Original Link: http://www.anandtech.com/show/1947
NVIDIA nForce4 Ultra: Biostar's Performance Surpriseby Gary Key on February 16, 2006 12:05 AM EST
- Posted in
"Mistakes are the portals of discovery."This quote by the accomplished Irish author, James Augustine Joyce, describes our experiences with the Biostar TForce4 U 775 motherboard. Biostar has been in existence since 1986 and has a long history of providing generally good products at inexpensive price points. They have branched out recently with video card and SFF products with success. In fact, the majority of Biostar's market success is tied to their ability to quickly deliver product based on current chipsets or designs at bargain pricing. Biostar's products might not be as polished or feature-rich as other manufacturers, but they typically offer mid-range performance at entry level pricing.
The T-series product line currently includes five AMD based products and two Intel based products. This product series is a departure for Biostar as they are catering to the gaming and enthusiast crowd with this product line. The boards feature BIOS enhancements specifically tailored to the casual overclocker along with upgraded hardware components and window's based utilities. More information about the entire line of T-series products can be found here.
The TForce4 U 775 motherboard is based on the recently announced nForce4 Ultra SPP and nForce4 Ultra Intel Edition MCP chipsets. Our initial impression of the motherboard's performance was that of mediocrity and disappointment with a chipset that was designed to compete directly against the excellent Intel 945P. We were harsh in our opinion of the board's performance to Biostar and NVIDIA. In the end, our initial opinion was a mistake and it led to a path of discovery about the performance potential of this chipset and board.
Biostar along with NVIDIA had discovered issues with the original BIOS design and worked together in creating a highly optimized BIOS release that is at the center of the performance results generated by this board. While the performance is nothing less than astounding in certain benchmarks, it is also very average in others. The board has an almost "Dr. Jekyll and Mr. Hyde" quality about it in configuration, performance, and options.
Once we discovered the right set up, the board's stability was exceptional and delivered impressive results in the latest synthetic and game benchmarks. However, the path that we traveled to find the limits of this board generally resulted in frustration and, at times, agony in the amount of work required to recover from errant settings.
The Self Recovery System (S.R.S.) within the BIOS would allow the system to recover from most errant settings, but we continually experienced random lockups upon entering the BIOS setup screen. These random lockups would require several power down and power up episodes or usage of the clear CMOS jumper. This process might be considered acceptable to an avid enthusiast who is discovering the limits of the board, but it is not for the intended buyer of this product.
The board was not very forgiving if we made a mistake with the settings. We would notice the board performing extremely well in both general applications and benchmarks at a given setting and then completely refuse to boot if we increased the FSB or Memory speeds by a couple of MHz. The majority of boards that we have tested recently would still boot, but would fail a benchmark or generate sporadic results until you finally reached the limit of the board. We attribute this to the Biostar BIOS optimizations and board level design due to cost constraints in meeting the US$95 price target.
Updated Bios Results
Biostar listened to our issues and provided an impressive bios update that has once again transformed this board. Biostar was able to replicate our Self Recovery System (S.R.S.) issues and recently provided bios update, NFUIA210.BST, which has eliminated random lockups once you enter the bios setup screen. This bios also allows for a graceful system recovery once you increase the FSB or Memory speeds past the board's limits for stable operation. We noted the board's overclocking performance and general stability at the increased FSB speeds has also improved. Overall the board feels more responsive and polished in our testing than before. We appreciate Biostar's ability to quickly resolve our reported issues while at the same time improving the board's performance.
The nForce4 Ultra SPP and nForce4 Ultra Intel Edition MCP are newly released chipsets for the mainstream market and will compete directly against the Intel 945P chipset. Unlike the Intel 945P chipset, the nForce4 Ultra SPP offers full support for all Socket 775 processors, including the Extreme Edition units. Further details on the release of this product can be found in our preview article located here.
