Dude, we wont see Deneb until 2009. I'm building a 6500BE rig in the meantime. This old opteron 185 rig needs to be retired.Quote:
Originally Posted by Mats
Printable View
Dude, we wont see Deneb until 2009. I'm building a 6500BE rig in the meantime. This old opteron 185 rig needs to be retired.Quote:
Originally Posted by Mats
I know, but that's not really far away.Quote:
Originally Posted by Mechromancer
You know AM2+ has been out for a while now I don't know why some people are bashing this chip especially when it gives people who have stuck with their mobos a fairly decent upgrade. I don't think AMD is bragging about this chip or making a big fuss so I don't see why theres a big deal over it?
People who bash AMD without even reading the fine print will always...Quote:
Originally Posted by Glow9
Bash AMD without reading the fine print.
6500+? That does not exist :ROTF:Quote:
Originally Posted by Mechromancer
So when is AMD gonna release a dual-c0re chip that does 3.5 on air? :shrug:
its called the 6400 BE tard
Have you considered that the 6500 is the model name, like 9500 or 8750. "Nice PR stunt to cheat customers". What a sad little Intel shill you areQuote:
Originally Posted by Shintai
i dont understand, if this chip overclocks to around 3.3 on air (or more im sure, with some tweaking), it will beat the snot out of any athlon 64 X2 there is.
end result: good for us overclockers, bad for the consumers (a 2.8-2.9ghz k8 performance is not too shabby though)
Shintai's point is that AMD enumerates their processors with 4 digit PR numbers (they have long since lost their 'Intel equivalent GHz performance' meaning even when pitted against the P4s).Quote:
Originally Posted by LightSpeed
People have become accustomed to the higher the PR number the higher the performance. The 6500 designation would imply to the less knowledgeable that it is AMD's highest performing dual core... when, in fact, it is not.
Same thing with naming the 9600, 9850, 9950 to an extent, playing against Intel's numbering scheme -- less knowledgeable people may think that the 9600 is almost as fast as a QX9650, and that the 9850 and 9950 are faster.
I completely agree, it is misleading to the general consumers, but for overclockers it is their highest performing dual core. hence i said its good for overclockers, and bad for the consumersQuote:
Originally Posted by JumpingJack
Black Editions are only sold in channel (NO OEM), and particularly for enthusiasts and overclockers who obviously do that.Quote:
Originally Posted by LightSpeed
The clock rate should be 100-200Mhz more for the price, but I'm not complaining. If Deneb suddenly sucks I might get this and run with it.
finally im sure we'l be seeing +1000mhz overclocks on average with these chips
some acc loving coming up...
i hope these are cheap in the UK
^^the thing is there is no cheap board with S750 and people who want to upgrade usually have older mobos. still it should oc pretty good since the open multiplier.
http://www.newegg.com/Product/Produc...82E16813138128Quote:
Originally Posted by dexman
Quite ironic, isn't it?Quote:
Originally Posted by dexman
Amd fans braged how cheap there boards where compared to intels one, now you pay for a 790FX-GX/SB750 the same as for a X48/ICH9R. :p:
This Biostar mobo looks interesting :)
In your bizzaro justification world, perhaps.Quote:
Originally Posted by Hornet331
790FX-GX+SB750 price = x48+ICH9R? have you got a proof.Quote:
Originally Posted by Hornet331
The Dually's name should have been Phenom X2 7650. Other than that it's fine, but I guess AMD don't want to get rid of their beloved Athlon brand.
If you take away the $100 higher price tag, the cheapest x48 on newegg is the same price as the cheapest 750SB amd board. :shrug:
+1Quote:
Originally Posted by Power5
X48 board pricing is ridiculous.
no its called europa. :rofl:Quote:
Originally Posted by Macadamia
790FX/GX + SB750 80-225€
http://geizhals.at/eu/?cat=mbamd2p;a...=on;v=e&sort=p
X48 - ICH9R DDR2 150-225€
http://geizhals.at/eu/?cat=mbp4_775_...317_X48&sort=p
X48 - ICH9R DDR3 125-300€
http://geizhals.at/eu/?cat=mbp4_775_...317_X48&sort=p
times of super cheap Amd boards are gone. :yepp:
Whatever you're smoking, I'd like it. Delusional...
Remove that ASRock even your mom wouldn't buy (probably won't OC too), and X48 is baseline 150 Euro.
The only 2 boards exceeding 150 Euro are niche 790FX designs with 3x PCIe. 790GXes are all cheaper than X48s.
But don't worry, it's already concluded that Spintel enthusiasts will even be willingly raped for the next generation of motherboards- X58- that ironically has even less crucial silicon than ever, and yet going to be even more expensive, especially after the SLI tax.
Haha, it's always the same. Some guys think NewEgg and the US are the only place where mobos, CPUs and such are sold. Bizzaro world... :ROTF:Quote:
Originally Posted by Hornet331
Dude, while the majority of 790FX are cheaper, that doesn't change the fact that AMD boards are no longer much cheaper than Intel's like they were before. As Hornet says, the days of cheap high-end AMD motherboards are gone.Quote:
Originally Posted by Macadamia
In greece they are way cheaper...
790gx = 120-130 euro( asus for example)
790fx = 170 (asus again)
Screw x48. even X38 is still fine and has no problem competing against an AMD rig. X38 boards are cheap nowadays.
AMD is simply not cheaper then Intel price/performance wise.
