TR Forums

question:
How the @#$3$^@!^% does someone tell BEFORE if a CPU will work with a Shuttle SFF? Going by CPU specs on Shuttle's website clearly isn't enough.

http://www.shuttle.com/hq/support/faq/sff/sff.asp

back story:
I just got bit in the butt building a SFF for a friend who wanted it as low budget as I would spend and have confidence in the system.

So I ordered a Shuttle SB51G (Intel chipset), 512 Meg PC2700 Corsair memory, and.... the big mistake... a Celeron "D" 320. Now, that Celeron is rated at all the proper specs for that shuttle cube. But it won't work. No post, no beeps, just lights on and nada. Memory tested out as good.

Googling produced this:
http://forums.sudhian.com/messageview.c ... SGDBTABLE=

So. How do I avoid this mistake (and having to eat the cost of this CPU from NewEgg) in the future? How do I know what freakin' GENERATION of CPU a cube will take? Just following the FSB and pinout specs aren't enough, obviously.

(And, conversely, how do I know what Shuttle SFF will take a Celeron "D" or other Prescot core processors?)

*sigh* Help? I do NOT want to make this mistake again.

Go AMD next time.

Check the CPU support list on the Shuttle website!

http://www.shuttle.com/hq/support/faq/s ... 1G%20(FB51)

Return it and get a real p4. Even a 2.26 P4 will beat a 2.6 celeron for not that much more money.

Also when building computers for people.....DO YOUR RESEARCH. People are trusting you with their money so don't let them down. I have built about 3 computers for friends of mine in the last few months and I swore I would never do it again. One wanted to spend only $1000 on a computer that could play Doom 3 and Half Life 2 with everthing on. Aghhh! I also built a shuttle system for my girlfriend but I chipped in the extra cash so that she could get her cute little case. Another one wanted a Cad/3dmax workstation for super cheap. Never again.

apsog33 wrote:

Return it and get a real p4. Even a 2.26 P4 will beat a 2.6 celeron for not that much more money.

Per
http://www.newegg.com/app/faq_return.asp
they do NOT cross-shop CPUs and they do not return RMA items for different items. I'm stuck with this CPU (as an unused spare of sorts until I buy another rig for someone else). So "return it" isn't an option.

apsog33 wrote:

Also when building computers for people.....DO YOUR RESEARCH.

Well, that's the question, isn't it. HOW would I do my research when the specifications of the CPU match that of the list on Shuttle's website?!? There's no mention of the critical issue of 'Prescot' vs 'Northwood' anywhere by anyone except in techie forums.

apsog33 wrote:

People are trusting you with their money so don't let them down. I have built about 3 computers for friends of mine in the last few months and I swore I would never do it again. One wanted to spend only $1000 on a computer that could play Doom 3 and Half Life 2 with everthing on. Aghhh! I also built a shuttle system for my girlfriend but I chipped in the extra cash so that she could get her cute little case. Another one wanted a Cad/3dmax workstation for super cheap. Never again.

This is lame lecturing that I can only assume is meant for general forum readers. I've built systems for people for over 6+ years (mostly myself) and never had an issue like this crop up. Checking CPU FSB and pinouts was always enough before. And there was nothing this time around to indicate to the contrary from NewEgg or Shuttle's websites; which is the nature of my complaint.

d_rogue wrote:

Check the CPU support list on the Shuttle website!

http://www.shuttle.com/hq/support/faq/s ... 1G%20(FB51)

The link I gave is the "mother" page to that one. I checked that list. It mentions NOTHING about CPU cores. The CPU I purchased legitimately fills the specifications on that webpage.

*pause*

Oh @#$@WDS. There, at the bottom, in tiny print, is the clarification. I just wonder if they put that up AFTER I did my purchase, or if I missed it the whole time and was a goofball. I'll never know.

lesson 1) Shuttle does recognize the problem and will address it on that techincal website at SOME point. (I'm not convinced of their speed in recognizing the issue and posting the clarification.)

lesson 2) I need to quit putzing with Intel stuff and just make the jump to AMD. There's just no excuse anymore...

ciRCuSSIdeSHOw wrote:

Go AMD next time.

