VGA sangat penting
Tapi butuhnya sampai mana nih
Pendapatnya seperti ini deh
Pertama ,
Notebooknya buat apa, kalau sekedar buka internet, buka facebook dan kerja kecil, ngetik , cari aja netbook
kalau buat presentasi, vga perlu lebih kenceng. Cari versi Non IGP , tentu lebih mahal
Kecuali buat main game tua, kaya DOTA. Notebook tua juga jalan.
Kedua , layar.
Cari selebar mungkin, kalau udah mau migrasi dari desktop ke notebook.
Kalau udah biasa main di 17 inch, 14 inch notebook itu rasanya sempitttt banget
Apalagi yang widescreen.
Untung sih bisa make dual screen, nah notebok di konek ke layar LCD yang gede.
Ketiga, memory
Begitu beli , langsung pentokin ke 2GB
Buang Vista dan ganti ke XP.
Keempat
Begitu beli notebook, backup semua OS dan isi sistem dari harddisk ke Flash
Bikin ghost file buat recovery.
Pecah isi harddisk jadi 2 atau 3 partisi
INI yang penting
Jangan berharap notebook bisa mengantikan desktop kelas gamer
Harga notebook gamer ya hampir setara desktop, atau bakalan lebih mahal.
Dulu temen gue rajin bawa notebook, buat download dan disain kecil
Lama lama bete juga, apalagi pas beli 22 inch.
Tuh notebook cuma buat download dan cek email doang.
Kamis, 08 Oktober 2009
Rabu, 16 September 2009
Athlon II X4 620: AMD's new Quad-Core CPU reviewed
Low priced quad-core tested
Athlon II X4 620: AMD's new Quad-Core CPU reviewed
Up to now quad-core CPUs could be found in the upper class only. But with the Athlon II X4 AMD introduces a processor with for cores for less than 100 Euros. PC Games Hardware reviews the new X4 620.
Athlon II X4 620
Athlon II X4 620 [Source: view picture gallery]
More and more quad-core gaming PCs are based on quad-core processors, but for the mass market such CPUs are still too expensive: For Intel's smallest quad-core - the Core 2 Quad Q8200 - you have to pay about 115 Euros and the AMD counterpart - the Phenom II X4 810 - costs about the same.
But this situation has come to an end no: With the Athlon II X4 AMD introduces the quad-cores to the entry-level market. The first model of this series, the X4 with 2.6 gigahertz, is currently offered for about 85 Euros - a direct challenge to Intel, whose forthcoming Core i3 processors are intended for the same market sector but have only two physical cores plus SMT. Another new CPU is the Athlon II X3 with three cores.
Athlon II X4 in Detail
The Athlon II X4 (codename Porpus) has four physical cores each of which has 512 KiByte L2 cache. Thus the L2 cache of the cores is only half as big as it has been in the dual-core Athlon II X2 (2x 1,024 KiByte), even if the overall size is the same. The X3 offers 3x 512 KiByte L2 cache - all in all only 1.5 MiByte. Nevertheless The Athlon II X3 and X4, just like the X2, don't have L3 cache - the only difference between a Propus and a Phenom II (Deneb).
So it suggests itself that AMD sells Deneb processors with a defective L3 cache as Athlon II X4 - reports about successful reactivations of L3 cache in Athlon II X4s apparently confirm this thesis. But actually the X4 should be based on the Propus, an own chip design without L3 cache, what would make such unlocking attempts impossible. Unlocking L3 cache of our own sample of the X4 640 did not work - the imprint "AADAC” identifies the CPU as a Propus. It is most likely that only few shipped X4 are not real Propus processors so unlocking the L3 cache won't work in most cases.
With the Athlon II X3/X4 AMD's new array of 45 nanometer based products is complete. The chart below gives an overview.
The Athlon II X4 600 will be available in four versions from the power efficient 600e with 2.2 GHz and 45W TDP to the 630 with 2.8 GHz and 95W TDP. The first model on the market will be the 620 with 2.6 GHz, which we also had a sample of four our tests, while the other versions will follow later this year. The X3 series contains four models, too. From the 400e (2.2 GHz) up to the 435 (2.9 GHz). None of the Athlon II processors will be offered as Black Edition with an open multiplier, even if this had been the case with the predecessor Athlon X2 7750/7850 BE (Kuma).
Athlon II X4 620: AMD's new Quad-Core CPU reviewed
Up to now quad-core CPUs could be found in the upper class only. But with the Athlon II X4 AMD introduces a processor with for cores for less than 100 Euros. PC Games Hardware reviews the new X4 620.
Athlon II X4 620
Athlon II X4 620 [Source: view picture gallery]
More and more quad-core gaming PCs are based on quad-core processors, but for the mass market such CPUs are still too expensive: For Intel's smallest quad-core - the Core 2 Quad Q8200 - you have to pay about 115 Euros and the AMD counterpart - the Phenom II X4 810 - costs about the same.
But this situation has come to an end no: With the Athlon II X4 AMD introduces the quad-cores to the entry-level market. The first model of this series, the X4 with 2.6 gigahertz, is currently offered for about 85 Euros - a direct challenge to Intel, whose forthcoming Core i3 processors are intended for the same market sector but have only two physical cores plus SMT. Another new CPU is the Athlon II X3 with three cores.
Athlon II X4 in Detail
The Athlon II X4 (codename Porpus) has four physical cores each of which has 512 KiByte L2 cache. Thus the L2 cache of the cores is only half as big as it has been in the dual-core Athlon II X2 (2x 1,024 KiByte), even if the overall size is the same. The X3 offers 3x 512 KiByte L2 cache - all in all only 1.5 MiByte. Nevertheless The Athlon II X3 and X4, just like the X2, don't have L3 cache - the only difference between a Propus and a Phenom II (Deneb).
So it suggests itself that AMD sells Deneb processors with a defective L3 cache as Athlon II X4 - reports about successful reactivations of L3 cache in Athlon II X4s apparently confirm this thesis. But actually the X4 should be based on the Propus, an own chip design without L3 cache, what would make such unlocking attempts impossible. Unlocking L3 cache of our own sample of the X4 640 did not work - the imprint "AADAC” identifies the CPU as a Propus. It is most likely that only few shipped X4 are not real Propus processors so unlocking the L3 cache won't work in most cases.
With the Athlon II X3/X4 AMD's new array of 45 nanometer based products is complete. The chart below gives an overview.
The Athlon II X4 600 will be available in four versions from the power efficient 600e with 2.2 GHz and 45W TDP to the 630 with 2.8 GHz and 95W TDP. The first model on the market will be the 620 with 2.6 GHz, which we also had a sample of four our tests, while the other versions will follow later this year. The X3 series contains four models, too. From the 400e (2.2 GHz) up to the 435 (2.9 GHz). None of the Athlon II processors will be offered as Black Edition with an open multiplier, even if this had been the case with the predecessor Athlon X2 7750/7850 BE (Kuma).
Selasa, 15 September 2009
Mainstream Quad-Core CPU Performance Comparison
Looking back, quad-core processors have had quite the run already. Intel's first quad-core CPU, the Core 2 Extreme QX6700 was released almost three years ago, with the much more popular Q6600 following up months later at a rather steep $850. As you may recall it was quite the luxury then to have one of these at your disposal.
AMD came to market with their own quad-core solution a year later, marking the release of the first Phenom X4.Though in all fairness, it took them about about one more year to get things going right.
