Not to be a Killjoy but I distinctively remember these words directly from the horses mouth........We at AMD Do Not do Soft Launches!!!!:rolleyes:
Printable View
Not to be a Killjoy but I distinctively remember these words directly from the horses mouth........We at AMD Do Not do Soft Launches!!!!:rolleyes:
they launched the chips and sold them already to OEMs,
but the OEMs won't launch them untill the BIOS is reliable.
if there is product available on the market within 6 weeks of the announcement, then, it's definitely not a soft launch in my book. a couple of months, yes, that's a problem.
dave
Hey Dave and others, I'll say thanks!! :up:Quote:
Originally Posted by dave_graham
I'm just looking forward to Phenom stuff. I can't wait to upgrade and put my S939 stuff in my wife's computer.
To bad Intel didn't take some pointers from AMD's 939 Platform :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap: :clap:
Checksums are valid because he probably DID run those times, but using a different CPU ;)Quote:
Originally Posted by KTE
I'll try to find some time to run a few more benches this weekend as I'm currently @ IDF (yeah, yeah, I know - just seeing what the competition has to offer :D ) which ends on Thurs.
i love specing on xs whenever something new comes out. you guys make my giggity giggity. lol.
BTW, forgot to mention - Colfax determined the problem with the setup I was running was an issue with a bug in the bios incorrectly recognizing the cores on the chips....problem solved :yepp:
s7: Any more details as to what they were being recognized as? Finally waiting for some fully working representative benches :D
Thanks doompc. I knew they were fake as I mentioned before asking why - but I was trying to establish how we the end guys can tell they're fake, because heck, that's just as good as anyone else's proof around anywhere online: 99% of it.
Which means for such a contest, we REALLY do need far better regulations and verifications (SPI).
And BTW justapost... look at what SOLDNER-MOFO64 said in that same thread: http://www.xtremesystems.org/forums/...&postcount=522
Wow, talk about diversity! 49s, 46s, 43s, 42s, 26s, 19s, 13s.. :confused: :shrug:
Something that may be interesting also:
Also: S7, those are (maybe?) good news! Would it be possible to say what exactly was going wrong?Quote:
Originally Posted by http://www.mail-archive.com/gmx-users@gromacs.org/msg09469.html
That, summed with the info that SM ("alleagedly", not quite sure how to write this word) has solkved these specific problem can bring us really closer to real benchies finally, correct?
For those who appreciate what we're doing with what we can, you're welcome. ;) To those of you bickering for no reason other than to start problems and flame the hardware, enough already. I think Dave's "gratis" comment pretty much sums it up. I actually took time out of my work day to get you guys results, and had to bust my ass and stay late to make up the time difference. :p Give us a break, we're doing what we can since we have access to the procs.
I don't know what to believe anymore. :cool:
I hope/think the majority of us are very grateful for the info you guys have given us. I also hope/think we realize how much of a sacrifice you guys are making. Unless you have the burning desire to do so, please don't sacrifice your homelife or work time to satisfy our selfish thurst for knowledge. As far as I'm concerned, the rest of us can just wait.Quote:
Originally Posted by tsuehpsyde
Anyway, THANKS for taking time out of your busy schedule to work on this system. :up:
++Quote:
Originally Posted by PhilDoc
Thanks for taking time to pique our nerdish hungers of knowledge :)
I guess the funny thing about it is we've all been bashing Super Pie as a comparitive bench but we love seeing an AMD chop come close. :D
Yup, with PhilDoc here entirely. Thank you :up: :clap:
Take your time... you know what I mean. :D
I think the credible benchmarks that we have seen indicate that a 1.9Ghz Barcelona scores around 43s in SPI. Which is fine because I'm sure AMD didn't set out with K10 to take back the SuperPi crown. I'm sure that with Phenom with DDR2 800 and tight latencies we'll be looking at sub 40s at 1.9Ghz.
Even if we see benchmarks with a PCIE graphics card, it doesn't really mean anything because the chips won't be overclocked and they'll be running with crippled memory. Even if K10 is fast at 2Ghz in games, it doesn't matter if they can't clock to where Intel is clocking right now.
Everyone needs to chill out. Thanks to dave and s7 for doing so much work to get us these benchmarks that people seem ungrateful for :).
SuperPi my ass......I just want a fully functional dual Barclona Server system thats not lacking on raw performance and memory bandwith and I have 0 intention of doing any overclocking what so ever!
