Slashdot Mirror
More Juicy Dual-Processor Goodness
ronmon writes: "I was cruising around bp6.com (I still want one) and happened to see a link to some pics of dual Socket A motherboards. It's in Japanese, so I can't read most of it, but this particular board caught my eye. It's a SuperMicro sporting two sockets and five DIMM slots, plus four drive connectors (IDE RAID?). Yummy!"
And credulous reader Jim writes: "This one gave me a start when I woke up this morning. 2CPU.com has a screenshot of sisoft sandra scores from a Dual 1.53GHz (11.5x133)T-bird box. Apparently from an anonymous email. The scores are nothing short of amazing. Check it out." Grain of salt, remember.
And what does this looks like?
* Dual Athlons
* 2 Ethernets (look near the parallel port)
* 2 SCSI connectors (in the extreme left side, the picture quality sucks)
* 4 64-bit PCI slots
So what does this looks to you? a typical workstation?
Hetz (Heunique)
You get for interrupt pins per slot, but they can map all into the same IRQ. Or more likely you have about 6 PCI slots and 5 IRQs total, and IRQ assignments roll (so Pin A on slot 1 is the same as Pin B on slot 2). There isn't a hardware problem with shared IRQs on a PCI bus, so as long as the drivers are happy, it is only a small loss of performance.
And where did this all come from anyway?
Or maybe years ago....The AMD K6-III (that was years ago, wasn't it?) had a L1 and L2 cache. It went into existing K6 motherbords that had an existing L2 cache, and it worked with the existing cache, making that an L3 cache.
Not that it was all that popular.
This really is what's going on, and everyone is doing it. Memory speeds are advancing much more slowly than CPU speeds. To fight this, we see more or faster cache, more levels of cache (expect L3 to start showing up soon on IA32) cleverer cache controllers and cleverer compilers. Real benchmarks (see for instance http://www.specbench.org) show improvements from memory OR CPU speed-ups, so neither is totally bottle-necking the other at this stage.
you forgot to time the operation :)
You'd probably do better with -j4, since even -j2 will speedup a single processor compile.
Thanks! Good point. The KT series are often advertised as having ECC, but, looking at VIA's spec sheets, the KX actually has it.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
A lot of 133Mhz motherboards based on VIA claim to "support ECC"... but don't actually do any error checking or correction. I'm not sure where this came from; the VIA specs don't list ECC as a feature, but the ABit VP6 specs *say* the board supports ECC. It doesn't.
So, if you want a nice, stable system, be very wary.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
Sure, but for a tinkerer, it's a pain to have to go and buy a new PCI card when you just want to plug that ISA card in for a minute...
I've got IDE now, but I wanted to grab data off of an old SCSI drive (old - 5.25, full height) that I had kicking around. I had to go set up my old 486 because I don't have a PCI SCSI card, just an ISA and a VLB one. That ISA SCSI card is nowhere near fast, but considering the drive and the short-term nature of the task, it would have been good enough.
I also wanted to test an old sound card, an Awe 64, that I was giving to someone. I couldn't plug it in. Major pain. If I wanted to bring one of the old A/D boards home from work and fiddle with remote sensors, I couldn't because they're all ISA.
If I could replace any ISA card with a PCI card, at midnight, without leaving the house, within five minutes, that'd be okay. But if I can't, it's not so handy.
This implies to me that if you want a dual-processor laptop, you'll have to have two CPU's with less power (each) than a single-CPU laptop.
Transmeta's chips (and the G4) both use much less power than Intel/AMD offerings. The G4 is also smaller (don't know about Crusoe). That's why I think AMD would need some new technology to pull it off. There's also all the cool shit IBM has (0.01 micron process in the lab, SOI, etc.).
Apple has demonstrated the possibilities of G4 laptops (with no fans, I believe), so a dual G4 or Crusoe laptop should be possible with clever engineering. Phase change heat pipes might be useful, for instance...
...but can you imagine a Beowulf Cluster of these? [Ducks for cover]
Those are very cool (figuratively), but I bet they put out tons of heat. Someone mentioned that the board appeared designed to fit in a 1U case. What would it take to properly cool it in such a small enclosure?