Today we find ourselves with very different and diversified offerings from both companies, that are not only cheaper but also significantly faster. Just take for example AMD's Phenom II X4 945 that can be purchased for as little as $170, not to mention Intel's most recent release, the Core i5 750, which is meant to crush its competitors offering top notch performance at the $199 price point.
With more powerful quad-core processors becoming mainstream, and with so many options currently available, we wanted to know which CPU provides users with the most value at under $300. That said, we won't just be evaluating the value of the individual processors, but also their accompanying platforms.
At the upper end of the scale we have the Intel Core i7 920 processor, which was just recently joined by the Core i7 860; both cost around $280. Although the Core i7 860 is clocked 140MHz higher, it is designed for the LGA1156 platform, and therefore lacks the triple-channel memory controller of the Core i7 920. Also, the Core i7 860 should make for a more affordable build, as P55 motherboards will be cheaper than comparable X58-based boards.
Moving down the ladder we have the Phenom II X4 965, which is the current AMD flagship processor. The Phenom II X4 965 operates at 3.40GHz, and with a retail price tag of $250, it’s quite a good value option. Also worth keeping in mind is the Phenom II X4 955, which is virtually the same processor at a 24% discount. Both are “Black Edition” processors featuring unlocked multipliers, so by simply increasing the multiplier of the X4 955 by one notch, you get the 3.40GHz of the Phenom II X4 965.
Finally we have the new Core i5 750 processor which costs just $200 and boasts similar specifications to those of the Core i7 920. The key difference between the two, besides the different socket/platforms, is that the Core i5 only features a dual-channel memory controller and Hyper-Threading is disabled. The Core i5 750 is designed for the new LGA1156 platform just like the Core i7 860, so it will be interesting to compare the performance of these two processors as well.
Test System Specs
LGA1366 System Specs
- Intel Core i7 920 (LGA1366)
- x3 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus P6T Deluxe (Intel X58)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7
Ultimate (64-bit)
- Nvidia Forceware 190.38
LGA1156 System Specs
- Intel Core i7 860 (LGA1156)
- Intel Core i5 750 (LGA1156)
- x2 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus P7P55D Deluxe (Intel P55)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7 Ultimate (64-bit)
- Nvidia Forceware 190.38
AM3 System Specs
- AMD Phenom II X4 965 (AM3)
- x2 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus M3A79-T Deluxe (AMD 790FX)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7 Ultimate (64-bit)
- Nvidia Forceware 190.38
Memory Bandwidth Performance
Surprisingly MaxxMem² reports very little difference in memory bandwidth between the Core i7 920 and the new Core i7 860 processor. Despite utilizing a triple-channel memory controller, the Core i7 920 produced the same write performance, while the read throughput was 5% higher. The Core i7 920 has a 140MHz clock speed disadvantage nonetheless.
The Core i5 750 processor, which does operate at the same frequency as the Core i7 920, only uses a dual-channel memory controller and does without Hyper-Threading. This reduced the write performance by 8.5% when compared to the 920, while read performance was 12% lower.
Although the AMD Phenom II X4 965 processor also features a dual-channel memory controller with a rather high operating frequency of 3.40GHz, it delivered by far the lowest memory bandwidth result. The read performance of the Phenom II X4 965 was 26.5% lower than that of the Core i5 750, making it considerably slower in this test when compared to the Intel competition.
The latency results follow the same trend as the bandwidth results, though as you can see the difference between the Phenom II X4 965 and Core i5 750 processors is much smaller here. The Core i7 920 is also only slightly better than the Core i7 860 in this test.
EVEREST Ultimate Edition shows slightly different performance margins when compared to MaxxMem². For instance, the Core i7 920 has a considerably higher read and write performance advantage over the Core i7 860 processor, but the margin separating the Core i7 860 and Core i5 750 is significantly smaller in this test. One thing did remain constant, and that was the huge margin by which the Phenom II X4 965 trails the Intel processors.
Synthetic Application Performance
The MaxxPI² benchmark is again very intriguing as the AMD Phenom II X4 965 processor cleans up in the multi-threading test, beating even the Core i7 920 processor with a calculation time of just 10.9 seconds. When running the single-thread test, the Phenom II X4 965 tanks with a 60.2 second calculation time, which doesn't seem to add up. Still, having run the test several times, this is the result we received each time.
AMD's Phenom II X4 965 was 12% faster than the Core i5 750 in the multi-threading test, 5.5% faster than the Core i7 860 and just 2% faster than the Core i7 920. Conversely, when measuring single threaded performance with the MaxxPI² benchmark, the Phenom II X4 965 delivered half the performance of the Core i5 750.
Taking a quick look at single-thread throughput in MaxxPI² we see how the Phenom II X4 965 falls way behind, as it managed just 136K/sec, whereas the Core i5 750 was capable of 236K/sec. However, something kicks the Phenom II X4 965 into gear when running the multi-thread test, as throughput increases to 748K/sec and beats the 736K/sec of the Core i7 920 processor.
WinRAR is another good program for measuring processors' single and multi-thread performance. Here we see that the AMD Phenom II X4 965 struggles regardless of whether we are using just one or multiple threads.
The Core i5 750 was 25% faster than the Phenom II X4 965 in the multi-thread test, while it was just 3% slower than the Core i7 860 processor, which is surprising given that it lacks Hyper-Threading support. Finally the Core i7 920 is able to make use of its additional bandwidth, as it was found to be 5% faster than the Core i7 86
Application Performance
Although the Phenom II X4 965 struggled in our synthetic tests, the first real-world test that we threw at it went very well. Using a complex spreadsheet that we created (not supplied by Intel or AMD) the Phenom II X4 965 completed the large calculation in 11:10 minutes. The Core i7 920 took 13:04 minutes, making it over a minute slower. There was very little difference in performance when comparing Intel CPUs to one another.
This time using WinRAR we simulate our own test by compressing two different archives. The first is the Windows 7 installer, which contains over 800 files and totals 3GB in size, and the second is a huge individual 12GB media file. Each test is run three times and we record the average result.
The Core i7 920 had a clear advantage over the dual-channel LGA1156 processors in this test, as it was 9% faster than the Core i7 860 and 10% faster than the Core i5 750. Interestingly, Hyper-Threading support on the Core i7 860 counted for very little here, as the chip was just a fraction faster than the slower clocked Core i5 750.
Finally we fired up Adobe Photoshop CS4, opened a 13MB (7000x6000) JPEG image file and proceeded to apply two filters, measuring the time each of them took. The extrude filter took each processor roughly 66 seconds, with the Core i7 920 delivering a marginally better result with a time of 65.5 seconds.
The Phenom II X4 965 was faster when applying the blur filter, taking just 41.2 seconds, followed by the Core i7 920 at 46.9 seconds.
Synthetic 3D Performance
FurMark results are boring to stare at, though we were surprised to see the Phenom II X4 965 dominating the minimum frames per second test with 119fps, compared to all three Intel processors that were limited to 101fps. The maximum frames per second were all capped at 186fps, clearly showing our GeForce GTX 285 was acting as the bottleneck.
Using the Unigine Engine, we ran the Sanctuary and Tropics benchmarks. The Phenom II X4 965 fares well producing the highest score in both tests, though margins were insignificant across the board.
Unlike the previous 3D synthetic benchmarks, SPECviewperf is more CPU dependent, and therefore we do see a bit of a difference when comparing these four processors.