The chart above lists the standard feature set available when utilizing the NVIDIA nForce4 Ultra SPP and nForce4 Ultra Intel Edition MCP chipsets. The nForce4 Ultra SPP enables full support for 1x16 PCI-E graphics support , up to four x1 PCI-E devices, 4GB memory addressability, and native DDR-2 667MHz memory support. The nForce4 Ultra Intel Edition MCP enables support for 8 USB 2.0 ports, HD Audio with S/PDIF connectors, Gigabit Ethernet, 5 PCI slots, 4 SATA 3.0Gb/s ports, and 4 IDE devices.
Unfortunately, Biostar chose to differentiate these options with AC-97 audio via the Realtek ALC-850 and 10/100Mb/s PCI Ethernet via the Realtek 8201CL PHY.
Let's find out what else we discovered with this board.
Basic Features: Biostar TForce4 U 775
|Specification||Biostar Tforce4 U 775|
|CPU Interface||LGA775-based Pentium 4, Pentium 4 XE, Celeron D, and Pentium D processors|
|Chipset||NVIDIA nForce4 Ultra - SPP (C19, rev. C1)
NVIDIA nForce4 Ultra - MCP (MCP51, rev A20)(nForce 430)
|Pentium D Support (Dual-Core)||820D, 830D, 840D, 840EE, 920D, 930D, 940D, 950D, 955EE|
|Front Side Bus||1066 / 800 / 533 MHz|
|Front Side Bus Speeds||Default, 450MHz ~ 1300MHz in 1MHz increments|
|Memory Speeds||Default, 400MHz ~ 1300MHz in 1MHz increments|
|PCI Bus Speeds||Default|
|PCI Express Bus Speeds||Default, 100MHz ~ 148.4375MHz in various increments|
|Set Processor Multiplier||12 to 60, (in 1 step increments)|
|LDT Multipliers||1x, 2x, 3x, 3.5x, 4x, 5x|
|Core Voltage||Default, 1.1000V to 1.8250V (in 0.0125V increments)|
|DRAM Voltage||Default, 2.05V, 2.10V, 2.25V|
|SPP Chipset Voltage||Default, 1.53V, 1.58V, 1.61V|
|FSB Voltage||Default, 1.30V, 1.40V, 1.50V|
|Memory Slots||(4) x DIMM, max. 4GB, DDR2 667/533/400, non-ECC, un-buffered memory|
|Expansion Slots||(1) x PCI-E x16 (operates in 1x16 mode)
(2) x PCI-E x1 (operates in x1 mode)
(4) x PCI 2.2
|Onboard SATA||NVIDIA nForce 430: (4) x SATA II|
|Onboard IDE||NVIDIA nForce 430: (2) x UltraDMA 133/100/66/33|
|SATA/IDE RAID||NVIDIA nForce 430:
(4) x SATA II
RAID 0, RAID 1, RAID 0+1, RAID 5
|Onboard USB 2.0||(8) USB2.0 ports|
|Onboard LAN||Realtek 8201CL PCI 10/100Mb/s LAN - PHY|
|Onboard Audio||Realtek ALC-850, 8-channel capable Audio|
|Power Connectors||24-pin ATX
4-pin ATX 12V
|Back Panel I/O Ports||1 x PS/2 Keyboard
1 x PS/2 Mouse
1 x Parallel
1 x Serial
1 x Audio I/O Panel
1 x RJ45
4 x USB
|Other Features||(O.N.E.) - Overclocking Navigator Engine
(C.R.P.) - CMOS Reloading Program
(M.I.T.) - Memory Integration Testing
(I.F.P.) - Integrated Flash Program
(S.R.S.) - Self Recovery System
Hardware Monitor - Windows Based
Overclocking Utility - Windows Based (Limited Functionality)
Smart Fan Function - Windows Based
Live Update - Windows Based BIOS Update Program
The Biostar TForce4 U 775 is a member of the TForce4 Series product family and, as such, is a value-based board targeted towards the enthusiast user who is concerned about price and performance. The board ships with an accessory package that includes the standard assortment of IDE/SATA cables and power connectors. Biostar also includes a driver CD along with several desktop utilities for monitoring and tuning your system.
This is the main BIOS screen and it displays the generic options available on the board.