Too bad AMD didn't take some pointers from their 939 platform. ;)Quote:
Originally Posted by Brother Esau
Edit
They are fake because I made them. I was just showing how easy it is to fake benchmarks with a little creativity and Photoshop :D I think I'll remove them, they causing more bad things than good.Quote:
Originally Posted by KTE
SP33DFR34K: Actually, thanks for posting them. I know you "concocted" them when I posted them :D but its good to have them so we know where the loopholes exist in establishing whats fake and original online. ;)
I'm going to die off now until tsuehpsyde, dave_graham or s7e9h3n post updates in their thread. :)
So anything new regarding the ASUS L1N64-SLI WS with Dual Barcelona
Here ya go: http://www.newegg.com/Product/Produc...82E16819105165Quote:
Originally Posted by Brother Esau
And another: http://www.ncix.com/products/index.p...anufacture=AMD
The first link =trueQuote:
Originally Posted by DeepThought86
The second link != true
http://www.ncix.com/products/index.p...anufacture=AMD
I don't know how to check stock at Newegg so I'll pressume they have 'em.Quote:
Taking Orders for this Product
This product is available but currently we do not have stock - order it today and we will ship when it's available.
Warehouse SKU City O.H. Qty. ETA Date On Order
Warehouse 1 57777G Vancouver 0 0
Warehouse 1 57777G Toronto 0 0
On the side: I'm company agnostic. However, if we're going to start throwing out terms like "it's available," I feel we should look at how available it really is. And shush, Dave, we already know that you have a tray or two. Lucky lil' **** :P
Back to Barcelona testing...........please? :)
Newegg is realtime so almost always if it says in stock then it is in stock.
not exactly true, Sparky. I've had SEVERAL orders that have shown as in stock and have bounced back and been voided. ;)
dave
no.Quote:
Originally Posted by kemo6600
dave
Newegg has 36 in stock atm
and i wouldn't advise running ANYTHING on a single plane board. :)Quote:
Originally Posted by stealth17
cheers,
dave
what stepping on egg stock? b1 or ba?
it's BA. AMD, again, is only shipping BA for sale, not B1.Quote:
Originally Posted by tictac
dave
Because of the higher NB speed only?Quote:
Originally Posted by dave_graham
no, because of the power management features, dual HT links, and requisite board refreshes that come along with it.Quote:
Originally Posted by justapost
if you haven't bought in already, it's more advisable to wait until the B2 announcement. If you have to have it now, please let me know and I can direct you.
regardless of purchase timeframe, the split plane boards will be a better platform moving forward.
dave
Dave, among those reasons for the re-spin of the MBs: which should be related to raw performance?
Also, are you too at IDF? That would be funny! :D
well, the respins are to provide support for the features I mentioned above.Quote:
Originally Posted by JohannesRS
nope, not at IDF. :)
cheers,
dave
http://farm1.static.flickr.com/58/18...953045fd58.jpg
this is me. :)
this is me after getting barcelona's to work on a 3992.
just needed to put SOMETHING here.
dave
Yes, but I would guess that only the dual HT links are directly related to raw performance, is it right?
Ok, the "extra-clock" in the IMC from the split-planes also... But anything else?
Btw: guess there is still no news with all the bioses problems? :(
Hehe ,Dave ,that's how a true happy geek looks like when his gear finally starts to work right :D
that smile made me smile too :)
lookin forward for results
Hahaha, just seen the picture now!
Is the long beard debugging stress related? :)
By the way: Is it fully working now so? What was wrong? Any pics for us to see that it's working now (we can wait for the benchs)?? :D
And, com'on, you must be joking: the same day it appears on newegg, you fix it? Confess, you all were only making us bite our nails! :D
nah, i keep the goatee as a habit.Quote:
Originally Posted by JohannesRS
well, i get consistent posts and i've verified that I'm not running into the P-states issues that S7 is. I've got some AMD tools that i'm using (sorry, can't post those as it's covered by my NDA) to verify platform stability, so...we'll be wrapping that up later tonite, i think.
still using B1's and have another BIOS load to go, but, progress is little by little. :)
what newegg does is their business. :)
dave
49yeras old?:shrug: :DQuote:
Originally Posted by dave_graham
hardly....try 29.
dave
good news now i(3992 owner) can sit back relax and wait till i get 2GHZ barcys fpr 200€. :up:Quote:
Originally Posted by dave_graham
you mentioned a new bios version, does this one allow fsb and multiplyer settings?
not on the Barcs. there's a checksum diff. between the bios that I have and the website. that's my project for tonite. :)
dave
hey i was close wasn't i?:hrhr:Quote:
Originally Posted by dave_graham
ah ok, so it is the 2.01 version you already emailed me?Quote:
Originally Posted by dave_graham
I'm working on a comparison between the native and xen-enterprise virtualized oses here and did alot of benchmarking with w2k3_64 since the weekend.