Duly noted. Nobody makes motherboard chipsets like Intel, unfortunately. I tried a VIA one a couple years ago and was roasted by their 4-in-1 drivers. I'll give nVidia a try when I go AMD. Neither of them seem to have people blazing with confidence and extacy from a motherboard driver perspective, however. *sigh* (not a perfect world)

Actually, the Celeron D doesn't match the specs listed on the compatibility chart. Note that the Celeron 2.6GHz listed there has a 100MHz bus speed. Celeron D's have a 133MHz bus speed.

within 15 days, you should be able to get your $$ back...or at least most of it, I hope for my own personal sake

Dr. Evil wrote:

Actually, the Celeron D doesn't match the specs listed on the compatibility chart. Note that the Celeron 2.6GHz listed there has a 100MHz bus speed. Celeron D's have a 133MHz bus speed.

Out of respect for the TR staff, I'm hesitant to argue...
Shuttle website has both 100FSB and 133FSB has happy possabilities

Item #6 (going from top to bottom) on that list on

http://www.shuttle.com/hq/support/faq/s ... 1G%20(FB51)

133MHz 18x Intel Pentium 4 socket 478 2.40 GHz

and item #7 on that list

100MHz 24x Intel Pentium 4 socket 478 2.40 GHz

looks to me like it matches the stats on a Celeron D from NewEgg:

http://www.newegg.com/app/ViewProductDe ... 197&depa=1

NewEgg doesn't mention the FSB speed, but that is irrelevant (I mean, I can do the math myself, it isn't hard) since both FSB speeds are listed. The only warning is the small print on the bottom of that page:

2. Prescott/Celeron D processor are not supported

and I don't know when that was put up there. I COULD have missed it. Honestly. I admit I screw up. I just think this was an EASY one to screw up.

I suppose I should just quit being a grumpy bastard and tip my hat for Shuttle noting the issue at all.

Usacomp2k3 wrote:

within 15 days, you should be able to get your $$ back...or at least most of it, I hope for my own personal sake

I'm just going to build a second Shuttle SFF for the 'rents for Xmas. They're still sitting on WinME. I'm almost morally obligated to move them along, so here comes lemonade from a lemon....

And yes... I will let this thread die.

Thanks to everyone who read, and especially those who bothered to reply to my posts in this thread. Please don't mistake my disgruntled demeaner with ingratitude. I'm very grateful for the TR forums and the goodness that happens therein.

MaceMan wrote:

ciRCuSSIdeSHOw wrote:

Go AMD next time.

Duly noted. Nobody makes motherboard chipsets like Intel, unfortunately. I tried a VIA one a couple years ago and was roasted by their 4-in-1 drivers. I'll give nVidia a try when I go AMD. Neither of them seem to have people blazing with confidence and extacy from a motherboard driver perspective, however. *sigh* (not a perfect world)

It's foolish to hold grudges, IMO, though people do it all the time. Whether it's nVidia vs ATI, AMD vs Intel, or nForce vs Via, the forums are full of people who had a bad experience with a particular product from a particular company -- sometimes years ago -- and have held it against them ever since. There's people who swear one company's products are the worst thing they've ever used and the other company's are fantastic, and there are other people who swear the opposite. And they're often both right, because they're talking about different products. They're overgeneralizing from their own limited experience. A bad experience is reason to be skeptical, sure, but not to dismiss the next thing out of hand. Particularly when it relates to drivers: virtually every hardware company has had to learn, the hard way, that their superiffic chip or board is worthless without an investment in software development. Sometimes they've had to learn more than once. But they do generally improve things, and drivers in particular are one area where the improvement can be dramatic (and, alas, regression more so). So to let a bad experience turn into a blind bias is simply irrational. Assume each generation starts from a clean slate (except perhaps for products from Creative, where the current generation never seems to end). Don't buy anything until there are reviews, then read the reviews without bias, and buy what seems best right now.