The Maya test had the Core i5 750 averaging 40.3fps, or half a frame faster than the Core i7 920 and 860 processors, while the Phenom CPU was a tad slower.
The 3dsMax test showed more homogeneous results with very small margins between processors.
Gaming Performance
Unreal Tournament 3 was tested at two different resolutions, though as we have seen time and time again at 1920x1200 you are almost certainly guaranteed to be GPU limited rather than CPU limited. At 1024x768 we did see a 10fps difference between the Phenom II X4 965 and the Core i7 920, a margin which favored the Intel processor.
In Street Fighter IV, frame rates were locked at around 166–167fps when using a 1920x1200 resolution, while at 1024x768 the Intel Core i7 920 managed to outperform AMD's Phenom by an 11fps margin.
Far Cry 2 is a game that we have always known for being quite demanding on the CPU, which is unusual for a first person shooter. This time at 1920x1200 there is 8fps separating the Phenom II X4 965 and Core i7 920 processors, while a 12fps margin can be seen at 1024x768.
Although the Core i5 750 processor was just 1fps faster than the Phenom II X4 965 at 1920x1200, it got away to a 9fps lead at 1024x768.
Power Consumption
We were impressed by the LGA1156 platform efficiency when we tested the Intel Core i5 750 for our review earlier this week. This graph brings some additional information as we have a second CPU using the exact same platform.
At idle the power consumption difference between the Core i7 860 and Core i5 750 CPUs is negligible, while at full load the Core i7 uses more power but stays way ahead of the Core i7 920 and its X58-based motherboard.
All in all, the Core i5 750 system used just 178 watts of power under full load which translated to 10% less power than the Core i7 860, and an incredible 34% less than the Core i7 920.
By comparison the Phenom II X4 965 is quite power hungry, consuming 41% more power than the Core i5 750, while in many cases failing to deliver the same level of performance.
The Best Mainstream CPU is...
Before we are able to determine which processor offers the most value, we must work out roughly how much each platform costs to put together.
The most expensive is obviously going to be the Core i7 920 ($280) with its accompanying X58 motherboards retailing anywhere from $170 (Foxconn FlamingBlade GTI, ASRock X58 Extreme and MSI X58M) while most cost around $220 - $250. Add to that some memory - most will go for a 6GB DDR3 kit which costs about $140 for something decent - for a total platform cost of ~$640.
In the past we have built entire budget systems for that kind of money. On the other hand, until just recently this was the best or only the way to get yourself a powerful Core i7 system on a budget. Given the performance this combination of hardware delivers, we don't hear anyone complaining thus far.
The newly launched Core i7 860 will also be priced at $280, but supporting P55 motherboards should be priced below the $200 mark, or in some extreme cases nearing the $100 range, making them a more affordable choice. Rather that 6GB of memory you must opt between 4GB or 8GB on a dual-channel platform. A good quality DDR3 4GB kit will cost you $80, making for a total of ~$500.
Also from Intel the new Core i5 750 is priced at just $200, and from the benchmarks you just saw you know it makes for an incredibly good buy. The Core i5 750 is supported by the same P55 motherboards as the Core i7 860, so expect to pay anywhere from $100 to $180 for the motherboard and $80 for a decent DDR3 4GB dual-channel memory kit. This brings the total platform cost at under $450.
Lastly we have the AMD Phenom II X4 965 which costs $245. In all fairness however, AMD hasn't had the time yet to react to Intel's CPU launch this week and so we expect them to adjust their CPU pricing accordingly. AMD will not only be forced to decrease the price of the flagship Phenom II X4 965, but we expect them to drop many other Phenom models in the process as there's no longer enough room to offer so many different models in the $100-200 range.
There is a massive range of motherboards to choose from supporting Phenom II X4 CPUs, and we feel that most users will seek out a board with the AMD 790FX chipset and AM3 socket (~$180). Together with 4GB of DDR3 memory ($80), the total cost for a Phenom II X4 965 equipped system comes to $500, and $450 for the slightly slower Phenom II X4 955.
Current pricing makes the Phenom II X4 a terrible choice. At $500 against the Core i7 860, it is a bad match up for AMD. And while it's hard to predict where exactly AMD will set their prices a few weeks from now, we know they have more value offerings coming very soon. The Phenom II X4 remains a pretty competent CPU but with such harsh competition from the new Intel CPUs, the X4 965 would have to drop at or below the $170 range to make it a worthwhile choice against the Core i5 750. And even at that price, it's hard not to pay a little extra for the Core i5.
The next question is whether you should spend the extra $80 on the Core i7 860 processor or save some of that money to get a better motherboard with the Core i5 750, or double your RAM. The Core i7 860 has a 140MHz frequency advantage and the ability to use Hyper-Threading. However, many of our benchmarks showed that this translated into very small performance gains. Ultimately those of you working with programs that use lots of threads will benefit more from the Core i7 860, as will users that heavily multi-task. For the rest of us that just do some light word processing, e-mailing and net surfing, when we are not gaming, then the Core i5 750 will keep you more than happy. Actually that is probably not fair, as the Core i5 750 will hang with the Core i7 860 in quite a few powerful applications. Then, when you are gaming, the Core i5 750 proved to be every bit as good as the Core i7 860.
The Core i7 920 and 860 comparison is an interesting one, as the Core i7 860 wasn't a great deal slower than the almighty Core i7 920. The only real difference between these two processors is the memory controller, and most applications don't seem need such huge amounts of bandwidth, and I guess the same goes for most users. Therefore, it does appear that the savings users can enjoy by going for the cheaper Core i7 860 setup are worth it.
Honestly, I never thought a few weeks ago I would be recommending two LGA1156 processors over anything the LGA1366 and AM3 platforms had to offer, but I am.
The Core i5 750 is the ultimate value quad-core processor. Then for those that require something a little more heavy duty, the Core i7 860 appears to be the perfect solution, with its ability to use Hyper-Threading technology.
AMD came to market with their own quad-core solution a year later, marking the release of the first Phenom X4.Though in all fairness, it took them about about one more year to get things going right.
Today we find ourselves with very different and diversified offerings from both companies, that are not only cheaper but also significantly faster. Just take for example AMD's Phenom II X4 945 that can be purchased for as little as $170, not to mention Intel's most recent release, the Core i5 750, which is meant to crush its competitors offering top notch performance at the $199 price point.
With more powerful quad-core processors becoming mainstream, and with so many options currently available, we wanted to know which CPU provides users with the most value at under $300. That said, we won't just be evaluating the value of the individual processors, but also their accompanying platforms.
At the upper end of the scale we have the Intel Core i7 920 processor, which was just recently joined by the Core i7 860; both cost around $280. Although the Core i7 860 is clocked 140MHz higher, it is designed for the LGA1156 platform, and therefore lacks the triple-channel memory controller of the Core i7 920. Also, the Core i7 860 should make for a more affordable build, as P55 motherboards will be cheaper than comparable X58-based boards.
Moving down the ladder we have the Phenom II X4 965, which is the current AMD flagship processor. The Phenom II X4 965 operates at 3.40GHz, and with a retail price tag of $250, it’s quite a good value option. Also worth keeping in mind is the Phenom II X4 955, which is virtually the same processor at a 24% discount. Both are “Black Edition” processors featuring unlocked multipliers, so by simply increasing the multiplier of the X4 955 by one notch, you get the 3.40GHz of the Phenom II X4 965.