Do plan to benchmark the barcys undel linux?
If so i can run those benchies here too for comparison.
Hmm just found that one, looks interesting and more server related.
http://www.mercury.com/us/products/p...er/loadrunner/
I was looking in this tread and finding it actually funny that people are all screaming for some 3d or super-pi benchmarks it’s not like the Barcelona is made for those kind of thinks it’s not desktop platform, so I’m not quite sure why people are interested in that. It’s not like you gone try and get some killer 3dmark numbers whit a xeon processor..;)
so there for I would love to see some more 3ds max, SPECjbb2005 or MySQL :rolleyes:
also thnx dave and s7e9h3n for all the info
BUT......Here in XS we are Overclockers......We are not "Servers".... :D :p:Quote:
Originally Posted by lookmomnobrains
I think it is the wrong forum posting things like you wrote.....We NEED OUR bencies.... ;)
WE here don't care about how it does in yourSQL, etc.....:D We need 3Ds and things like that you know.... ;) :)
i'll be receiving an Asus KFSN4-DRE tomorrow.
link here
cheers,
dave
Anyway to move the thread to 2cpu then? ;)Quote:
BUT......Here in XS we are Overclockers......We are not "Servers"....
I think it is the wrong forum posting things like you wrote.....We NEED OUR bencies....
WE here don't care about how it does in yourSQL, etc..... We need 3Ds and things like that you know....
Quote:
Originally Posted by hipro5
But it’s not like you can expect some great 3dmarks numbers from a server platform if that was the case roll in a big LN2 tank and start benching some on a cpu cluster..
If this was the desktop version aka phenom I couldn’t agree more, but it seems kind of useless to me to see some 3dmark benches on a server platform sins its clearly not gone produce some killer numbers.. I much rather would like to see how it does in some server benchmarks to see how it hold its own compared to the new xeon line
lookmomnobrains: Most people at XS are very interested in hardware, overclocking, bandwidth, RAM timing, CPU & GPU benchmarks, etc... and they keep asking for more as long as someone here posts info about something new, like a new CPU type.
It's not like their logical thinking (like the fact that this is a server CPU) stops them from asking for more, because there's a chance that even a server board can tell us something about desktop K10 that we don't know. A WS board would be better of course. Besides, it's not like any of us had enough of K10 benches yet. For the most of us, this is the best we can get for now. The K10 users in this thread may have other/more reasons though. The possible results in this thread may also help some people decide wether they should get a C2 now or a X4 later, even if they don't get all the answers here.
I personally don't care much about most benches, I'm just interested in the new hardware.
I think you'll find the benches you asked for somewhere else before they shows up here.
Dave, S7 and all the other K10 users: Keep up the good work!:toast:
Dave, can you tell us what did they modified in that new bios?
Thanks! ;)
Yea, we all care about something... lets call it squitz there. ;)
Dave: DRE is supposed to be awesome, AMD uses it for official benchmarking. Lets hope it plays out well in your hands. :D
when i get there, i'll try to determine that.Quote:
Originally Posted by JohannesRS
dave
Foud that article on a german site.
http://www.heise.de/newsticker/meldung/96237
They got their hands on hapertown and barcelona cpu's and ran spec on both machines.
The rate benchmarks are faster on an Barcelona at 2,5GHZ than on an xeon running at 3,0GHZ.
Just to make it easier for people 2360SE (2.5ghz)Quote:
Originally Posted by justapost
SPECint_rate
SPECfp_rate
Whoa. So a 3.2GHz/130W TDP Penryn scores 89.8 in SPECfp_rate 2006 according to Intel at IDF and a 2.5GHz/120W TDP K10 scores 80.6 and the 3GHz Penryn is beat on both scores by a 2.5GHz K10 :eek:
Ermm.. what was that about scaling again?
EDIT
Yeah. Intel Harpertown 3.2GHz 1600FSB results by Intel at IDF for comparison: http://www.betanews.com/article/Inte...ron/1190217688
Awsome. Finally some nice results :)
Quote:
Originally Posted by KTE
Its the same board Anand used for their Barcelona review as well, so you'd think it would run adequately. I don't know how good of a board it is and I don't know from the manual if you can overclock it, but it looks like you can upgrade the video card. So, for those hungry for 3dmark scores, it should give us some indication. If this is the case, then considering how much Dave hates Asus, maybe we should thank him for throwing this board into the mix. Its possible he did it because of our continued demand for gaming scores.
hrm.....hasnt the barcelona already proved itself in the RATE scores already? but it kinda lacks in the non RATE scores...I thought weve seen all this already?