The Overclock Navigator section allows you to change the majority of the system performance settings. The system features the Auto Overclock System that will automatically adjust most system parameters based upon the components installed. The V6-Tech setting will overclock the system around 10~15%, V8-Tech setting will allow up to a 25% overclock, and the V12-Tech system will allow up to a 30% system overclock.
The balance of the Overclock Navigator section allows you to set individual parameters manually for voltage, multipliers, and chipset settings.
The board fully supports manual memory timing adjustments or allows for a Default setting that will set the memory to the SPD settings.
An additional feature on the board is the CMOS reloading program that allows users to save up to 50 sets of CMOS settings into the BIOS. The user is able to personalize each setting before saving it. The feature worked well in our testing, provided that we were able to reach the BIOS setting.
Overall, the BIOS options are very good for a chipset that NVIDIA has stated is not optimized for overclocking. The board's bios even includes a built in memory tester based upon Memetest86+ version 1.55. We would like to see more memory voltage options included in the BIOS along with the ability to adjust the PCI bus speeds.
Biostar Tforce4 U 775: Features
Biostar designed a very well laid out board with all major connections easily reached. The board is lacking most clearance issues and was simple to install in a mid-size ATX case. Although the board features a 3-phase voltage regulator power design, it provided very good stability and allowed a decent level of overclocking. However, considering the power requirements of the current Intel processor lineup, we would like to have seen a 4-phase or 5-phase design considering this board is targeted to the gamer and computer enthusiast.
The DIMM module slots' color coordination is correct for dual channel setup. The memory modules are slightly difficult to install with a full size video card placed in the first PCI Express x16 slot. The floppy drive port connector is conveniently located on the edge of the board. The CPU fan header is located above the first DIMM slot and will require proper placement of the heat sink assembly due to the distance from the CPU socket.
The NVIDIA nForce4 IDE port connectors are located to the right of the battery and did not present any connection issues in our mid-size ATX case. The system fan header is located between the battery and nForce4 SATA ports. The NVIDIA nForce4 Ultra Intel Edition MCP chipset is passively cooled with a low rise heat sink that did not interfere with cards installed into the PCI slots.
The NVIDIA nForce4 SATA ports are conveniently located below the MCP chipset and to the left of the IDE connectors. The SATA ports feature the new clamp and latch design. Unlike other recently reviewed boards, the SATA ports are not color-coded for primary and secondary operation. We found the positioning of the SATA ports to be excellent when utilizing the PCI slots or IDE port connectors.
The NVIDIA nForce4 USB connectors, S/PDIF out connector, and chassis panel are located to the left of the MCP chipset along the edge of the board. The BIOS chip is located below the fourth PCI slot. The clear CMOS jumper block is a traditional jumper design located below the number one SATA port connector. The location of this jumper required the removal of the SATA drive cable during repeated usage.
The board comes with (1) physical PCI Express x16 connector, (2) PCI Express x1 connectors, and (4) PCI 2.2 connectors. The layout of this design offers an exceptional balance of expansion slots for a mainstream board.
The physical x16 connector is located next to the 24-pin ATX power connector. The (2) PCI Express x1 connectors are located next, followed by the (4) PCI slots.
We did not have any issues installing our EVGA 7800GTX 512MB or ATI X1900XTX video cards in the x16 PCI Express slot. This configuration will physically render the first PCI Express x1 slot useless. We did not have any issues utilizing this slot with video cards containing single slot cooling systems.
Returning to the CPU socket area, we find ample room for alternative cooling solutions. We utilized the stock Intel heatsink, but also verified that several aftermarket cooling systems such as the Thermaltake Big Typhoon would fit in this area during our tests. However, due to the large SPP heatsink/fan, the installation of large air or water-cooling solutions could be problematic.
The NVIDIA nForce4 SPP chipset is actively cooled with a large heat sink/fan unit that did not interfere with any installed peripherals. In fact, this unit kept the normally hot SPP chipset cool enough that additional chipset voltage was not a factor in our overclocking tests. Our only concern is the lifespan of the fan as it typically runs at 5500rpm and does generate a slight whirling sound during operation.