Finally we have the new Core i5 750 processor which costs just $200 and boasts similar specifications to those of the Core i7 920. The key difference between the two, besides the different socket/platforms, is that the Core i5 only features a dual-channel memory controller and Hyper-Threading is disabled. The Core i5 750 is designed for the new LGA1156 platform just like the Core i7 860, so it will be interesting to compare the performance of these two processors as well.
Test System Specs
LGA1366 System Specs
- Intel Core i7 920 (LGA1366)
- x3 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus P6T Deluxe (Intel X58)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7
Ultimate (64-bit)
- Nvidia Forceware 190.38
LGA1156 System Specs
- Intel Core i7 860 (LGA1156)
- Intel Core i5 750 (LGA1156)
- x2 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus P7P55D Deluxe (Intel P55)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7 Ultimate (64-bit)
- Nvidia Forceware 190.38
AM3 System Specs
- AMD Phenom II X4 965 (AM3)
- x2 Kingston HyperX 2GB DDR3-1333 (CAS 8-8-8-24)
- Asus M3A79-T Deluxe (AMD 790FX)
- OCZ GameXStream (700 watt)
- Seagate 500GB 7200-RPM (Serial ATA300)
- Asus GeForce GTX 285 (1GB)
Software
- Microsoft Windows 7 Ultimate (64-bit)
- Nvidia Forceware 190.38
Memory Bandwidth Performance
Surprisingly MaxxMem² reports very little difference in memory bandwidth between the Core i7 920 and the new Core i7 860 processor. Despite utilizing a triple-channel memory controller, the Core i7 920 produced the same write performance, while the read throughput was 5% higher. The Core i7 920 has a 140MHz clock speed disadvantage nonetheless.
The Core i5 750 processor, which does operate at the same frequency as the Core i7 920, only uses a dual-channel memory controller and does without Hyper-Threading. This reduced the write performance by 8.5% when compared to the 920, while read performance was 12% lower.
Although the AMD Phenom II X4 965 processor also features a dual-channel memory controller with a rather high operating frequency of 3.40GHz, it delivered by far the lowest memory bandwidth result. The read performance of the Phenom II X4 965 was 26.5% lower than that of the Core i5 750, making it considerably slower in this test when compared to the Intel competition.
The latency results follow the same trend as the bandwidth results, though as you can see the difference between the Phenom II X4 965 and Core i5 750 processors is much smaller here. The Core i7 920 is also only slightly better than the Core i7 860 in this test.
EVEREST Ultimate Edition shows slightly different performance margins when compared to MaxxMem². For instance, the Core i7 920 has a considerably higher read and write performance advantage over the Core i7 860 processor, but the margin separating the Core i7 860 and Core i5 750 is significantly smaller in this test. One thing did remain constant, and that was the huge margin by which the Phenom II X4 965 trails the Intel processors.
Synthetic Application Performance
The MaxxPI² benchmark is again very intriguing as the AMD Phenom II X4 965 processor cleans up in the multi-threading test, beating even the Core i7 920 processor with a calculation time of just 10.9 seconds. When running the single-thread test, the Phenom II X4 965 tanks with a 60.2 second calculation time, which doesn't seem to add up. Still, having run the test several times, this is the result we received each time.
AMD's Phenom II X4 965 was 12% faster than the Core i5 750 in the multi-threading test, 5.5% faster than the Core i7 860 and just 2% faster than the Core i7 920. Conversely, when measuring single threaded performance with the MaxxPI² benchmark, the Phenom II X4 965 delivered half the performance of the Core i5 750.
Taking a quick look at single-thread throughput in MaxxPI² we see how the Phenom II X4 965 falls way behind, as it managed just 136K/sec, whereas the Core i5 750 was capable of 236K/sec. However, something kicks the Phenom II X4 965 into gear when running the multi-thread test, as throughput increases to 748K/sec and beats the 736K/sec of the Core i7 920 processor.
WinRAR is another good program for measuring processors' single and multi-thread performance. Here we see that the AMD Phenom II X4 965 struggles regardless of whether we are using just one or multiple threads.
The Core i5 750 was 25% faster than the Phenom II X4 965 in the multi-thread test, while it was just 3% slower than the Core i7 860 processor, which is surprising given that it lacks Hyper-Threading support. Finally the Core i7 920 is able to make use of its additional bandwidth, as it was found to be 5% faster than the Core i7 86
Application Performance
Although the Phenom II X4 965 struggled in our synthetic tests, the first real-world test that we threw at it went very well. Using a complex spreadsheet that we created (not supplied by Intel or AMD) the Phenom II X4 965 completed the large calculation in 11:10 minutes. The Core i7 920 took 13:04 minutes, making it over a minute slower. There was very little difference in performance when comparing Intel CPUs to one another.
This time using WinRAR we simulate our own test by compressing two different archives. The first is the Windows 7 installer, which contains over 800 files and totals 3GB in size, and the second is a huge individual 12GB media file. Each test is run three times and we record the average result.
The Core i7 920 had a clear advantage over the dual-channel LGA1156 processors in this test, as it was 9% faster than the Core i7 860 and 10% faster than the Core i5 750. Interestingly, Hyper-Threading support on the Core i7 860 counted for very little here, as the chip was just a fraction faster than the slower clocked Core i5 750.
Finally we fired up Adobe Photoshop CS4, opened a 13MB (7000x6000) JPEG image file and proceeded to apply two filters, measuring the time each of them took. The extrude filter took each processor roughly 66 seconds, with the Core i7 920 delivering a marginally better result with a time of 65.5 seconds.
The Phenom II X4 965 was faster when applying the blur filter, taking just 41.2 seconds, followed by the Core i7 920 at 46.9 seconds.
Synthetic 3D Performance
FurMark results are boring to stare at, though we were surprised to see the Phenom II X4 965 dominating the minimum frames per second test with 119fps, compared to all three Intel processors that were limited to 101fps. The maximum frames per second were all capped at 186fps, clearly showing our GeForce GTX 285 was acting as the bottleneck.
Using the Unigine Engine, we ran the Sanctuary and Tropics benchmarks. The Phenom II X4 965 fares well producing the highest score in both tests, though margins were insignificant across the board.
Unlike the previous 3D synthetic benchmarks, SPECviewperf is more CPU dependent, and therefore we do see a bit of a difference when comparing these four processors.
The Maya test had the Core i5 750 averaging 40.3fps, or half a frame faster than the Core i7 920 and 860 processors, while the Phenom CPU was a tad slower.
The 3dsMax test showed more homogeneous results with very small margins between processors.
Gaming Performance
Unreal Tournament 3 was tested at two different resolutions, though as we have seen time and time again at 1920x1200 you are almost certainly guaranteed to be GPU limited rather than CPU limited. At 1024x768 we did see a 10fps difference between the Phenom II X4 965 and the Core i7 920, a margin which favored the Intel processor.
In Street Fighter IV, frame rates were locked at around 166–167fps when using a 1920x1200 resolution, while at 1024x768 the Intel Core i7 920 managed to outperform AMD's Phenom by an 11fps margin.
Far Cry 2 is a game that we have always known for being quite demanding on the CPU, which is unusual for a first person shooter. This time at 1920x1200 there is 8fps separating the Phenom II X4 965 and Core i7 920 processors, while a 12fps margin can be seen at 1024x768.
