Original Link: https://www.anandtech.com/show/1079
VIA Responds with KT400A, Five Months Later
by Anand Lal Shimpi on March 10, 2003 12:19 PM EST- Posted in
- CPUs
Although NVIDIA launched their nForce2 chipset back in July of last year, it took another four months before it evicted VIA from the top of the Athlon XP market. After the lackluster performance of the original nForce, we took no chances in wholeheartedly recommending a NVIDIA chipset before seeing it in action; NVIDIA failed to deliver on their first attempt, why should we believe that they would deliver on their second try?
It just so happened that NVIDIA's second try was an extremely strong effort, and the nForce2 quickly dethroned the KT333/KT400 as the chipset to have for Athlon XP systems.
Since those days last October, we haven't heard much from VIA. On the Pentium 4 side, VIA's chipsets continue to be outsold by their SiS counterparts, entirely thanks to the legal issues surrounding VIA's ability to manufacture Pentium 4 chipsets. For Socket-A platforms, VIA's market share was slowly eroding as new users began purchasing nForce2 motherboards instead of VIA based solutions. What kept VIA around all this time was actually their lesser performing products; the C3 and all its integrated glory continue to sell very well in countries like India, and thus VIA continued to thrive, without the public presence they had in previous years.
If you remember back to how VIA got into the position of a leader in the chipset industry, you'll note that it wasn't innovation, superior performance or engineering excellence that finally thrust VIA into the limelight - it was a series of poor decisions on Intel's part to force the high-end market towards RDRAM that gave VIA the opportunity to be more than just a low-cost manufacturer. After being put in a leadership position seemingly overnight, VIA managed to adapt quite well and ended up producing some very solid designs; the KT266A & KT333 were quite possibly the most impressive chipsets to ever come out of VIA, but after being knocked out of the ring by NVIDIA, would VIA ever return?
At the beginning of November of last year we mentioned that VIA had just started the KT400A project just one month prior. Having started the project that late in the year, it was almost a certainty that we wouldn't see anything until this year, but we still saw a few reference boards come Comdex time. VIA contemplated going dual-channel DDR with KT400A and contemplated it seriously enough that when confronted about it, they even mentioned that it may be necessary, a complete 180 from their old policy on a 128-bit memory interface.
The dual-channel DDR version of the KT400A was eventually scrapped and VIA went ahead with their usual methods of updating a chipset - fine tuning the memory controller and bringing a new South Bridge to market. Designed by the same woman that has done virtually all of VIA's memory controllers, the KT400A was born - equipped with a faster memory controller and with a new South Bridge coming down the pipe.
KT400A North Bridge
In their very motherboard-manufacturer friendly nature, VIA has once again made the 'A' update to their latest chipset pin-compatible with its predecessor; the KT400A North Bridge is a drop in replacement for the KT400, which isn't really a surprise considering that the basic feature set of the North Bridge remains unchanged.
The KT400A still boasts support for AGP 8X, 266/333MHz FSB, DDR200/266/333/400 and features a 533MB/s link between the North and South Bridge. You'll realize that this feature list is virtually unchanged from the KT400, which makes sense considering the only changes to the North Bridge are confined to the memory controller itself.
Putting the marketing machine to work, VIA coined a new term this time to describe the improvements they made to the memory controller. If NVIDIA can brand dual channel DDR as DualDDR then VIA saw no harm in explaining the high-performance of their single channel memory controller with the term FastStream64 Technology.
There's nothing terribly unique about FastStream64 compared to previous North Bridge improvements VIA has made, the difference this time around is that they are giving the improvements a marketing name - not a bad idea, considering the competition they're up against. The improvements to the memory controller mainly come in the form of deeper buffers, which in turn allow data to stream in a more uninterrupted fashion from main memory to the CPU.
Remember that all memory requests must pass through the North Bridge, and every single bit of data that the CPU needs from main memory must be sent through the North Bridge. The deeper the command and data buffers in the North Bridge, the more requests can be actually stored in the North Bridge and quickly fed to the CPU, instead of causing cycle eating delays in the CPU's pipeline. The most extreme example of this is a situation where the CPU is waiting for a piece of data from main memory before completing an add instruction, if that data can't be pulled into the North Bridge because all of its data buffers are full the CPU not only has to wait the usual time for the data to come from main memory, but it must now wait longer because it must wait for the North Bridge to finally evict data from its buffer before the CPU can continue on. This obviously increases overall memory access latency because it now takes even longer to get data from main memory, which we already know is incredibly slow.