Biostar places the four-pin 12v auxiliary power connector at the top of the CPU socket area, but out of the way of aftermarket cooling solutions. The 24-pin ATX power connector is located above the SPP chipset and next to the x16 PCI Express connector. This 24-pin connector is located in a difficult position and can hamper airflow with cabling that crosses over the SPP heatsink/fan or directly over the CPU heatsink/fan.
The rear panel contains the standard PS/2 mouse and keyboard ports, parallel port, serial port, LAN port, and 4 USB ports. The LAN (RJ-45) port has two LED indicators representing Activity and Speed of the connection. The audio panel consists of 6 ports that can be configured for 2, 4, 6, and 8-channel audio connections.
FSB Overclocking Results
|Front Side Bus Overclocking Testbed|
|Processor:||Pentium 4 Smithfield LGA 775
840EE Dual Core 3.2GHz
|CPU Voltage:||1.4825V (1.4000V default)|
|Memory Settings:||3-2-2-8 at 667MHz|
|Cooling:||Thermaltake Big Typhoon|
|Power Supply:||OCZ Power Stream 520|
|Maximum CPU OverClock:||250fsb x 17 (4250MHz) +33%|
|Maximum FSB OverClock:||255fsb x 16 (4092MHz) +27%|
This board is an average overclocker when comparing it to the Intel 975X boards. At these overclock settings, the system was able to complete all of our benchmark test suites three consecutive times and run Prime95 and SuperPI without issue. We lowered the CPU multiplier to 14, but could not overclock past a 252FSB setting. We changed the LDT ratio to several different settings without success. These results continue the pattern in which we see with the nForce4 Intel Edition chipsets not reaching high FSB speeds, but being successful in overclocking the CPU at standard FSB settings.
We installed our 955EE CPU and were able to reach the highest CPU overclock recorded for our sample unit. The board performed flawlessly at this setting and reaching 4.55GHz on air with this chip is impressive. However, all attempts at lowering the multiplier and extending the bus speed past 1180MHz resulted in failure.
Updated 955EE Testing
We flashed the board to the latest bios, NFUIA210.BST, installed our 955EE CPU, and were able to reach a very stable 4.62GHz with air cooling. Our previous attempts at this multiplier resulted in a stable 4.50GHz at a bus speed of 1060MHz. We were able to boot the machine at 4.76GHz but could not complete our benchmark testing.
Memory Stress Testing
Memory stress tests look at the ability of the Biostar TForce4 U 775to operate at the officially supported memory frequencies of 667MHz DDR2, at the best performing memory timings that the Corsair CM2X512A-5400UL revision 1.3 will support.
|Biostar TForce4 U 775
Stable DDR667 Timings - 2 DIMMs
(2/4 slots populated - 1 Dual-Channel Bank)
|Clock Speed:||200MHz (800FSB)|
|Timing Mode:||667MHz - Default|
|RAS to CAS Delay:||2|
|RAS Cycle Time:||8|
The Biostar TForce4 was very stable with 2 DDR2 modules in Dual-Channel at the settings of 3-2-2-8 at 2.1V. We will now install all four available memory slots that result in more strenuous requirements on the memory subsystem than testing 2 DDR2 modules on a motherboard.
|Biostar TForce4 U 775
Stable DDR667 Timings - 4 DIMMs
(4/4 slots populated - 2 Dual-Channel Banks)
|Clock Speed:||200MHz (800FSB)|
|Timing Mode:||667MHz - Default|
|RAS to CAS Delay:||3|
|RAS Cycle Time:||8|
The Biostar TForce4 was completely stable with 4 DDR2 modules in Dual-Channel operation at the settings of 3-3-3-8, but required the command rate to be increased to 2T. We tried several combinations of memory settings at the 3-2-2-8 settings, but the board was not stable enough to complete our test suite.