Although the Core i5 750 processor was just 1fps faster than the Phenom II X4 965 at 1920x1200, it got away to a 9fps lead at 1024x768.
Power Consumption
We were impressed by the LGA1156 platform efficiency when we tested the Intel Core i5 750 for our review earlier this week. This graph brings some additional information as we have a second CPU using the exact same platform.
At idle the power consumption difference between the Core i7 860 and Core i5 750 CPUs is negligible, while at full load the Core i7 uses more power but stays way ahead of the Core i7 920 and its X58-based motherboard.
All in all, the Core i5 750 system used just 178 watts of power under full load which translated to 10% less power than the Core i7 860, and an incredible 34% less than the Core i7 920.
By comparison the Phenom II X4 965 is quite power hungry, consuming 41% more power than the Core i5 750, while in many cases failing to deliver the same level of performance.
The Best Mainstream CPU is...
Before we are able to determine which processor offers the most value, we must work out roughly how much each platform costs to put together.
The most expensive is obviously going to be the Core i7 920 ($280) with its accompanying X58 motherboards retailing anywhere from $170 (Foxconn FlamingBlade GTI, ASRock X58 Extreme and MSI X58M) while most cost around $220 - $250. Add to that some memory - most will go for a 6GB DDR3 kit which costs about $140 for something decent - for a total platform cost of ~$640.
In the past we have built entire budget systems for that kind of money. On the other hand, until just recently this was the best or only the way to get yourself a powerful Core i7 system on a budget. Given the performance this combination of hardware delivers, we don't hear anyone complaining thus far.
The newly launched Core i7 860 will also be priced at $280, but supporting P55 motherboards should be priced below the $200 mark, or in some extreme cases nearing the $100 range, making them a more affordable choice. Rather that 6GB of memory you must opt between 4GB or 8GB on a dual-channel platform. A good quality DDR3 4GB kit will cost you $80, making for a total of ~$500.
Also from Intel the new Core i5 750 is priced at just $200, and from the benchmarks you just saw you know it makes for an incredibly good buy. The Core i5 750 is supported by the same P55 motherboards as the Core i7 860, so expect to pay anywhere from $100 to $180 for the motherboard and $80 for a decent DDR3 4GB dual-channel memory kit. This brings the total platform cost at under $450.
Lastly we have the AMD Phenom II X4 965 which costs $245. In all fairness however, AMD hasn't had the time yet to react to Intel's CPU launch this week and so we expect them to adjust their CPU pricing accordingly. AMD will not only be forced to decrease the price of the flagship Phenom II X4 965, but we expect them to drop many other Phenom models in the process as there's no longer enough room to offer so many different models in the $100-200 range.
There is a massive range of motherboards to choose from supporting Phenom II X4 CPUs, and we feel that most users will seek out a board with the AMD 790FX chipset and AM3 socket (~$180). Together with 4GB of DDR3 memory ($80), the total cost for a Phenom II X4 965 equipped system comes to $500, and $450 for the slightly slower Phenom II X4 955.
Current pricing makes the Phenom II X4 a terrible choice. At $500 against the Core i7 860, it is a bad match up for AMD. And while it's hard to predict where exactly AMD will set their prices a few weeks from now, we know they have more value offerings coming very soon. The Phenom II X4 remains a pretty competent CPU but with such harsh competition from the new Intel CPUs, the X4 965 would have to drop at or below the $170 range to make it a worthwhile choice against the Core i5 750. And even at that price, it's hard not to pay a little extra for the Core i5.
The next question is whether you should spend the extra $80 on the Core i7 860 processor or save some of that money to get a better motherboard with the Core i5 750, or double your RAM. The Core i7 860 has a 140MHz frequency advantage and the ability to use Hyper-Threading. However, many of our benchmarks showed that this translated into very small performance gains. Ultimately those of you working with programs that use lots of threads will benefit more from the Core i7 860, as will users that heavily multi-task. For the rest of us that just do some light word processing, e-mailing and net surfing, when we are not gaming, then the Core i5 750 will keep you more than happy. Actually that is probably not fair, as the Core i5 750 will hang with the Core i7 860 in quite a few powerful applications. Then, when you are gaming, the Core i5 750 proved to be every bit as good as the Core i7 860.
The Core i7 920 and 860 comparison is an interesting one, as the Core i7 860 wasn't a great deal slower than the almighty Core i7 920. The only real difference between these two processors is the memory controller, and most applications don't seem need such huge amounts of bandwidth, and I guess the same goes for most users. Therefore, it does appear that the savings users can enjoy by going for the cheaper Core i7 860 setup are worth it.
Honestly, I never thought a few weeks ago I would be recommending two LGA1156 processors over anything the LGA1366 and AM3 platforms had to offer, but I am.
The Core i5 750 is the ultimate value quad-core processor. Then for those that require something a little more heavy duty, the Core i7 860 appears to be the perfect solution, with its ability to use Hyper-Threading technology.
Jumat, 11 September 2009
Highest-end gaming computer
Fujitsu Celsius Ultra: World's fastest gaming PC on air cooling
The Clesius Ultra from Fujitsu is said to be the fastest, air cooled gaming system that has been released up to now. PC Games Hardware got a review sample and delivers specifications as well as pictures of the rig.
Up to now there only had been rumors and a video of the PC that is called Celsius Ultra. Only an SLI sticker on the case gave a slight idea of the components used for the system. But now Fujitsu provided PC Games Hardware with a sample and we take a look at the specifications of this PC that is said to be the World's fastest gaming system on air cooling.
The two Evga Geforce GTX 295 are linked in a 4-way SLI setup and have the new single PCB design. Although the graphics cards have stock cooling the frequencies have been increased slightly.
Manufacturer : Fujitsu
Product : Celsius Ultra
CPU : Intel Core i7 Extreme 975
Graphics card : 2x EVGA Geforce GTX 295 Ultra Clocked
Motherboard : FTS D2778-X1, Chipset: Intel X58
Memory : 6 x Samsung 2,048 DDR3-1333
PSU : 1,000 W
Hard Drive : 1x WD Veloci Raptor 150 GByte and 10,000 rpm, 2x Seagate 1 TByte and 7,200 rpm
Sound device : Creative Titanium X-Fi
Optical drive : Blu-Ray drive + DVD writer combo (Multi Format)
Operating system : Windows Vista Business 64 Bit
According to Fujitsu the Celsius Ultra will become available after the Gamescon for 3,999 Euros in retail stores and in Fujitsu's online shop. For more information see the Celsius website.
There's a few picture that you can see above :
The Clesius Ultra from Fujitsu is said to be the fastest, air cooled gaming system that has been released up to now. PC Games Hardware got a review sample and delivers specifications as well as pictures of the rig.
Up to now there only had been rumors and a video of the PC that is called Celsius Ultra. Only an SLI sticker on the case gave a slight idea of the components used for the system. But now Fujitsu provided PC Games Hardware with a sample and we take a look at the specifications of this PC that is said to be the World's fastest gaming system on air cooling.
The two Evga Geforce GTX 295 are linked in a 4-way SLI setup and have the new single PCB design. Although the graphics cards have stock cooling the frequencies have been increased slightly.
