SPCR’s review of a pre-production 1 GHz VIA EPIA-M10000 Mini-ITX board with the new Nehemiah C3 core. Ten thousand is the next step after M9000, the last released EPIA board reviewed here late last year. It is more than a marginal clock speed upgrade, says VIA. Our sample turns out to run at 933 MHz, so we had a head-to-head between Nehemiah and Ezra-T at the same speed.
March 18, 2003 by Mike Chin
| Product | VIA EPIA-M10000 Mini-ITX Mainboard |
| Manufacturer | VIA Technologies Inc |
| Supplier | VIA Mainboard |
| Market Price | ~US$175 (Pricewatch) |
We’ve been playing with a pre-production sample of the VIA EPIA-M10000 Mini-ITX board. Ten thousand seems like a pretty big model number; it’s the next step after M9000, the model number of the last released EPIA board reviewed here late last year. As many of you may already know, the M10000 features a 1 GHz version of the C3 variant processor used in VIA’s mini-ITX boards. It is more than a marginal clock speed upgrade, however. It features a new core code-named Nehemiah that VIA says is faster in many ways than the previous Ezra-T core. More on that later.
No visible surprises with the M10000; still the same tiny 6.5″ square footprint.
For the benefit of readers unfamiliar with the Mini-ITX platform by VIA, here is a summary from VIA:
A low power, high performance, and highly integrated x86 platform that provides the most flexible, compatible, and cost-effective solution for building the emerging new generation of connected digital information & entertainment devices. It combines a proven ultra low power sixth generation processor core with a choice of a variety of highly integrated North Bridge and South Bridge chips, as well as a broad spectrum of expansion options for enhanced communications, connectivity, and multimedia functions.
The EPIA M series boards, driven by the newer VIA CLE266 chipset, are designed to tackle a different market segment than the original EPIA series with PLE266 chipset, which will continue to be produced and sold. The multimedia performance improvements in this second generation Mini-ITX board allows VIA to position the EPIA-M series for use in a home entertainment PC.
Detailed specifications on the M boards, straight from VIA’s web site:
| VIA EPIA-M Mainboard Specification | |
| Processor | – VIA C3/EDEN EBGA Processor |
| Chipset | – VIA CLE266 North Bridge |
| System Memory | – 1 DDR266 DIMM socket |
| VGA | – Integrated VIA CastleRock AGP graphics with MPEG-2 decoder |
| Expansion Slots | – 1 PCI |
| Onboard IDE | – 2 X UltraDMA 133/100/66 Connector |
| Onboard Floppy | – 1 x FDD Connector |
| Onboard LAN | – VIA VT6103 10/100 Base-T Ethernet PHY |
| Onboard Audio | – VIA VT1616 6 channel AC’97 Codec |
| Onboard TV Out | – VIA VT1622 TV out |
| Onboard 1394 | – VIA VT6307S IEEE 1394 Firewire |
| Onboard I/O | – 1 USB connectors for 2 additional USB 2.0 ports |
| Back Panel I/O | – 1 PS2 mouse port |
| Application | – VIA FliteDeck? Luxurious Utility |
| BIOS | – Award BIOS |
| System Monitoring & Management | – CPU temperature monitoring |
| Form Factor | – Mini-ITX (4 layer) |
The EPIA M10000 is virtually identical to the M9000 except for the HSF and CPU. The visible changes:
- A lower airflow, quieter fan than was found on the M9000, a detail that can be discovered only by examining the fan label.
The M10000 uses a quieter, lower airflow 40mm fan than the one used on the original M9000.
Apparently this quieter fan will be used on all the EPIA boards.
- A slightly larger HS than the one used in the M9000.
Stacked for comparison: The M10000 (top) uses a larger HS than the M9000.
NEW HEATSINK FAN DETAILS
The Bi-Sonic BS401012M is a 40 x 10 mm 12VDC fan rated by the manufacturer for 6.5 CFM airflow at 5000 RPM and a noise level of 24.5 dBA @ 1 meter. This is some 8~9 dBA quieter than the fan on the M9000, so it is a welcome change. A quick SPL check at the anechoic chamber at UBC confirmed the reading: ~25 dBA @ 1 meter was obtained.