It seems like we've been over this many times, even K8 was very competitive in spec_rate and spec_fp. Its nice to focus on the two benchmarks that AMD wins, but they lose pretty much everything else until they can get clock speeds of 3G+.Quote:
Originally Posted by GenTarkin
I think the only usefull benchmarks in that article are the barcelona 2359 and the xeon E5472.
Both where build for 64bit and without the smart heap library, so there are no big compiler difference in the way here.
I even guess it was exactly the DRE board that "was not manufactured by asus", if I remember it well... ;)
Anyway, didn't got what KTE meant on his post.
Can you guys test it with Intel
Code Demo: Destroy the Castle
Download Link:
http://softwarecommunity.intel.com/a...s/eng/1363.htm
Video Preview
http://youtube.com/watch?v=QdCDq6snQoU
Specs:
AMD x2 3600 @ 2.72Ghz
Biostar 550
Consair and Kingston = 512x3 DDR2 775Mhz
EVGA 7900GS KO @ 570Mhz/1.6Ghz
that's pretty slick, nice linkQuote:
Originally Posted by tictac
http://img229.imageshack.us/img229/1...armultifr0.gifQuote:
Originally Posted by dave_graham
I'm interested in getting a 2S SP board now with a single 1.7GHz, which I'll toss for two 2.5GHz as soon as those parts are stable - do you think a S2932-E purchased now would work well with the 2.5Ghz chips? - or will even newer boards come out then?Quote:
Originally Posted by dave_graham
Thanks,
Morphic
No news from the front? ;) Still problems with the bioses, or the DRE MB have failled completelly?
The system I have access to is still working, the fun part is trying to explain to Supermicro what the issues are, since they seem to think the motherboard runs the processors fine. *groans*
That's why I love supermicro... :P
Anyway: tsue, can you tell us what are the issues you are having specifically with this MB? Still the same (evil) problems with memories and processors clocks?
Any advances fellas?
Have to fulfill my comparison quest in the meantime :p: lol!
Harpertown-Stoakley 3GHz (2CPU/TR):Quote:
Originally Posted by KTE
http://img166.imageshack.us/img166/9776/htwnjbboa4.gifhttp://img241.imageshack.us/img241/9111/specjbbfn9.gif
Harpertown-Stoakley 3GHz (2CPU):Quote:
Originally Posted by s7e9h3n
http://img215.imageshack.us/img215/1...wnssmempm6.gif
Harpertown-Stoakley 3GHz (TR):Quote:
http://img166.imageshack.us/img166/4...cachebwrk1.gifhttp://img237.imageshack.us/img237/7...hebw1gbmc6.gif
Nobody found or got any new bench for the Barcellonas, guys? :(
Anandatech posted a few in an comparison with hapertown last week.Quote:
Originally Posted by JohannesRS
http://www.anandtech.com/IT/showdoc.aspx?i=3099&p=1
Thanks! and that seems that Barcellona can fight direct against Hapertown clock per clock, isn't it?
Price, Watts for servers, and also OCs for desktops. Those are the only stuff that we are still awaiting to know how the game will go. ;)
Thanks, nice article. Looks like they use a Tyan mb this time. The results lead me to the same conclusion as before. It all depends on how much AMD can ramp up Barcelona. We also don't know which stepping of Barcelona they used, which leaves things a little up in the air. Either way, if they can get up to 3G reasonably soon, things could get very interesting.Quote:
Originally Posted by justapost
Hope they will do a clock per clock comparison soon.Quote:
Update: For those that are looking for more details and wondering why certain other chips aren't included, at the time testing was conducted we did not have any of the faster 2.5GHz Barcelona chips (or the slower Harpertowns). That situation has been remedied in terms of AMD's CPUs, and we will have some update articles looking at how the faster Barcelona compares with other processors. Stay tuned....
Quote:
Originally Posted by justapost
Sounds like it. Hopefully, within the next week or two.
AMD has it down, less power consumption, less utilization to complete tasks.
harpertown will need one heck of a psu over 400 watts on a few of those tests!
Those 2 bench comparisons I posted compared to what we've seen here were both new from TechReport and 2CPU, as well as the AnandTech that justapost later posted. They're all pretty much the same in conclusion, although the absolute performance figure in all of them are different; clock-per-clock Barcelona server releasing Q4 07 looks better than Penryn server (1600FSB) releasing Q1 08 more often than not. Altogether, it even wins a few whilst with a 500MHz deficit and loses many due to raw MHz power.Quote:
Originally Posted by JohannesRS
I remember Henri saying this last year with regards to competition... "it's a race, of course it's a race." ;)
But how long will AMD take to get the Barcelona speeds to the 2.5-3.0 GHz?
