of DDR2-533 modules.
Final Words
Having summed up all the test results, we note that we
expected more from Sempron CPUs for Socket AM2 but the miracle hasn't
worked and we got a bit less than that. AMD has overdone with ratings
– honestly, to produce a true comparable result, 100 units of
the rating for each model should have been deducted. Anyway, there is
no reason to worry especially if you are short for funds - there is no
drop in performance, the CPUs will allow for a comfortable work and
even play if a good video card is installed. At the same time, prices
for Sempron AM2 are traditionally affordable, faster and cheaper RAM is
used, and a system assembled on the base of Socket AM2 is open for
future upgrades.
We appreciate LLC PF "Servis"
(Dnepropetrovsk, Ukraine) for the processors and motherboards presented
for tests.
 |
Top Stories: |
 |
 |
|
MoBo:
| |
|
|
VGA Card:
|
|
|
|
CPU & Memory:
|
|