What would be even cooler is if AMD bought Transmeta and made dual-CPU laptops (or at least chips suitable for use in said laptops). Then there might be something out there cooler than Apple's Titanium PowerBook G4. Multiprocessor laptops... Drool.... For those occasions when you need to perform advanced simulations of nuclear weapons tests while on that long redeye flight. (Can anyone else actually think of a good use for that much power?)
Rader
So I'm sure that as long as there's a market for putting an ISA slot in, some MFG's will continue to do so.
As long as there are options to buy boards with ISA slots on them, *I* want the option to buy boards with PCI only on them. Well, I guess this really isn't an argument until the MFG's take our choices away from us........see you then! :)
Rader
Please excuse my ignorance, I haven't followed hardware developments very closely for some time. ... could somebody enlighten me what news there are other than the raw clock speed? (and no, I'm not trolling, I really want to know. otherwise I would have checked 'post anonymously' ...)
So those Thunderbirds have a multiplication factor of 11.5 over the system clock? And over the RAM, if it's SDRAM133 (or whatever that is called). Even with DDR that still makes a factor of about 6, which imho is downright ridiculous. How well do these things score in real applications, not just cycle counting in first level cache (aka MIPS and MFLOPS)? I can't imagine you get much over a 800 MHz system
EagerEyes.org: Visualization and Visual Communication
This SUCKS
There are many ISA cards that cannot be found in PCI form. The RealWeasel is one of them. Many aquisition cards and even do-it-yourself ISA boards (as the ISA signaling/integration is MUCH SIMPLER than the PCI way)
We DEMAND an PCI2ISA converter with an external ISA card cage. (alltrough i'm not sure it will work as the ISA bus has access to all irq's and the PCI slots only to the four INTs assigned to them.)
--
1% APY, No fees, Online Bank https://captl1.co/2uIErYq Don't let your $$$ sit in a no-interest acct.
-----------------------
Nicotine free Amish .sig.
-----------------------
Nicotine free Amish .sig.
The only difference I can see is that mine has 4 DIMM slots as opposed to 5. I can't tell from the picture in this article whether it's a socket for PIII or Athlon though. Considering it's got a VIA chipset that looks just like mine, I'd say this motherboard is just a dual PIII board.
However, it totally rocks. :) To sum up a few of its features:
- Dual IDE buses (yes, it does have 4 IDE ports), one at ATA 66, one at ATA 100.
- Firewire and USB onboard
- Nice clocking facilities
As well as the usual neat bits, and onboard sound. Recommended if you need a PIII board, specially dual processor.Fross
I'd have to agree. You can clearly see that the two dimm slots nearest to the processors are different than the other three. The two closest to the processor appear to be used for DDR SDRAM, due to the single notch, whereas the other three appear to be for standard SDRAM, due to the two notches.
Not likley, I dont know about the pc motherboard market specifically but in other fields of electronics manufacturing it isn't terribly uncommon to have 6, 8, or more layers to a pcb. It may be a little more tricky for somthing the size of a full ATX mobo but it's not uncommon. It shouldn't be teribly expensive in production quantities either (~$20 per board in ten-thousand quantities at a guess).
"Prefiero morir de pie que vivir siempre arrodillado!"
Ok, I would like to say that no one thought of using the babel fish japanese to english translator. That site was not entirely about a dual socket A motherboard. The supermicro board you saw was the 370ddi board, a new dual p3/celeron board that has 5 SDRAM slots, not DIMM slots. The actual dual socket A part was very small and only mentioned the Tyan board. Please translate before you go posting something like this, because it just creates a lot of hype for things that are false.
When I worked for Corel (in Ireland on Linux Support, not Canada on code, remember that let's you know the sort of people I could get too and hence how much the following is about personal opinions rather thatn corporate policy, it's just what I think), I tried (in the limited ways that I could) to get them to do just this! I suggested that if they can give away free to use WP8 than surely they can dig back somewhere and give away a Free version (say the first draw+wp versions released for unix). Alas, while a few people thought perhaps, they generally couldn't grasp any benefits to them. As for GPL or even Open Source WP9/10 ... I don't think they had the stomach for that AND I don't think they thought anyone else would have a clue how to work with the code ... I'm not even sure if they knew if they could work with the code. When I heard talk of a re-write of some of the programs I always suggested it again (if anyone who might be worth saying it to and I knew wasn't against the idea was around) because it could offer someone like them the opportunity to ditch being toolkit writers and let them focus on the important stuff.