The NVIDIA nForce4 Ultra chipset fully supports all dual core Pentium D processors in both stock and overclocked conditions. This chipset also supports the upcoming Cedar Mill processor range. Dual core really makes a difference in certain multi-tasking scenarios, as was demonstrated in the dual core performance preview. If you are interested in how the various chipsets perform in a real world multitasking setup, please take another look at that review.
|Performance Test Configuration|
|Processor(s):||Intel Pentium 840EE (3.2GHz, 800FSB, Dual-Core, 2x1MB L2, HT) utilized for all tests|
|RAM:||2 x 512MB Corsair CM2X512A-5400UL revision 1.3 Settings- DDR2-667 as noted at (CL3-2-2-8)|
|Hard Drive(s):||2 x Maxtor MaXLine III 7L300S0 300GB 7200 RPM SATA (16MB Buffer), 1 x Maxtor MaXLine III 7L300R0 300GB 7200 RPM IDE (16MB Buffer)|
|System Platform Drivers:||Intel Chipset Software - 22.214.171.1246
NVIDIA Platform Driver - 8.22
|Video Cards:||1 x XFX 7800GTX OC (PCI Express) for all tests|
|Video Drivers:||NVIDIA nForce 81.98 WHQL|
|Cooling:||Thermaltake Big Typhoon|
|Power Supply:||OCZ Power Stream 520|
|Operating System(s):||Windows XP Professional SP2|
|Motherboards:||Biostar TForce4 U 775
Asus P5N32-SLI Deluxe
Gigabyte GA-G1 975X
Asus P5LD2 Deluxe
General Performance & Encoding
The Biostar TForce4 U 775 absolutely dominates the other boards in PCMark 2005 and 3Dmark05. While the Biostar board scores near the bottom in the balance of the synthetic benchmarks it still leads the 945P based board in these benchmarks. We ran the tests several times and on a second board with the same results. It is obvious the BIOS optimizations have made a difference in these particular benchmarks and if the memory performance of the board improves, we are sure that it would outscore the 975x based boards. The Biostar board also performs very well in the encoding tests. Our encoding tests will soon change to the DivX 6.1 codec and additional multimedia tasks.
We recently switched to version 2.50 of Everest, so these scores are not comparable to previous tests with version 2.20. The memory latency test continues to show a 7% advantage for the nForce4's memory controller on the Asus board. The Biostar board's performance was near the bottom of the field and mirrors the results in the more memory and cpu sensitive 3DMark03 and 3DMark01 tests. The surprising number is the difference between the Asus P5N32-SLI and Biostar TForce4 U 775 in the read performance results, which results in the Asus being 4% faster.
The overclocking performance graphs have been added to the standard benchmark test suite and should allow for a better comparison on the overclocking capabilities of tested boards. For more details on the specific overclocking abilities of this board, please refer to the Overclocking and Memory Stress Test section in the Basic Features section.
The Biostar TForce4 U 775 is an average overclocker, but falls within the normal range of most NVIDIA based Intel boards. We still recommend the Intel chipsets if higher overclocks are required.
The overall gaming performance of the Biostar TForce4 U 775 is phenomenal compared to our recent Intel based test results. The Biostar board took first place in five of the seven tests and scored well in the two that it did not win. In the benchmarks where the minimum and maximum scores are listed it should be noted the Biostar board consistently had the best minimum scores reported. This performance pattern directly led to the board's impressive scores. We contacted NVIDIA and Biostar about the performance of the board and were assured they have witnessed excellent results in their internal testing. We ran the tests on our second board and actually witnessed higher results in the Serious Sam II and Battlefield 2 benchmarks. Although we did not report Doom 3 and Quake 4 scores, the Biostar board was generally 4% faster than the Asus and Intel boards. A combination of the optimized BIOS and chipset revisions has certainly turned this board into an excellent Intel platform for gaming.
Disk Controller Performance
With the variety of disk drive benchmarks available, we needed a means of comparing the true performance of the wide selection of controllers. The logical choice was Anand's storage benchmark first described in Q2 2004 Desktop Hard Drive Comparison: WD Raptor vs. the World. The iPeak test was designed to measure "pure" hard disk performance, and in this case, we kept the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of a hard drive controller with a consistent hard drive.