Along with deeper command and data buffers, VIA also outfitted the KT400A with deeper prefetch buffers that allow even more data to be pre-fetched from main memory into the North Bridge without specifically being requested by the CPU. The way prefetching works depends on a lot of the same principles as a CPU's cache, and the idea is that if the North Bridge prefetches correctly (it correctly guesses what data will be needed by the CPU in the very near future), data will only have to travel half as far to get to the CPU as it will be pulled from the North Bridge's prefetch buffers instead of from main memory.
More prefetch buffers means more room to store prefetched data, it is the equivalent of moving to a larger cache on a CPU; hit rate (and thus performance) inevitably improves because there's more data in the cache, or in this case, in the buffer. We're not sure if VIA actually improved the algorithms in which they use to prefetch data into the North Bridge, and given the tight lipped nature of these companies and their IP (understandably so) it is virtually impossible to make a quantified comparison between the prefetch unit of NVIDIA's nForce2 and that in VIA's KT400A.
Just like the nForce2, the KT400A will run faster with the memory and FSB frequencies in sync with each other (e.g. 333MHz FSB and DDR333, instead of 333MHz FSB and DDR400).
New Southbridge - Not yet Ready
With the KT400A VIA will eventually be introducing a new South Bridge (VT8237) sometime in Q2, however until then all KT400A motherboards will be based on the current VT8235CE South Bridge - the KT400 South Bridge.
The VT8237 South Bridge will offer native Serial ATA support, which is its main claim to fame. Native Serial ATA support means that instead of having a discrete Serial ATA controller running on the PCI bus, the Serial ATA controller will be integrated into the South Bridge and will not reside on the PCI bus, meaning that it won't be confined by the PCI bus' limited 133MB/s of bandwidth (although current drives are no where near reaching those limits). Intel will also have native Serial ATA support with their ICH5 due out in Q2 with their 865 & 875 chipsets.
The VT8237 also expands the number of USB ports supported from 6 to a total of 8 USB 2.0 ports, which will probably be the stopping point for both Intel and VIA for quite some time. Now it's up to motherboard manufacturers to make sure we have enough headers and enough ports on board to take advantage of what the South Bridges will offer.
The new VT8237 is supposed to be pin-compatible with the current VT8235CE South Bridge, however the presence of a Serial ATA interface on the new South Bridge makes us wonder if maintaining pin compatibility is possible, after all you do need the new interface pins for the two Serial ATA channels on the South Bridge.
The final improvement to the VT8237 South Bridge is support for VIA's Ultra V-Link interconnect, which doubles the North/South Bridge bandwidth to 1.06GB/s. This won't be a performance improving technology since we're not even coming close to saturating the 533MB/s of the present day V-Link interconnect, but it's not too difficult to implement and is definitely a marketing plus.
The VT8237 uses the same 6-channel audio controller that's present in the current VT8235CE South Bridge, although VIA is showing off a higher quality DAC with their VT1616 AC'97 codec that resides outside of the chipset. VIA is claiming superiority to nForce2 with their VT1616, however that's not exactly a valid comparison as the VT1616's counterpart on nForce2 motherboards is the ALC650 AC'97 codec that we see on most motherboards. In theory, a nForce2 motherboard maker could use the same VT1616 AC'97 codec on a nForce2 motherboard. VIA does not have the same processing power of the nForce2's APU because of the lack of dedicated DSP logic in the South Bridge, however as we've seen, the processing power in the nForce2 APU goes unused in most usage scenarios. What it all boils down to in the end is that VIA has a potentially higher quality AC'97 codec than Realtek, which happens to be on most nForce2 motherboards.
Reference Board
VIA sent us a total of three reference boards based on the KT400A chipset. The first two we received immediately after IDF, which were unfortunately quite rushed and were the victims of some stability issues. Because of these issues, VIA actually pushed out the launch day of the KT400A chipset and supplied reviewers with another reference board, this time devoid of all stability problems.
Our motherboard testing lab is currently receiving retail KT400A motherboards from VIA's partners, so we'd suggest waiting until seeing those reviews before deciding whether or not the KT400A is the platform for you. With this review we're aiming to provide a quick look at performance to give you an idea of whether or not VIA can live up to their claims as offering a cheaper nForce2 competitor.