Manufacturer : Fujitsu
Product : Celsius Ultra
CPU : Intel Core i7 Extreme 975
Graphics card : 2x EVGA Geforce GTX 295 Ultra Clocked
Motherboard : FTS D2778-X1, Chipset: Intel X58
Memory : 6 x Samsung 2,048 DDR3-1333
PSU : 1,000 W
Hard Drive : 1x WD Veloci Raptor 150 GByte and 10,000 rpm, 2x Seagate 1 TByte and 7,200 rpm
Sound device : Creative Titanium X-Fi
Optical drive : Blu-Ray drive + DVD writer combo (Multi Format)
Operating system : Windows Vista Business 64 Bit
According to Fujitsu the Celsius Ultra will become available after the Gamescon for 3,999 Euros in retail stores and in Fujitsu's online shop. For more information see the Celsius website.
There's a few picture that you can see above :
Selasa, 25 Agustus 2009
PhotoFast G-Monster CFast Gen1
Compact Flash dari PhotoFast G-Monster CFast Gen1 menawarkan kecepatan diatas standar Compact Flash lain.
Memory card yang satu ini dibuat dengan kombinasi CF dan ATA serial transport (AST). Didalam memory card PhotoFast G-Monster CFast Gen1 sudah dipasangkan SATA Controller dengan 7+17 pin konektor dimana 7 pin untuk signal dan 17 pin untuk power dan controller konektor.
PhotoFast CFast Gen1 card
Read: up to 100 MB/sec
Write: up to 50 MB/sec
Features :
Available in 8GB / 16GB / 32GB
Read: up to 100 MB/sec
Write: up to 50 MB/sec
Average Seek time: <0.2 ~ 0.3ms
Supports enhanced endurance by dynamic/static wear-leveling
Supports Reed-Solomon ECC 6/8/10 bytes in 512 data bytes
Supports dynamic power management
Supports SMART (Self-Monitoring, Analysis and Reporting Technology)
Data integrity under power-cycling
Low Power Consumption
O/S Support: Windows 2000/XP/Vista/Linux/Mac OSX
Fully Compliant with RoHS directive
CE and FCC compatibility
Physical Specifications:
Connector / Interface: CFast
SATA-II speeds with MLC Flash
Capacity: 8GB - 32GB
Dimensions: 36.40mm x 42.80mm x 3.30mm (the same size as a CF Type 1 card)
Weight: 13-25g
Power Supply: 3.3VDC +/-5%
MTBF: > 1,000,000 Hours
Environmental Specifications:
Shock (operating): 1500G
Vibration (operating): 16G
Altitude (operating): 120,000 ft
Operating Temperature: 0°C ~ +70°C
Storage Temperature: -25°C ~ +85°C
Memory card yang satu ini dibuat dengan kombinasi CF dan ATA serial transport (AST). Didalam memory card PhotoFast G-Monster CFast Gen1 sudah dipasangkan SATA Controller dengan 7+17 pin konektor dimana 7 pin untuk signal dan 17 pin untuk power dan controller konektor.
PhotoFast CFast Gen1 card
Read: up to 100 MB/sec
Write: up to 50 MB/sec
Features :
Available in 8GB / 16GB / 32GB
Read: up to 100 MB/sec
Write: up to 50 MB/sec
Average Seek time: <0.2 ~ 0.3ms
Supports enhanced endurance by dynamic/static wear-leveling
Supports Reed-Solomon ECC 6/8/10 bytes in 512 data bytes
Supports dynamic power management
Supports SMART (Self-Monitoring, Analysis and Reporting Technology)
Data integrity under power-cycling
Low Power Consumption
O/S Support: Windows 2000/XP/Vista/Linux/Mac OSX
Fully Compliant with RoHS directive
CE and FCC compatibility
Physical Specifications:
Connector / Interface: CFast
SATA-II speeds with MLC Flash
Capacity: 8GB - 32GB
Dimensions: 36.40mm x 42.80mm x 3.30mm (the same size as a CF Type 1 card)
Weight: 13-25g
Power Supply: 3.3VDC +/-5%
MTBF: > 1,000,000 Hours
Environmental Specifications:
Shock (operating): 1500G
Vibration (operating): 16G
Altitude (operating): 120,000 ft
Operating Temperature: 0°C ~ +70°C
Storage Temperature: -25°C ~ +85°C
Radeon seri 5000 bulan depan
2 VGA baru lagi akan muncul dari AMD, dengan teknologi DirectX 11
Dari VRzone mengatakan HD seri 5000 akan tampil di bulan September 2009.
Ada 2 VGA baru ditampilkan oleh AMD yaitu HD 5870 dan HD 5850. Tidak dijelaskan detailnya lebih lanjut, tetapi kedua VGA memiliki 1GB GDDR5.
* 1GB GDDR5 memory
* ATI Eyefinity technology with support for up to three displays
* ATI Stream technology,
* Designed for DirectCompute 5.0 and OpenCL
* Accelerated Video Transcoding (AVT)
* Compliant with DirectX 11 and earlier revisions
* Supports OpenGL 3.1
* ATI CrossFireX multi-GPU support for highly scalable performance6
* ATI Avivo HD video and display technology
* Dynamic power management with ATI PowerPlay technology
* DL-DVI, DL-DVI, DisplayPort, HDMI
* PCI Express 2.0 support
Dari VRzone mengatakan HD seri 5000 akan tampil di bulan September 2009.
Ada 2 VGA baru ditampilkan oleh AMD yaitu HD 5870 dan HD 5850. Tidak dijelaskan detailnya lebih lanjut, tetapi kedua VGA memiliki 1GB GDDR5.
* 1GB GDDR5 memory
* ATI Eyefinity technology with support for up to three displays
* ATI Stream technology,
* Designed for DirectCompute 5.0 and OpenCL
* Accelerated Video Transcoding (AVT)
* Compliant with DirectX 11 and earlier revisions
* Supports OpenGL 3.1
* ATI CrossFireX multi-GPU support for highly scalable performance6
* ATI Avivo HD video and display technology
* Dynamic power management with ATI PowerPlay technology
* DL-DVI, DL-DVI, DisplayPort, HDMI
* PCI Express 2.0 support
Jumat, 21 Agustus 2009
ATI Power Saving dengan Overdrive
Power VGA terbaru membutuhkan arus DC besar, adakah cara untuk menghemat pemakaian power VGA. Produsen VGA boleh saja mengatakan bahwa mereka mengunakan chip dengan teknologi lebih kecil. Selain menghemat biaya produsksi, memakai ukuran silikon lebih kecil akan dapat menekan pemakaian power tanpa meninggalkan kecepatan VGA itu sendiri
Tetapi seberapa besar pemakaian power dapat ditekan. Apakah kita harus menerima produk mereka begitu saja. Seberapa besar kemampuan VGA yang dibutuhkan. Atau ada trik tertentu agar konsumsi power dapat lebih dihemat.
Antara 2D dan 3D memiliki dampak berbeda
Kapan sebuah VGA mengkonsumsi power paling besar. Pastinya ketika VGA bekerja pada aplikasi gaming, proses 3D akan dilakukan penuh oleh GPU dan pemakaian power yang dibutuhkan juga maksimal.
Produsen VGA berupaya agar GPU mereka tidak terlalu banyak mengkonsumsi power. Seperti produk ATI memberikan 2 kecepatan antara 2D dan 3D. Contoh saja pada test pengujian dengan PowerColor Radeon HD 4850, GPU ATI 4850 memiliki 2 kecepatan berbeda pada kodisi 2D dan 3D.