Subjectively, the new fan seems even quieter than the substantial -8 dBA measured difference, because the earlier fan on the M9000 has a distinctive whine while this one whines a lot less. It sounds a bit louder at 12V than our reference Panaflo 80mm L1A fan. The audible whine level is very similar! This is kind of crazy because the Panaflo is only spinning at 1900 RPM and the Arkua is at 5000 RPM. Perhaps the larger diameter of the Panaflo blades actually have higher velocity at the tips than the small 40mm fan’s blade tips, and thus causes more whistling air turbulence?
Still, we do not consider the Panaflo quiet enough until undervolted by 3-4V or lower (ideally 7V or lower). The new 25 dBA BI-Sonic fan is a big improvement over the noisy AVC first used in the M9000, and quieter than 90% of current CPU cooling fans or even video card cooling fans. In our view, it’s not enough for VIA to rest its laurels on. That fan probably needs to be undervolted by 2-3V for most silent PC enthusiasts to consider quiet enough. In the coolish 20C lab, undervolting the fan to 9V using a Zalman Fanmate 1 fan voltage controller made no apparent difference in stability or performance, even inside a very compact mini-ITX case (yes, we finally have one! — more on that in another article), but it does bring the noise down to a hush.
VIA informed us that the fanless solution used in the EPIA M6000 could not be adapted to the faster CPU models. As the photo below shows, the M6000 sports a much larger aluminum heatsink. A copper version of this device might have sufficed for the M9000 and M10000.
M6000 board, with a much larger aluminum HS, does not sport a fan.
Interestingly, Mini-ITX.com reported in late February,
The long awaited EPIA M10000 has apparently reached some US online stores, but not in the form expected. The C3 powering them has the Ezra-T core as found in the EPIA M9000, not the newer Nehemiah core as initially advertised (retailers have since changed their specifications). The FSB runs at 100Mhz, and not 133Mhz – and LVDS support does not appear to be present. Cooling is improved with a double North/Southbridge heatsink and meatier CPU heatsink. It looks like VIA hit manufacturing delays with the Nehemiah and decided to quietly release a M10000 anyway, with an “M10000-II” (our name, not theirs) to follow later, with more fanfare. We’ll wait for the real one…
The link above to the photos on the Japanese Akiba PC website shows the same CPU heatsink used on our sample. One assumes that the same BI-Sonic fan is used. All this suggests that VIA is juggling fast and hard to resolve both cooling and noise issues optimally. They have a reputation to maintain. After all, VIA have been pushing the quiet computing envelope for some 2 years now.
NEW NEHEMIAH CORE
Physically, the CPU change from the M9000 is found only by removing the CPU heatsink: details are printed on the CPU casing (“133 x 7.5” and so on). Curiously, the review sample was identified by both Windows XP and Sisoft Sandra 2003 as a VIA C3 933. It’s likely that in their haste to get this pre-production board to us, VIA neglected to update the multiplier in the board. Flashing the BIOS to the latest version,10B, did not change this incorrect data. Windows XP does identify it clearly as a Nehemiah.
Taking some advice from mini-itx.com, WCPUID by H.Oda was used to try and change the CPU multipler. The utility did not recognize the board or CPU, even though the correct information was displayed. The change multiplier feature would not work on this EPIA M10000.
An email was fired off to VIA to determine how and whether the CPU could be made to run at the correct speed, but everyone there are likeLy too busy at CeBit in Germany, reputed to be the biggest computer show in the world. Any response from VIA pertaining to this issue will be published later as a postscript.
For the time being, it is probably safe to say that the benchmark results for a production-run 1 GHz M10000 Nehemiah would probably be a bit better than our 933 MHz sample. This problem is actually something of a benefit: we get to pit the Nehemiah core against the Ezra-T core at the same speed!
IMPROVED PERFORMANCE
In one of their features / benefits pages, VIA provides a comparative graph showing the difference between a 1 GHz Ezra-T core C3 and a 1 GHz Nehemiah C3 on a VIA C3M266 main board, which uses the same Apollo CLE266 chipset with integrated VGA as the EPIA M boards. The differences appear significant, especially in 3D Mark 2001 SE.
