Palit has started manufacturing motherboards only recently. As is well known, it's a big manufacturer of graphics cards, mostly on NVIDIA GPUs (initially, the company produced only NVIDIA cards, but ATI Radeon cards also appeared two years ago, even though GeForces are still in majority). Such preferences naturally affected the choice of motherboard chipsets and earlier release dates of products on new chipsets. Today we are going to examine a motherboard on GeForce 8200, previously known as MCP78. As this motherboard is an engineering sample, this is actually a preview of the new chipset by the example of this motherboard. The GeForce 8200 chipset is equipped with an integrated graphics core supporting DirectX 10, its system buses are updated to Hyper-Transport 3.0 and PCI Express 2.0, it supports HDMI, DVI, and VGA outputs, but it can use only two outputs simultaneously - digital and analog ones. Just like GPUs in modern graphics cards, the chipset has an integrated unit to decode HD video in MPEG2, H.264, and VC-1 formats. To raise 3D performance, the integrated graphics core can work with an external graphics card in Hybrid SLI mode. And when you don't play games, the system uses only the integrated graphics core to reduce noise and power consumption (both options are not supported by the drivers yet). NVIDIA is not going to abandon the single chip architecture even in such complex solutions, where peripheral functionality commands respect: 6 x SATA ports supporting RAID 0, 1, 0+1, and 5, integrated gigabit MAC adapter supporting traffic shaping, and 12 x USB 2.0. But let's proceed from theoretical characteristics to practical tests of the motherboard based on this chipset. In this case, engineers apparently tried to design a motherboard with minimum multimedia features. It lacks additional controllers. Fortunately, the chipset offers all necessary features (and six SATA ports are more than enough) for a compact motherboard with integrated graphics for inexpensive office and home computers. DVI-D and VGA connectors on the rear panel also indicate that this product is designed for serious multi-monitor configurations rather than for media centers. This motherboard has a good PCB layout for a compact product. The only gripe we have is an additional power connector in the center of the board (as a rule, a corresponding cable in this case goes through a CPU cooler, and owners of bulky devices may be disappointed). Designers tried to take into account interests of users that own even large graphics cards. At least three SATA drives can be plugged, even if a graphics card takes up all room from the graphics slot down. On the other hand, the chipset heat sink may not allow to install a card with additional heat sinks on the back side. The heat sink on the chipset gets quite hot only under heavy load. That may be the reason why the official TDP is 14.3 W, which is relatively high for a modern chipset for the AMD platform. However, even lengthy game tests did not reveal any failures (in an open testbed without additional ventilation). Besides, if you want to raise efficiency of chipset cooling, you may take into account that fins on the chipset heat sink are perpendicular to the processor socket and buy a proper CPU cooler. This PCB has a unique design, there are no empty seats. The 4-phase switching voltage regulator of the processor incorporates three field-effect transistors per channel and three 820 uF capacitors, 5 x 680 uF, 1 x 1000 uF, 4 x 1800 uF. Our engineering sample used capacitors from little-known manufacturers. However, production-line motherboards may use different components, probably from United Chemi-Con (traditional Palit's partner). Motherboard dimensions - 244x220 mm (truncated microATX), six-screw mount, the right edge is not fixed at the level of connectors for storage devices. As the motherboard is a tad smaller than standard microATX, power connectors and IDE/SATA slots are quite far from the retention crews. Perhaps, engineers should have preserved these several centimeters of PCB. System monitoring (ITE IT8716F-S, according to BIOS Setup):
Onboard ports, sockets, and connectors
Back panel (left to right, blockwise)
Integrated Controllers
We couldn't evaluate the integrated audio quality, because we didn't even manage to install Realtek HD Audio driver, although we tried several versions (we got a device initialization error) in Windows XP. It's most likely a problem of our engineering sample. Settings