Ketika VGA digunakan untuk aplikasi biasa, seperti kinerja office dengan kinerja 2D. Maka VGA akan menurunkan kecepatan clock GPU lebih rendah dibandingkan kecepatan 3D.Untuk 2D misalnya kinerja GPU hanya bekerja pada speed 500Mhz. Sedangkan kinerja GPU pada aplikasi 3D seperti gaming, maka VGA akan meningkat pada kecepatan 625Mhz.
Tentu menjadi pertanyaan bagi anda, misalnya apakah konsumsi power sebuah VGA PowerColor Radeon HD 4850 dapat diturunkan. Jawabnya tentu saja bisa. Tetapi dengan sedikit pengorbanan pada performa untuk menekan pemakaian power VGA.
Caranya dibawah ini
Anda dapat masuk ke bagian control VGA atau Catalyst Control Center. Dan clik bagian ATI Overdrive, disana akan informasi untuk kinerja GPU seperti kecepatan memory dan GPU sampai suhu dan putaran fan yang terjadi.
PowerColor Radeon HD 4850 memiliki 2 kecepatan berbeda. Ketika proses 2D ditangani oleh GPU maka kinerja GPU menjadi 500Mhz dengan kecepatan memory 993Mhz. ketika kinerja GPU menangani 3D seperti gaming, GPU akan naik menjadi 625Mhz dengan kecepatan memory tetap yaitu 993Mhz. Sehingga untuk proses 3D akan mengkonsumsi power lebih besar.
Apa yang dapat dilakukan untuk menghemat power pada VGA PowerColor Radeon HD 4850. Caranya dengan memainkan kinerja clock GPU dan memory.
Tetapi tidak disarankan menurunkan kecepatan putaran fan. Mengingat kinerja fan sangat berhubungan dengan panas GPU, dan bekerja secara dinamis mengikuti perkembangan temperatur dari GPU dan memory.
Hasil test antara 2D dan 3D
Untuk percobaan dilakukan 2 cara pada dibawah ini, sehingga VGA PowerColor Radeon HD 4850 dapat lebih minimal mengkonsumsi power
Hardware yang digunakan dibawah ini :
CPU : Intel E8500 3.16Ghz, speed default
VGA : PowerColor HD4850.
Board : Gigabyte GA-X48T-DQ6 DDR3 Intel® X48
Cooler: Zalman CNPS 9700.
Memory : 4GB Kingston DDR3 PC1066 (clock 800Mhz)
Power : Corsair HX 520
Harddisk : Hitachi 250B SATA, Seagate 120GB SATA data
Monitor : Acer P223W
Untuk recording konsumi power , mengunakan digital ampere meter dan mencatat seluruh kebutuhan power listrik untuk computer.
Tweak menurunkan kinerja VGA dengan 2D
VGA dengan kekuatan besar tentu akan memboroskan pemakaian power computer. PowerColor Radeon HD 4850 memiliki kecepatan default 2D dengan 500Mhz untuk GPU dan 993Mhz untuk memory
Bila anda sering mengunakan VGA dengan perfoma papan atas, seperti PowerColor Radeon HD 4850. Tetapi lebih banyak memanfaatkan untuk aplikasi biasa seperti video editing, office dan sebagainya. Sementara kinerja 3D jarang digunakan , maka VGA secara tidak langsung VGA akan membuang power cukup besar.
Untuk menurunkan kinerja VGA, untuk konsumsi power 2D memang tidak akan berdampak besar. Karena bagian terbesar sudah diminimalkan oleh produsen chip. Misalnya PowerColor Radeon HD 4850 memang sudah mengatur kecepatan minimal pada GPU dengan 500Mhz untuk performa 2D.
Tetapi satu cara agar PowerColor Radeon HD 4850 lebih hemat memakai power. Karena PowerColor Radeon HD 4850 mengunakan clock speed memory pada kecepatan 993Mhz. Pada setting Overdrive terdapat pilihan kinerja memory agar diturunkan sampai 750Mhz. Kunci pada kecepatan memory minimum menjadi 750Mhz, sehingga memory VGA tidak terlalu besar mengunakan power.
Walaupun dampaknya tidak terlalu besar, tetapi kinerja VGA tidak terasa berbeda dibandingkan kecepatan default yang diberikan pabrikan. Selisih penghematan dengan kecepatan VGA asli dan setelah diturunkan, menjadi 135W dari seluruh kebutuhan power PC. Sementara dengan kecepatan standar akan mencapai 141W untuk total konsumsi power computer.
Perbedaan selisih 6W memang tidak berdampak besar, tetapi kinerja memory yang tidak perlu dapat ditekan. Selain menghemat power, maka panas memory tidak akan menambah beban kerja fan VGA.
Tweak menurunkan kinerja VGA dengan 3D
Apakah akan berdampak juga dengan konsumsi power VGA ketika digunakan untuk 3D. Jawabannya tentu dan berdampak lebih besar . Kami mencoba menurunkan kecepatan default 625Mhz (GPU) / 993Mhz (Memory) menjadi 500Mhz (GPU) / 750Mhz (Memory) melalui software Overdrive dari ATI. Selisih konsumsi power lebih rendah 20W. Seperti gambar grafik dibawah ini, ketika kinerja VGA dilakukan penuh maka konsumsi power mencapai 230W. Bila kecepatan VGA diturunkan akan berada antara 210W.
Pertanyaan lain, apakah kinerja VGA akan ikut turun. Sudah pasti menurun tetapi hasil performa penurunan nantinya akan relatif. Dengan mengikat kecepatan VGA pada titik minimum untuk GPU dan memory agar menghemat konsumsi power, sehingga performa VGA akan dampaknya pada game dan menjadi lebih lambat.
Pada test dibawah ini adalah catatan dari kecepatan VGA dengan game COD5. Dengan mengunakan resolusi 1680 x 1050 widescreen, kinerja VGA akan menurun antara 10-15 frame.
Perlu tidaknya menurunkan kinerja VGA agar lebih hemat
Kapan kita perlu berhemat dan tidak perlu membuang energi listrik pada computer
* Bila computer mengunakan VGA power besar. Sementara computer lebih banyak digunakan pada posisi stand-by dan aplikasi biasa, maka langkah menurunkan kecepatan memory dari ATI Overdrive akan menghemat konsumsi power.
* Sedang untuk aplikasi game 3D yang tidak terlalu intensif, seperti RTS game. Mengunakan setting minimal untuk VGA adalah cara yang bijaksana untuk menghemat power computer.
* Kombinasi antara menurunkan kecepatan memory VGA dan GPU, menjadi alternatif untuk sebuah computer agar lebih hemat bekerja. Misalnya tetap mempertahankan kinerja memory , sementara kecepatan GPU ditekan seminimal mungkin. Gunakan setting kekuatan VGA yang anda butuhkan sehingga konsumsi power dapat ditekan.
* Bila computer masih mengunakan power supply yang boros, dampaknya akan lebih besar dalam menghemat power listrik.
Tetapi seberapa besar pemakaian power dapat ditekan. Apakah kita harus menerima produk mereka begitu saja. Seberapa besar kemampuan VGA yang dibutuhkan. Atau ada trik tertentu agar konsumsi power dapat lebih dihemat.
Antara 2D dan 3D memiliki dampak berbeda
Kapan sebuah VGA mengkonsumsi power paling besar. Pastinya ketika VGA bekerja pada aplikasi gaming, proses 3D akan dilakukan penuh oleh GPU dan pemakaian power yang dibutuhkan juga maksimal.