We played back Anand's raw files that recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPEAK utility was then used to play back the trace file of all IO operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. To try to isolate performance differences to the controllers that we were testing, we used the Maxtor MaXLine III 7L300S0 300GB 7200 RPM SATA drive in all tests . The drive was formatted before each test run and a composite average of 5 tests on each controller interface was tabulated in order to ensure consistency in the benchmark.
iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each IO operation. In order to make the data more understandable, we report the scores as an average number of IO operations per second so that higher scores translate into better performance. This number is meaningless as far as hard disk performance is concerned, as it is just the number of IO operations completed in a second. However, the scores are useful for comparing "pure" performance of the storage controllers in this case.
The performance patterns hold steady across both Multimedia Content IO and Business IO, with the on-board NVIDIA nForce4 SATA 2 still providing the fastest IO, followed closely by the Intel ICH7R, Silicon Image 3132, and Marvell 88SE6141 SATA 2 controllers.
Firewire and USB Performance
After looking at many options for Firewire and USB testing, we finally determined that an external USB 2.0, Firewire 400, and Firewire 800 hard disk would be a sensible way to look at USB and Firewire throughput.
Our first efforts at testing with an IDE or SATA drive as the "server" yielded very inconsistent results, since Windows XP sets up cache schemes to improve performance. Finally, we decided to try a RAM disk as our "server", since memory removed almost all overhead from the serving end. We also managed to turn off disk caching on the USB and Firewire side by setting up the drives for "quick disconnect" and our results were then consistent over many test runs.
We used 1GB of fast 3-2-2-8 system memory set up as a 450MB RAM disk and 550MB of system memory. Our standard file is the SPECviewPerf install file, which measures 432,533,504 bytes (412.4961MB). After copying this file to our RAM disk, we measured the time for writing from the RAM disk to our external USB 2.0, Firewire 400, or Firewire 800 drive using a Windows timing program written for AnandTech by our own Jason Clark. The copy times in seconds were then converted into Megabits per second (Mb) to provide a convenient means of comparing throughput. Higher Rates therefore mean better performance in this particular test.
Possibly the most striking finding in our Firewire and USB throughput tests is the continued performance of an external hard drive connected to Firewire 800. Our benchmarks show Firewire 800 is up to 46% faster than a drive connected to the more common Firewire 400, and about 29% faster than USB 2.0.
The Biostar TForce4 U 775 board does not offer a Firewire option. The USB 2.0 performance is consistent with other NVIDIA based boards and continues to lead the Intel chipset solutions in throughput.
The current motherboard test suite includes LAN performance measurements. All of these boards utilize PCI Express controllers with the only difference being the supplier of the core logic.
The Windows 2000 Driver Development Kit (DDK) includes a useful LAN testing utility called NTttcp. We used the NTttcp tool to test Ethernet throughput and the CPU utilization of the various Ethernet Controllers used on the Intel motherboards.
We set up one machine as the server; in this test, an Intel system with an Intel CSA Gigabit LAN connection. Intel CSA has a reputation for providing fast throughput and this seemed a reasonable choice to serve our Gigabit LAN clients.
At the server side, we used the following Command Line as suggested by the VIA whitepaper on LAN testing:
Ntttcpr -m 4 ,0,‹server IP› -a 4 -l 256000 -n 30000On the client side (the motherboard under test), we used the following Command Line:
Ntttcps -m 4 ,0,‹client IP› -a 4 -l 256000 -n 30000At the conclusion of the test, we captured the throughput and CPU utilization figures from the client screen.
The choice of the PCI based Realtek 8201CL 10/100Mb/s Ethernet PHY is a severe disappointment for a board that natively supports PCI-E Gigabit Ethernet operations. We do not understand why Biostar decided to utilize the 10/100Mb/s option for their T-Series product offering at a time when even most budget boards at least offer 10/100/1000Mb/s options. The performance of the Realtek 8201CL is abysmal compared to the other solutions.