The Test
|
Windows
XP Professional Test Bed
|
|
|
Hardware
Configuration
|
|
| CPU |
AMD
Athlon XP 3000+ (2.167GHz) Barton
|
| Motherboard |
ASUS
A7N8X - NVIDIA nForce2 Chipset
VIA KT400A Reference Board |
| RAM |
2 x 256MB DDR400
CAS2 Corsair XMS3200 DIMMs |
| Sound |
None
|
| Hard Drive |
120GB
Western Digital Special Edition 8MB Cache ATA/100 HDD
|
| Video Cards |
ATI
Radeon 9700 Pro w/ CATALYST 3.1 Drivers
|
General Usage Performance
Although not as performance-critical as content creation applications, it is the set of every day applications like Office and other general usage programs that the majority of users find themselves interacting with the most, thus performance here is also very important.
We start with VeriTest's Business Winstone 2002:
The Business Winstone tests are "market-centered" tests. Business applications are the popular applications employed by most users every day.
Five Microsoft Office 2002 applications (Access, Excel, FrontPage, PowerPoint, and Word)
Microsoft Project 2000
Lotus Notes
WinZip 8.0
Norton AntiVirus
Netscape Communicator
|
39.9
37.0The biggest performance advantage the nForce2 will have over the KT400A exists in business/general usage environments, where NVIDIA's excellent IDE drivers improve performance tremendously. Here the nForce2 holds a nice 8% advantage over the KT400A.
Next we have the Office portion of SYSMark 2002; the applications tested include:
Microsoft Word 2002
Microsoft Excel 2002,
Microsoft PowerPoint 2002
Microsoft Outlook 2002,
Microsoft Access 2002,
Netscape Communicator® 6.0
Dragon NaturallySpeaking Preferred v.5
WinZip 8.0
McAfee VirusScan 5.13.
|
Here we have a 7% performance advantage in favor of NVIDIA.
Content Creation Performance
For "Content Creation" performance we use two benchmarks - the new Content Creation Winstone 2003 and Internet Content Creation SYSMark 2002. We start with Content Creation Winstone 2003 and a description of the benchmark from its creators, VeriTest:
Multimedia Content Creation Winstone is a system-level, application-based benchmark that measures a PC's overall performance when running top, Windows-based, 32-bit, multimedia content creation applications on Windows 2000 (SP2 or higher), Windows 98, Windows ME, and Windows XP. Multimedia Content Creation Winstone 2003 uses the following applications:
Adobe® Photoshop® 7.0
Adobe® Premiere® 6.0
Macromedia® Director 8.5.1
Macromedia® Dreamweaver 4
Microsoft® Windows MediaTM Encoder 7.01.00.3055
Netscape® 6.2.3
NewTek's LightWave® 7.5
Sonic Foundry® Sound Forge® 6.0Following the lead of real users, Multimedia Content Creation Winstone 2003 keeps multiple applications open at once and switches among those applications. Multimedia Content Creation Winstone 2003 is a single large test that runs the above applications through a series of scripted activities and returns a single score. Those activities focus on what we call "hot spots," periods of activity that make your PC really work--the times where you're likely to see an hourglass or a progress bar.
|
The mature nForce2 platform has less than a 2% performance advantage over VIA's reference board in CCWS2K3. We can only assume that retail boards will offer similar if not better performance to what we're seeing from VIA's reference board.
SYSMark 2002 isn't the world's best CPU benchmark for comparing the Athlon XP to the Pentium 4, but it serves us beautifully as a CPU-independent platform comparison. Here's a list of the applications found in the Internet Content Creation suite of the benchmark:
Adobe Photoshop® 6.01
Adobe Premiere® 6.0
Microsoft Windows Media Encoder 7.1
Macromedia Dreamweaver 4
Macromedia Flash 5
|
Once again, a negligible difference in performance (less than 1% in this case) - the KT400A is just as fast as the nForce2 running in DualDDR333 mode.
Gaming Performance - Unreal Tournament 2003 (Flyby)
With this review we continue to use the final retail version of Unreal Tournament 2003 as a benchmark tool. The benchmark works similarly to the demo, except there are higher detail settings that can be chosen. As we've mentioned before, in order to make sure that all numbers are comparable you need to be sure to do the following:
By default the game will detect your video card and assign its internal defaults based on the capabilities of your video card to optimize the game for performance. In order to fairly compare different video cards you have to tell the engine to always use the same set of defaults which is accomplished by editing the .bat files in the X:\UT2003\Benchmark\ directory.