The Nehemiah core does much better at 3D imaging than the previous Ezra-T.
This data needed to be verified, so a system was assembled by putting components onto an open frame platform, and then the motherboards tested with these components, one at a time. As the chipsets and video cards and all “peripheral devices” are identical, there was no need to make any driver adjustments when the motherboards were swapped.
At the last moment, an M6000 was also thrown into the fray for curiosity’s sake. Let’s check out the whole M family at once, shall we?
| EPIA M10000, M9000 or M6000 |
| PC2700 DDR RAM, 512 meg |
| Seagate Barracuda IV 40G hard drive |
| Nexus NX3000 power supply |
| Hitachi DVD 16X drive |
| Windows XP Pro SP1 |
| Samsung 955DF monitor |
| PowerDVD XP 4.0 |
The EPIA test rig.
BENCHMARKS
Sisoft Sandra 2003 was used to run some benchmarks. The improvement between the M10000 and M9000 in the SANDRA multimedia benchmark is 89%, even higher than the 74% obtained by VIA in 3D Mark 2001. The 22% gain in the CPU Arithmetic score is also interesting. This is despite the 2 boards running at the SAME clock speed.
Between the M6000 and M9000, one would expect to see only straight line gains based on the 50% increase in clock speed, but this is not the case. With CPU Arithmetic, the gain was 57%; with Multimedia Integer, it was 76%
Memory Bandwidth, which is one of the most often cited Sandra benchmarks, was pretty much the same for all three boards: 220/250. Wiping the floor in terms of comparisons against current chipsets for Intel and AMD. (An aside: Why did VIA move the board from SDRAM to DDRAM when the M series boards’ bottleneck is clearly not memory bandwidth? The reason is a practical one: DDRAM is becoming the default standard for memory everywhere. Volume determines price and availability. Already SDRAM prices are a bit higher than DDRAM.)
| PARAMETER | M6000 | M9000 | M10000 | % gain* |
| CPU Arithmetic / FPU | 770 / 211 | 1212 / 330 | 1483 / 343 | + 22% / + 4% |
| MM Integer / Float iSSE | 631 / 1195 | 1113 / 1836 | 2104 / 2126 | + 47% / + 14% |
*NOTE: % gain is the improvement of M10000 scores over M9000
The Cache / Memory benchmark results are really interesting. Sandra contains typical benchmarks of a broad range of CPUs and chipsets in its database so that the system or CPU under test can be compared. Because all the current CPUs from AMD and Intel are running at speeds well in excess of 1 GHz, older comparison CPU/chipset combinations were chosen from Sandra’s database.
933 MHz Ezra-T EPIA M9000 (blue line) scores 2.5 times worse than its closest competitor, a long discontinued P3-750 on the Intel 815 chipset. (Click on the image for a large view in a new window.) The EPIA M6000 scores about 40% lower.
The 1GHz Nehemiah EPIA M1000 (blue line) makes a miraculous improvement over the M9000, moving up to the middle of the pack, a 300% better performance than the M9000. (Click on the image for a large view in a new window.)
REAL WORLD PERFORMANCE
So how do all the above benchmark improvements translate in actual use?
Since the biggest improvements are in multimedia, DivX and DVD playback were used to see, hear and otherwise perceive difference between the Nehemiah and Ezra-T core versions. A 19″ ADI monitor (a good performer with up to 1600 x 1200 resolution) and a Panasonic GAOO 27″ TV were used. High resolution Grado SR-60 headphones were used, along with an assortment of typical multimedia PC speakers.
Generally, no significant differences were observed between the two boards for video performance on the monitor. Divx and DVD playback were both very nice, smooth and detailed without frame dropouts or artifacts. Better than TV, in general, though because one sits further back with the large screen of a TV, it is hard to say for sure. In any case, if the monitor was just a bit bigger so I could sit further back comfortably, I’d personally be happy to use the EPIA M for watching DVD movies. No attempt was made to multitask — i.e., check email while watching a movie, etc. (It just seems silly…)
Only a small number of DVD movies and Divx files were used. Perhaps these were not demanding enough of the playback system. It is possible that a very demanding video clip could have showed more of the difference.