All standard Ethernet tests were performed with standard frames and the NVIDIA Active Armor suite disabled unless otherwise noted. Gigabit Ethernet supports Jumbo frames as well and provides a further reduction in CPU overhead. We added another test scenario in which ActiveArmor and Jumbo frames were enabled on the Asus P5N32-SLI Deluxe board via the 6.82 WHQL platform driver set. This is shown for illustrative purposes and shows the favorable impact of this technology.
We limited audio testing to the Rightmark 3D Sound version 2.2 CPU utilization test and tested with sound enabled to show the performance effects on several games. The Rightmark 3D Sound benchmark measures the overhead or CPU utilization required by a codec or hardware audio chip.
The Realtek ALC-850 was tested with the recently released 3.82 driver set. The Realtek audio drivers do not support more than 25 3D buffers at this time, so the scores cannot be directly compared to the HDA Mystique 7.1 and Creative Labs Sound Blaster X-FI cards in the benchmarks.
The Realtek ALC-850 codec has average CPU utilization rates with reductions of up to 37% in the 3D tests compared to the previous driver release. The HDA Mystique 7.1 Gold has the highest overall utilization rates of the audio solutions tested, but its scores improved up to 20% with the January driver release. BlueGears is no longer supporting the card directly, but HDA continues to offer support. The Realtek ALC-882 performance is equal to the ALC-850 and offers significantly better audio quality. The Sound Blaster X-FI has the lowest rates, with the Sound Blaster Live! 24-bit solution on the Gigabyte board following closely. Let's find out how these results translate into real world numbers.
The audio performance numbers remain consistent as the Realtek ALC-850 consistently finishes behind the HDA Mystique 7.1 and SoundBlaster X-FI except in the Serious Sam II and Splinter Cell benchmarks. Serious Sam II has an average loss of 21%, Half Life 2 at 23%, Splinter Cell at 2%, Battlefield 2 at 26%, and F.E.A.R. at 7%. While the output quality of audio with the Realtek ALC-850 is acceptable for desktop usage, it is not for games that rely on 3D sound for effects and atmosphere. The audio quality in music and video remains somewhat hollow and flat in our subjective testing.
Obviously, if you are a serious gamer, then a dedicated sound card is still a requirement to ensure consistent frame rate averages across a wide variety of games. We noticed in previous testing of our Battlefield 2 and Half Life 2 benchmarks that the ALC-850 would stutter momentarily in intensive scenes. The 3.82 driver release has now eliminated stuttering in the BF2 and Half Life 2 benchmarks.
The Biostar TForce4 U 775 is certainly a performance leader for the Intel market if gaming is your priority and SLI capability is not important. The performance of the board in most of the synthetic and memory benchmarks was average, but still competitive with the Intel 975X offerings while outpacing the 945P solution. The stability of the board was excellent once it was set up correctly. The slightest change in settings when exploring the limits of the board would result in a constant reboot and CMOS clear activity that became unnecessarily irritating at times.
With that said, let's move on to our performance opinions regarding this board.
In the video area, the inclusion of a single PCI Express x16 slot does not provide SLI capability. If you require SLI capability, then the upcoming NVIDIA Intel Edition SLI-XE chipset or the recently released NVIDIA Intel Edition SLI-16 chipset will be required. The board fully supported our ATI X1900XTX video card in limited testing.
In the on-board audio area, this board offers the Realtek ALC-850. The audio output of this codec in the music, video, and DVD areas is average at best. The audio quality in gaming was acceptable for basic sound generation. If you plan on utilizing this board for gaming, then our only recommendation is to purchase an appropriate sound card. One of the new features of the NVIDIA nForce4 Ultra Intel Edition MCP is full support for HD audio. We believe it was a mistake on Biostar's part not to include an HD audio codec on their top-line board series.
In the storage area, the Biostar board offers the standard storage options afforded by the NVIDIA nForce4 Ultra MCP. The board fully supports NVIDIA's Media Shield technology and offers RAID 0, 1, 0+1, 5 capability, NCQ, and 3Gb/s support. The board also offers the standard eight NVIDIA USB 2.0 ports, but does not come with Firewire capability.