And about Phenom´s launch date and speeds? Is there more information?
K10 supply has improved greatly:
http://www.amdzone.com/modules.php?o...ticle&sid=8453
:)Quote:
AMD Opteron Quad Core 2344 HE CPU OS2344PAL4BGE, 1.7GHz, Socket F (1207), 1000MHz HT, 2MB L2 Cache $236.00
Wonder if this means that Phenom will be a little early, as barcelona supplies look good?:DQuote:
Originally Posted by informal
Sorry I haven't been able to get back to this thread until now. At the moment, it seems that this is about as good as it's going to get with the current hw/sw support available. The bios on my Tyan s3993 mobo still has no OC options, no voltage options (including dpp), no HT3 options (yes, this board supposedly supports it), and no memory timing options. Add to that the fact that neither Broadcom, Tyan, nor Serverworks seems to have windows chipset drivers available for the Broadcom/Serverworks Ht2100 Chipset and the memory is stuck in single channel mode and you end up with this totally, utterly, stock-as-can-be dual Barcelona rig that I have here. Here's Cinebench and Everest's cpu-tests with the quads running @ the correct speeds:
http://img509.imageshack.us/img509/1...ench64akl3.jpg
http://img210.imageshack.us/img210/2/evcpuqueenqs9.jpg
http://img210.imageshack.us/img210/1...puphotorp0.jpg
http://img210.imageshack.us/img210/145/evcpuzlibtm0.jpg
http://img509.imageshack.us/img509/2522/evcpuaesoa7.jpg
http://img210.imageshack.us/img210/1...pujuliaja0.jpg
http://img210.imageshack.us/img210/7...mandrelxc4.jpg
http://img210.imageshack.us/img210/3937/evfpusinog5.jpg
I'll see what I can figure out about the memory problem and run some in-depth benches later.......
FP power looks weak... teh same B1 eh?:P
thanks Stephen for your patience with these CPUs!
I noticed that as well, but also note that on all the FP tests, Everest gave that error message so something may be wrong there....Quote:
Originally Posted by alayashu
Yep Steven,if Everest deals with unsupported HW,it gives wrong results in some tests(and warns with that message)
If you do a Sandra memory bandwidth test what are those scores looking like on single channel?, and thanks for all the hard work.
Very nice, Thanks, S7. Looks like you finally got the bios work around or was there a differenet problem?
Better, but still crap :p:
http://img106.imageshack.us/img106/7587/sandrabwfo6.jpg
It was a bug in the original bios that somehow didn't get caught. Basically pinned core0 @ max power-save mode. A bios revision corrected that problem ;)Quote:
Originally Posted by PhilDoc
Thanks S7 for the quick response. Sandra doesn't look too bad for single chanel. Hard to tell what the problems is there. I'm sure you've tried different ram modules. Could be the mb itself.
Wow yes thanks for the quick response, looks very good for single channel.
great. So, now we have the cores at ptoper speeds, so.
But unfortunatelly memory still at single channel, and I would expect that to give a hit on opterons performance. Big one.
S7, do you think it worths trying a linux installation at this point, or it more wise to wait for a fully working bios? ;)
And, of course, thanks a lot for sharing those results with us! :D
Sigel channel and DDR-667 @ 5-5-7-15 ain´t helping either.
Wonder how much 3-3-3-8 @ ddr800 dual would give in therms of sandra memory bandwith? As always thanks for sharing S7:up:
Those are the expectations for Phenoms in terms of memory. Don't know the timings of ECC memories, but 667 MHz are the specificated ones.
The main problem for server parts, for me, now looks like only the single channel problem still. Is it right?
I really hope the corporate types aren't having the problems (ie...dual channel memory problems, cores not recognized properly) discussed in this thread. Even though it's not AMD's fault (nothing wrong with the chips that we know of), the motherboard manufacturers can hurt AMD just as bad with flaky bioses.
For reference from a dual-dual Opteron rig...Octal AMD should be MUCH, MUCH faster:
http://img440.imageshack.us/img440/5...dwthfr53ce.jpg
The problem seems to be with the chipset not allowing ccNuma to function properly along all 8 cores as the only two IMC's that get loaded during bw benchmarks are core0 of each physical proc.
I wonder if its the chipset itself or a faulty mb. Tried to look the reviews that ran sandra, but they were running supermicro boards. I think the tech report got 11,500ish so they must have gotten dual channel to work. Probably no way to try a different mb.