Never underestimate the dark side of the Source
IDE RAID on the MOBO has already been done.
:)
:(
Check it out here: Abit KT7A-RAID. It's got the on-board IDE RAID you dreamed of, AND it's been acclaimed on Anandtech... their mobo pick for many of their high-end systems. Ultra ATA-100, 3 DIMMs, 1 AGP, 6 PCI and you even get to keep your one ISA card
Sadly, no dual processor support yet
Download your mp3s any way you want, and support the artist via FairTunes
I used to talk directly to the head of the tech support and he was also in charge of overseeing development. We had some customers want some really particular bios settings and tech support said they would burn us custom bios if we really needed it. Have you actually tried updating the bios? I built a Lightwave animation box with a P6DBU and a 3d Labs Oxygen RPM video card. There was a conflict with the bios. I jumped on their site and found the updated bios right away. Their tech site is good with updated bios, pdf manuals, etc.
Actually all those Abit boards (& Epox & IWill, etc) that have an extra IDE controller, so that they support 2x4 IDE drives (instead of the normal 2x2 IDE drives) suport RAID (0,1 & 0+1) on the extra 2 IDE connectors.
Abit & Epox used a Highpoint controller, IWill uses an AMI controller, while Asus uses Promise (which on some Asus boards has to be bios hacked for RAID). All 3 brands of controller have their good & bad points.
Looks like it has 5 x 64bit PCI slots, and the four DIMM sockets are slanted, perfect for a 1U case.
What I am interested in is some info on some older multi processor boards, just so that I can get the last bit of milage from some older cpus I have kicking around.
But then I remember advice I used to give some retail customers, half in jest:
Remember, if you can buy it here at (mass market store), technically it is obsolete already.
But then, some people love living on the trailing edge of technology.
"It is a greater offense to steal men's labor, than their clothes"
I presently have 5 PCs, 4 of which were, when new, the biggest heat generators available. (Yup, it includes fast Athlons and an Alpha). It's -20 (centigrade, that's -5F) outside today, yet _none_ of my radiators are on! I heat my flat with PCs!
FatPhil
-- Real Men Don't Use Porn. -- Morality In Media Billboards
Also FatPhil on SoylentNews, id 863
Take a closer look. The first two slots closest to the processor are slightly larger. The keying notches are also in different places. Most likely DDR slots, but the board still has SDRAM slots for those who don't want DDR; now that's a nice feature.
Now, all of this stuff is well and good (dualie Athlons), but AMD needs server-class boards. That means things like built in SCSI, 64 bit PCI, etc. AMD has wanted the server market since they originally introduced the Athlon (and everyone wanted it since then) so it should be here soon.
..not faster cpu's that hide the flaws. Case in point "Wordperfect 5 (or 4.2)".
--
Jon - TheSpork
There this 760MP based board from Tyan shown in the article. Only four DIMM's though...
All these boards look fine to me. But what is more apperent now is that SMP is truly a dead end when it comes to multi processor systems. Intel has bled it to the last drop and others such has Sun and IBM have taken SMP to new bounds that are still not matched in the PC industry.
But, right now we are heading in the direction of massively parallel comuputing done using non SMP meothods. Good clustering practices has made SMP nearly obsolete. Mix that with the non ability for non intel vendors to come up with a motherboards for non-intel CPU's on the PC market, I feel that the end of SMP has been well written in concreate and hardended.
Mix that with good software such as Mosix which lets you parallel process any software without specfically having to compile the software into a traditonal beowulf type messging system (PVM/MPI), we are now on the verge of having home grown Cray system in every basement.
Enjoy
Trust the source!
I know there's a lot of hype behind supermicro, I know they have really cool feature lists.
But I know too many people who have paid a premium to own a supermicro product only to find that they were slightly flaky, that the super nifty features didn't work properly, and that the boards were quickly orphaned (support discontinued) when newer boards were released.
I won't buy 'em, and don't recommend 'em.
This is just like television, only you can see much further.