In the performance area, the Biostar TForce4 U 775 generated outstanding benchmark scores in the gaming and media encoding areas. The overall performance of the board in other areas was average, but was still competitive with 975X boards while outdistancing the 945P board. The stability of the board was excellent during testing, provided that we had properly configured the BIOS settings. At stock speeds, there were no issues, but once we started overclocking the board, it became twitchy as we explored the limits of the board.
The Biostar TForce4 U 775 is a board designed and marketed for the Intel enthusiast on a budget, yet it excels as a gaming solution. Biostar is the first manufacturer to market with the NVIDIA nForce4 Ultra solution and their results have certainly produced a board with a performance potential that belies the US$95 price target. The Biostar board performed better overall than our 945P based board and in the gaming area simply dominated it.
However, we feel that Biostar made the following errors in the design and execution of the board. The choice of the Realtek 8201CL 10/100Mb/s PCI Ethernet PHY is a grave mistake when the NVIDIA MCP chipset supports native PCI-E Gigabit operation with the proper PHY. Although most home network users will not exceed 100Mb/s operation, the lack of Gigabit capability in a board targeted to the gamer and enthusiast is not acceptable. The use of the Realtek ALC-850 audio codec is an error when the NVIDIA MCP fully supports HD audio now. The audio quality of the Realtek ALC-850 does not compare with their ALC-882 series in any area.
Our first tests with the board produced results that were below the Intel 945P boards, which is the intended competition for the NVIDIA nForce4 Ultra Intel Edition chipset. In some benchmarks, the results were up to 14% lower and the board lacked stability in most overclocking situations. Biostar and NVIDIA went back to the drawing board and released a highly optimized BIOS that improved some benchmark scores by 30%.
We were very skeptical of the performance increases that we witnessed as the new BIOS arrived with a replacement board for testing. We began to believe that we had a "cherry" board for testing and set out to discover if this was correct or not. We ran our entire benchmark test suite along with several other benchmarks through eight cycles with the same results. We switched out the video and platform drivers to previous releases and noticed at most, a 3% decrease in scores in the synthetic benchmarks while the game benchmarks were within a testing error of 1%. We also reflashed the BIOS with the public release to ensure there were no optimizations made to our version number. The test scores remained consistent.
We tested our other NVIDIA Intel based boards with the same drive image and the latest manufacturer's BIOS and driver updates that resulted in scores very similar to the previous test results. We then took the first board that we received and swapped out the BIOS chip. We utilized the same testing components in addition to the drive image and discovered (much to our surprise) the same or better results. We are confident the optimized BIOS is the differentiator in our performance results and not the board. We have been in contact with both NVIDIA and Biostar about our results. They both agree the performance increases we experienced are due to the BIOS optimizations based upon their internal test results. We received an updated bios from Biostar today that should improve the Self Recovery System and improve overclocking. We will report our results in the near future with this new bios release.
We believe that Biostar has done a masterful job in extracting as much performance as possible from the NVIDIA nForce4 Ultra Intel Edition chipset at this time. However, we feel like the optimizations in some instances have hurt the board's performance in the area of progressive overclocking and graceful recovery from going past the limits of the board. We certainly are excited about the performance potential of the NVIDIA nForce4 Ultra Edition chipset as this particular board outperforms the Intel 975X chipset in several areas while maintaining a comfortable performance lead over our 945P board. We are anxiously awaiting the arrival of more NVIDIA nForce4 Ultra boards before coming to a final conclusion if it is a serious competitor to the Intel 945P and possibly the Intel 975X products.
Sometimes, mistakes are the portals to discovery and pleasant surprises.
As mentioned on the front page, Biostar recently provided bios update, NFUIA210.BST, which solved our reported Self Recovery System issues while improving overall performance and stability at overclocked settings. We believe the bios improvements provided certainly showcases Biostar's customer service ability in this situation while giving a better insight into the true performance potential of NVIDIA's nForce4 Ultra Intel Edition chipset.