Observing CPU utilization, the M10000 might be very slightly more efficient (maybe 5% average?) — it was hard to tell, as the difference was small. AC power consumption, as measured with a Kill-a-Watt meter, seemed very slightly better for the M10000, and likely directly proportionate with the possibly reduced CPU load.
The sound quality on both boards was the same on all the media tried, with all speaker options. Again it is possible that more demanding sources — or a full multi-channel speaker system — could have shown more of a difference. Suffice it to say it was certainly satisfactory for the purpose.
The same generalization can be made using a TV for video: Not much difference between the two boards. Neither were very good, not any better than a VCR. But this is what I have also found with non-integrated video cards on AMD XP and Intel P4 systems. Playback over TV of DVD or Divx is not very good through the S-video link.
The bottleneck may be the Panasonic TV, considered fairly high end a couple of years ago but lacking the higher definition of the latest and greatest TVs today. It could also be the generic S-video cable. I have been told that a better quality S-video cable may improve matters. I have yet to see good enough PC-to-TV video performance in my admittedly limited experience to believe it desirable in any form. It would be most interesting to see what a really good PC-to-TV setup could do.
CONCLUSION
The Nehemiah EPIA M is a welcome advance in the growing Mini-ITX world. Its higher processing power and efficiency will surely be beneficial, especially as clock speeds increase. The new quieter fan is a much needed improvement, and as we have seen, VIA is likely going to continue juggling cooling solutions to keep noise to a minimum as their platform ramps up in speed and power.
When all is said and done, the sheer space and electrical power efficiency of this little device is hard to beat. The photo below shows an EPIA M10000 system encased in a slim black Morex Cubid 2699 case along side a standard size mouse and the ADI 19″ monitor. The case stands only a foot tall, and it is hardly bigger than a thick 3-ring binder, or about the size of two notebook computers stacked.
The comparison to a notebook is not a stretch. The Morex case sports a PSU that uses a small external transformer encased in plastic. The highest AC power draw, during a cinematic climax in an action movie DVD, was 49W. A reasonable guesstimate on the best power efficiency of this PSU is 60%. This means the entire system was pulling just 29W (DC). In idle, the AC power draw dropped to just 27W or ~16W DC power draw. This is in notebook territory! My Dell P3-866 notebook draws 25W on active idle and 35W peak on DivX playback (which consumes less power than DVD because the DVD drive is not working).
That wraps things up for now. Look soon for a review of two mini-ITX cases, including the Morex shown above.
Much thanks and appreciation to VIA for providing us the review sample and for their assistance with relevant information.
* * * * *
POSTSCRIPT – March 21, 2003
The odd clock speed of the sample Nehemiah EPIA-M board and my inability to see any real world improvements over the Ezra-T M9000 board raised questions for some readers. This led to a lengthy exchange with VIA personnel now back from Cebit, exhausted. It helped to clarify a few key issues:
There is absolutely no question that the sample IS a Nehemiah core erroneously locked at the wrong clock speed. They agree that this made for a better apple-to-apple comparison against the M9000 board, allowing the cores to be compared without the interference of differing clock speeds.
This page about VIA C3 power efficiency tells us that that the peak power dissipation of Nehemiah core at 1 GHz is 15 Watts. In contrast, an Intel Celeron at the same speed is said to draw 27.5W. What the page does not tell us is the fact that the Ezra-T at 1 GHz draws 18W. The Nehemiah draws 3W less power; it is more powerful yet consumes less electrical power — a neat trick. My observations above about slightly decreased power were not imagined.
VIA says the improvement in multimedia processing power is most easily seen in the creation of video files. The time it takes with the Nehemiah is considerably shorter, apparently, than with the Ezra-T. I cannot confirm this at this time as such software is unfamiliar to me, but have little reason to doubt the claim. When time allows, we will return to this issue in future.
Finally, VIA says:
“The Nehemiah is a new more efficient core that provides more performance for general tasks while still maintaining quiet operation, even higher power efficiency, a low profile and great DVD and audio. The fact of the matter is also that the Nehemiah core will take us to higher CPU frequencies later this year, where the Ezra-T had basically reached its ceiling.” (emphasis mine)
* * * * *
Discuss this article in our forums.