Produsen VGA berupaya agar GPU mereka tidak terlalu banyak mengkonsumsi power. Seperti produk ATI memberikan 2 kecepatan antara 2D dan 3D. Contoh saja pada test pengujian dengan PowerColor Radeon HD 4850, GPU ATI 4850 memiliki 2 kecepatan berbeda pada kodisi 2D dan 3D.
Ketika VGA digunakan untuk aplikasi biasa, seperti kinerja office dengan kinerja 2D. Maka VGA akan menurunkan kecepatan clock GPU lebih rendah dibandingkan kecepatan 3D.Untuk 2D misalnya kinerja GPU hanya bekerja pada speed 500Mhz. Sedangkan kinerja GPU pada aplikasi 3D seperti gaming, maka VGA akan meningkat pada kecepatan 625Mhz.
Tentu menjadi pertanyaan bagi anda, misalnya apakah konsumsi power sebuah VGA PowerColor Radeon HD 4850 dapat diturunkan. Jawabnya tentu saja bisa. Tetapi dengan sedikit pengorbanan pada performa untuk menekan pemakaian power VGA.
Caranya dibawah ini
Anda dapat masuk ke bagian control VGA atau Catalyst Control Center. Dan clik bagian ATI Overdrive, disana akan informasi untuk kinerja GPU seperti kecepatan memory dan GPU sampai suhu dan putaran fan yang terjadi.
PowerColor Radeon HD 4850 memiliki 2 kecepatan berbeda. Ketika proses 2D ditangani oleh GPU maka kinerja GPU menjadi 500Mhz dengan kecepatan memory 993Mhz. ketika kinerja GPU menangani 3D seperti gaming, GPU akan naik menjadi 625Mhz dengan kecepatan memory tetap yaitu 993Mhz. Sehingga untuk proses 3D akan mengkonsumsi power lebih besar.
Apa yang dapat dilakukan untuk menghemat power pada VGA PowerColor Radeon HD 4850. Caranya dengan memainkan kinerja clock GPU dan memory.
Tetapi tidak disarankan menurunkan kecepatan putaran fan. Mengingat kinerja fan sangat berhubungan dengan panas GPU, dan bekerja secara dinamis mengikuti perkembangan temperatur dari GPU dan memory.
Hasil test antara 2D dan 3D
Untuk percobaan dilakukan 2 cara pada dibawah ini, sehingga VGA PowerColor Radeon HD 4850 dapat lebih minimal mengkonsumsi power
Hardware yang digunakan dibawah ini :
CPU : Intel E8500 3.16Ghz, speed default
VGA : PowerColor HD4850.
Board : Gigabyte GA-X48T-DQ6 DDR3 Intel® X48
Cooler: Zalman CNPS 9700.
Memory : 4GB Kingston DDR3 PC1066 (clock 800Mhz)
Power : Corsair HX 520
Harddisk : Hitachi 250B SATA, Seagate 120GB SATA data
Monitor : Acer P223W
Untuk recording konsumi power , mengunakan digital ampere meter dan mencatat seluruh kebutuhan power listrik untuk computer.
Tweak menurunkan kinerja VGA dengan 2D
VGA dengan kekuatan besar tentu akan memboroskan pemakaian power computer. PowerColor Radeon HD 4850 memiliki kecepatan default 2D dengan 500Mhz untuk GPU dan 993Mhz untuk memory
Bila anda sering mengunakan VGA dengan perfoma papan atas, seperti PowerColor Radeon HD 4850. Tetapi lebih banyak memanfaatkan untuk aplikasi biasa seperti video editing, office dan sebagainya. Sementara kinerja 3D jarang digunakan , maka VGA secara tidak langsung VGA akan membuang power cukup besar.
Untuk menurunkan kinerja VGA, untuk konsumsi power 2D memang tidak akan berdampak besar. Karena bagian terbesar sudah diminimalkan oleh produsen chip. Misalnya PowerColor Radeon HD 4850 memang sudah mengatur kecepatan minimal pada GPU dengan 500Mhz untuk performa 2D.
Tetapi satu cara agar PowerColor Radeon HD 4850 lebih hemat memakai power. Karena PowerColor Radeon HD 4850 mengunakan clock speed memory pada kecepatan 993Mhz. Pada setting Overdrive terdapat pilihan kinerja memory agar diturunkan sampai 750Mhz. Kunci pada kecepatan memory minimum menjadi 750Mhz, sehingga memory VGA tidak terlalu besar mengunakan power.
Walaupun dampaknya tidak terlalu besar, tetapi kinerja VGA tidak terasa berbeda dibandingkan kecepatan default yang diberikan pabrikan. Selisih penghematan dengan kecepatan VGA asli dan setelah diturunkan, menjadi 135W dari seluruh kebutuhan power PC. Sementara dengan kecepatan standar akan mencapai 141W untuk total konsumsi power computer.
Perbedaan selisih 6W memang tidak berdampak besar, tetapi kinerja memory yang tidak perlu dapat ditekan. Selain menghemat power, maka panas memory tidak akan menambah beban kerja fan VGA.
Tweak menurunkan kinerja VGA dengan 3D
Apakah akan berdampak juga dengan konsumsi power VGA ketika digunakan untuk 3D. Jawabannya tentu dan berdampak lebih besar . Kami mencoba menurunkan kecepatan default 625Mhz (GPU) / 993Mhz (Memory) menjadi 500Mhz (GPU) / 750Mhz (Memory) melalui software Overdrive dari ATI. Selisih konsumsi power lebih rendah 20W. Seperti gambar grafik dibawah ini, ketika kinerja VGA dilakukan penuh maka konsumsi power mencapai 230W. Bila kecepatan VGA diturunkan akan berada antara 210W.
Pertanyaan lain, apakah kinerja VGA akan ikut turun. Sudah pasti menurun tetapi hasil performa penurunan nantinya akan relatif. Dengan mengikat kecepatan VGA pada titik minimum untuk GPU dan memory agar menghemat konsumsi power, sehingga performa VGA akan dampaknya pada game dan menjadi lebih lambat.
Pada test dibawah ini adalah catatan dari kecepatan VGA dengan game COD5. Dengan mengunakan resolusi 1680 x 1050 widescreen, kinerja VGA akan menurun antara 10-15 frame.
Perlu tidaknya menurunkan kinerja VGA agar lebih hemat
Kapan kita perlu berhemat dan tidak perlu membuang energi listrik pada computer
* Bila computer mengunakan VGA power besar. Sementara computer lebih banyak digunakan pada posisi stand-by dan aplikasi biasa, maka langkah menurunkan kecepatan memory dari ATI Overdrive akan menghemat konsumsi power.
* Sedang untuk aplikasi game 3D yang tidak terlalu intensif, seperti RTS game. Mengunakan setting minimal untuk VGA adalah cara yang bijaksana untuk menghemat power computer.
* Kombinasi antara menurunkan kecepatan memory VGA dan GPU, menjadi alternatif untuk sebuah computer agar lebih hemat bekerja. Misalnya tetap mempertahankan kinerja memory , sementara kecepatan GPU ditekan seminimal mungkin. Gunakan setting kekuatan VGA yang anda butuhkan sehingga konsumsi power dapat ditekan.
* Bila computer masih mengunakan power supply yang boros, dampaknya akan lebih besar dalam menghemat power listrik.
Langganan:
Postingan (Atom)