We used BIOS 0.9I, flashed by the manufacturer. The mentioned BIOS parameters are available in this version, but the viability of non-standard settings hasn't been tested. There are some overclocking options, but you cannot raise CPU core voltage. BIOS offers appropriate features for its class, you can hardly ask for more. We were a little confused by the lack of memory timing control, but only until our performance tests (below). OverclockingIn order to evaluate motherboard and its BIOS, we overclock our testbed processor to a maximum stable level. We use all features of the motherboard in this test, including raising CPU voltage and adjusting multipliers and frequencies of system and peripheral buses, if necessary. But if, for example, reducing Hyper-Transport frequency does not improve overclocking, we leave the default multiplier. Memory is set to the standard frequency for a given memory module (multiplier correction), if a manufacturer does not publish any ways to improve memory overclocking. Otherwise, we analyze their efficiency as well. In order to evaluate stability of the overclocked system, we load Windows XP and run WinRAR performance test for 10 minutes (Tools - Benchmark and hardware test). As overclocking potential is an individual property of a given motherboard sample to some degree, we don't set the task to determine overclocking potential to within a single MHz. In practice, we are to find out whether CPU overclocking will be limited by a motherboard as well as to evaluate its behavior in non-standard modes, including automatic restoration of a correct frequency after a failed overclocking attempt, etc.
It does not seem right to draw conclusions on overclocking for an engineering sample. But we should note that the limiting factor in this case is the integrated graphics core. When we raised frequencies further, there appeared image defects and the system would freeze under any graphics load. PerformanceTestbed configuration:
We decided to compare our motherboard under review with the Biostar TF7050-M2 on the previous NVIDIA GeForce 7050 chipset with integrated graphics.
First of all, we should compliment chipset engineers for a twofold gain in frame rate! Of course, we shall test this chipset in modern games and in Windows Vista, try Hybrid SLI in action (its drivers are not ready yet), and touch upon video decoding issues. Besides, it's nice to see some progress in tests with a discrete graphics card as well. In cases, when performance used to be limited by the platform (in all resolutions except the maximum, in our game tests), the problem was not in a CPU, as you might have thought, but in the efficiency of data exchange along the bus. Now we can see that results in all tests with discrete graphics are limited by performance of the graphics card itself. Moreover, it happens even though the controller operates in the PCI Express 1.0 compatibility mode, and BIOS does not even have a release number. Finally, proceeding to results of the motherboard itself, we should note maximum performance in arithmetic tests for our combination of CPU/RAM. We are used to getting such results on motherboards that allow manual control of memory timings and work well with the minimum value of the 1T/2T timing. In this case, you cannot specify timings in BIOS. So we can only assume that optimal (probably even aggressive) timings were set automatically. Well done, Palit! ConclusionsIt's not quite correct to draw conclusions on an engineering sample. We publish this article, because we want to show our readers excellent test results of this chipset, which has been already launched. Frankly speaking, we haven't expected such a performance gain. If you keep tabs on the market of graphics cards, you may remember that the upgrade from DirectX 9 to DirectX 10 in inexpensive graphics cards was accompanied by a significant regress in performance in old games. In this case we get DirectX 10 support (it won't break any records, of course, and it will probably be insufficient for dynamic games) and additional power for "quiet" games (they are in majority, by the way). I repeat that we are not through with reviews of chipsets for the AMD platform supporting DirectX 10. On the contrary, it's only a beginning. Along with further tests of the GeForce 8200, we'll publish test results of motherboards on AMD 780G in a couple of days. "Neutral" manufacturers of motherboards (those that use chipsets from different manufacturers) are unanimous that it's even better. As for now, we can only say that the 780G indeed demonstrates excellent results in those tests that we've already completed. Motherboard provided by the manufacturer
Dmitry Laptev (lpt@ixbt.com) March 14, 2008 Write a comment below. No registration needed!
|
Platform · Video · Multimedia · Mobile · Other || About us & Privacy policy · Twitter · Facebook Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved. |