Add the following parameters to the statements in every one of the .bat files located in that directory:
-ini=..\\Benchmark\\Stuff\\MaxDetail.ini -userini=..\\Benchmark\\Stuff\\MaxDetailUser.ini
For example, in botmatch-antalus.bat will look like this after the additions:
..\System\ut2003 dm-antalus?spectatoronly=true?numbots=12?quickstart=true -benchmark -seconds=77 -exec=..\Benchmark\Stuff\botmatchexec.txt -ini=..\\Benchmark\\Stuff\\MaxDetail.ini -userini=..\\Benchmark\\Stuff\\MaxDetailUser.ini -nosound
Remember to do this to all of the .bat files in that directory before running Benchmark.exe.
|
The retail nForce2 board holds a 4% performance advantage over the KT400A reference board here. Assuming that the final shipping versions of the KT400A are actually of higher performance, this isn't a bad showing for VIA. Even if performance doesn't change, a 4% differential is not noticeable in real world usage scenarios, so for all means and purposes the two are equal performers.
An extremely dated benchmark, Quake III Arena has become much more of a CPU and platform test than anything because of the fact that current generation graphics cards are no where near stressed by it. We used our old 1.29f build of the game with the classic demo "four" at High Quality defaults, with everything maxed out at 1024x768.
|
Another 3% advantage for the nForce2, the KT400A continues to remain on the heels of the mature nForce2.
|
The gap remains just under 3% under Jedi Knight II, once again, not a bad showing from VIA.
|
For our final gaming test, the KT400A comes within 1.5% of the nForce2 - using only a single 64-bit memory controller.
Video Encoding Performance - DiVX/XMpeg 4.5
What was once reserved for "professional" use only has now become a task for many home PCs - media encoding. Today's media encoding requirements are more demanding than ever and are still some of the most intensive procedures you can run on your PC.
We'll start off with a "quick" conversion of a DVD rip (more specifically, Chapter 40 from the Star Wars Episode I DVD) to a DiVX MPEG-4 file. We used the latest DiVX codec (5.03) in conjunction with Xmpeg 4.5 to perform the encoding at 720 x 480.
We set the encoding speed to Fastest, disabled audio processing and left all of the remaining settings on their defaults. We recorded the last frame rate given during the encoding process as the progress bar hit 100%
|
Despite all of the close performance calls in the gaming tests, the nForce2 still exudes dominance in the MPEG-4 conversion test, outperforming the KT400A by a significant margin of 12%.
Video Encoding Performance - Windows Media Encoder 9.0
For our next video encoding test we took Windows Media Encoder 9.0 and encoded the same chapter from the Star Wars Episode I DVD into a 2Mbps VBR WMV file using Media Encoder's built in 2Mbps DVD VBR settings. The time reported is in minutes to encode, lower being better obviously:
|
The performance difference between the two platforms shrinks back to nil in the WME9 test.
Video Encoding Performance - Quicktime 6.0 Professional
Our final encoding test takes the same source file and encodes it to a MPEG-4 Quicktime file for streaming over the Internet. We used the Normal MPEG-4 settings for the conversion, and the time was recorded in minutes (lower being better).
|
Quicktime shows a 3% performance gap, but nothing as severe as what we saw in the Xmpeg test.
3D Rendering Performance - 3dsmax R5
For our 3ds max 5 benchmarks we chose all of the benchmark scenes that ship with the product - SinglePipe2.max, Underwater_Environment_Finished.max, 3dsmax5_rays.max, cballs2.max and vol_light2.max.
|
|
|
|
|
All 3D rendering scenarios seem to show equivalent performance between the two platforms; so long as there are no glaring issues in the memory controller, it's rare that two otherwise identical platforms would perform differently under these sorts of tests because of the nature of 3D rendering applications and the relatively small working set size of the scenes we're testing.
3D Rendering Performance (continued)
|
Under Maya, the performance is identical to what we saw under 3dsmax, the two platforms are identical.
|
|
|
The KT400A manages to slightly outperform the nForce2 in the Lightwave tests, but once again the performance advantage is negligible at best.
Final Words
VIA's KT400A reference board has managed to keep up with the performance of a mature and high-performing ASUS nForce2 platform that we've used for our tests in virtually all of the tests. There are a few glaring instances in which the nForce2 manages to outperform the KT400A significantly, which could be situations where NVIDIA's prefetch unit (DASP) ends up being better than VIA's solution.
Assuming the performance of retail KT400A motherboards can equal that of VIA's reference board, the KT400A will definitely be a viable competitor to NVIDIA's nForce2. The major attraction to the KT400A is the fact that it is a single channel solution and thus reduces motherboard manufacturing costs.
With the KT400A, VIA could potentially be back in the high-end game, however it will be an uphill battle to remove NVIDIA from the pedestal that nForce2 has placed them upon.
Facebook
Twitter
RSS