I found this over at amdzone I agree:
2CPU has what they are calling Dual 1.53GHz Athlon scores on a Tyan board. Well, there are a couple of problems with that which make me very, very unsure that these are legit. First of the Tyan 760MP board does not have overclocking features. I know, I held one in my hands at Comdex. So you think that they might add clock multiplier features in the meantime? Nope, what they had was the final revision of the board which only needs a final chipset from AMD to be complete. Second, it is very difficult to get a 1.2GHz Athlon to even 1.4GHz, much less 1.53GHz, and then you are telling me they got two to go that high? Lastly there are no details about this system at all except that it is using the Tyan board. Who knows, maybe it could be right, and I'm not saying 2CPU is making it up, but there are not enough details and not enough evidence, and there is too much logic keeping me from believing it. And for the fools that will say that I am jealous because I don't have dual scores to post here, don't even bother e-mailing me. That is ridiculous.
Pretty much says it all regarding the benchmarks we "saw".
Try to hack my 31337 firewall!
Comment removed based on user account deletion
________
Does anyone actually have a Java program designed to control air traffic, or for the operation of a nuclear facility?
- Installing the latest RPM or DEB packages, doing a bit of config work via rsh, and being able to scale to 64 way by putting 64 boxes in the rackmount unit, or
- Designing a motherboard that integrates together 64 CPUs?
SMP certainly provides the benefit of very fast communication between CPUs, so if processing is strongly dependent on that, SMP wins.But it is most certainly a lot more expensive to scale it up. The vendors may sell 2-way and 4-way SMP motherboards for not overly princely sums, but moving on to higher multiples guarantees pretty monstrous prices, because you're simultaneously mandating:
- Small production runs, compared to the teeming hordes of single CPU motherboards, and
- Sitting at the bleeding edge, because nobody wants a souped up version of a CPU that was "state of the art," LAST YEAR.
Kernel compiles aren't liable to benefit all that much from parallelism, moving forward; consider that with a Pentium III or Athlon, it doesn't take long to recompile Linux even from scratch. It's just getting to be less and less an issue.As for benefiting from other forms of parallelism, it is entirely likely that the toolsets surrounding Beowulf and PVM will improve over time to make it easier to manage doing "clustered tasks" in much the same way that we have progressed from having rather primitive "package management" tools to having stuff like AutoRPM, apt-get , and BSD Ports.
If you're not part of the solution, you're part of the precipitate.
The 5 DIMM slots are almost certainly a combination of 2/3 with 2 being one type, and 3 of another (between SDRAM and DDR-SDRAM).
The 760MP has been more than a passing interest to me lately, and I've been digging up information/rumors about it daily for the past week. This is what is sounds like:
One EV6 bus requires a hefty chunk of PCB. Two will require even more (of course). This means that DP Athlon boards will require as many as 8 layers to fit into an ATX form factor. IIRC, the industry standard is 4 or 6, so this would be a new (and likely expensive) manufacturing process that may require new tooling to produce in bulk.
Athlon boards are already more expensive than P3 boards, and I think the overall DP price/performance comparison isn't going to be that bad for Intel when/if 760MP ships.
--
Maybe peyote juice?
But then again we must remember that this story has been promoted to front page material by the same group that brought us nanopants.
That's not a grain of salt the editor mentions, that's a rock of crack ...
But I digress ...
Why do I strongly suspect this is non-authentic? Does anyone else remember the photograph of the modified Duron that was supposedly being produced by AMD to thwart overclockers. A pin was physically "removed" from the pin interface.
The picture was posted all over the place. Everyone was all up in arms. It was the end of all things.
And then someone noticed that this "modified" processor had the same serial number as an unaltered promotional processor photo from another web site.
NEW FLASH: Overzealous Tech Sites Taken in by Paintshop Pro Forgery ...
Many Japanese tech sites are notorious for posting outrageously altered faux benchmark screens. This would appear to be another one of those posts ...
Multiprocessor Thunderbirds will rock, I am certain. But this is as ridiculous as the recently reported SMP Thunderbird Linux kernel compile that supposedly demonstrated a greater than two times increase in speed between one processor and two ... Yup.
Why couldn't I find people who believe things like this when I was selling electronics? I'd have made a fortune in commission. :)
Get off my virtual lawn, you damned virtual kids!