I see no reason to buy this 6500+. Wait for Deneb or Nehalem. Don't waste your money on older hardware like C2D or Phenom. Both have been around for a long time now.
I live in Malaysia. The difference is even bigger (1.5-2X pricing). Wanna compare? :rolleyes:Quote:
Originally Posted by STaRGaZeR
Gone my arse.
Plus AMD boards were cheap, but they ALWAYS had expensive models too. DFI LanParty etc, all taking up the highend NF3/4 market. Now you have cheap and expensive models (more brands going upclass) but the cheap models still exist. There's no cheap X38/X48/P45 cop-out. Hornet's fallacy hardly applies.
im not the one thats complaining about price, tbh i never cared how much I pay for my hardware. I buy that what im looking for, else I wouldn't have bought a QX9650 + X38 ddr3 board with ddr3 modules for 550$ back in 2007. :yepp:Quote:
Originally Posted by Macadamia
Also i think the new nehalem rig im going to build will cost me approx 2k€ (for cpu, mobo and ram alone).
Plus if you remove the ASRock you could also remove the Biostar from the AMD board selection, which leave the cheapest dual/tri slot 16x pcie board at 110€. (Gigabyte, since ASrock is a nogo for you and biostar for me).
If you look back how much a top of the line AM2 board back in 2006 was (~160€ ASUS M2N32-SLI Deluxe) and now a top of the line AM2+ board (250€ Foxcon Destroyer) its quite a difference. :yepp:
But that also goes for not so highend boards, back in 2006 you could find good AM2 boards for 70€ now you have to pay 40€ more.
what, no cheap p45 boards?Quote:
Originally Posted by Macadamia
Where do you live... if we start factoring in P45 boards the SB750 boards look even more expensive. -> MSI P45 Neo3-FR for sub 90€ and the plain vanilla MSI P45 Neo-F for sub 80€....
qfe :up:Quote:
Originally Posted by ownage
Many of us were expecting such advantages in performance, the 6500+ BE has the same IPC like Phenom X4/X3.Quote:
Originally Posted by FlanK3r
The 3.3GHz wprime stable OC is very promising. If the 6500+ BE is priced properly I will buy one for sure. It'd be my first K10. :)
@To the guys from expreview:
Thank you very much for the benchmarks, but can you please repeat the same benchmarks on the same system with an Athlon64 6400+. The product number "6500+" already have misled some people to think it will be faster than the 6400+. It'll fool far more people when it hit the shelves. So a hard proof would be nice to inform people who consider to buy it.
does k8 have 2meg level3 cache?
I know k8 has 4x512 same as this 2x512 level2 '1x512 per core', but maybe like with core2 arch the cache is the bonus with this new cpu?
It's true, even every single PC component recommendation on this site is automatically linked to Newegg, regardless if the person asking is located in Botswana or Norway.Quote:
Originally Posted by STaRGaZeR
And I also completely agree to the second point.
If you mean K10, then:Quote:
Originally Posted by CCC
1) Yes, the K10 CPUs up to date have 2MB of L3 cache.
2) Yes, K10 has 4x512kB of L2, 512kB of L2 per core.
Larger caches are always welcomed, but to them applies the law of diminishing benefit. How a CPU will benefit from a given cache capacity depends of the CPU architecture, the CPU (and cache) clock, the software(OS and apps), the cache size, etc.Quote:
but maybe like with core2 arch the cache is the bonus with this new cpu?
I was just thinking. They should have released a dualcore phenom based cpu series from the beginning instead of only releasing sucky quadcores.
They could have compete much better with dualcore phenome against C2D then old tech K8 based cpu's.
I can't believe it is hard to make dualcore phenoms. Should be easy for them.
So you think duals based on the 'sucky' quads don't suck? :confused:Quote:
Originally Posted by ownage
Seriously, even if Kuma was launched concurrently with Agena, it wouldn't have made a lot of difference to AMD in terms of competitiveness in the dual core market, especially at the modest 2.3GHz clockspeed. We already had 3GHz+ Wolfdales by then...
Weren't Wolfdales only released in early 2008 while Phenoms were released in 2007 November?Quote:
Originally Posted by Epsilon84
Sure but it's not like Conroe is chopped liver.Quote:
Originally Posted by Karolis
Chopped liver? :eh: :sick: All I was saying is that when Phenoms were released, Wolfdales weren't on shelves yet. :)Quote:
Originally Posted by Donnie27
This was AMD's fatal miscalculation. They never even bothered to design a dual-core next-gen architecture, they went straight to quads when the mainstream market didn't need 4 cores (still doesn't). So Brisbane has hung around for 2 years too long, and counting. Yet another disadvantage in the desktop space for the monolithic quad design. K10 is a server design through and through and AMD doesn't have the $$ to develop a separate desktop-optimized design which concentrates on high-clocked dual cores.Quote:
Originally Posted by ownage
Regarding the prices on mobo for X48 vs SB750, it made me lol tbfh.
Also how can someone compare ASRock and Biostar? Biostar is a whole lot better than ASRock. And the most expensive SB750 board is from Foxconn. And those are way overpriced and not worth it, probably because they had the 'release right' or however you name it. Also the RD790/RS780D are high-end chipsets and available for anyone, from mainstream (maybe value?) to highend for not too much. You pay a little bit more for SB750, but that's marginal but worth it a lot.
I do agree AMD should have released K10 X2's earlier, a lot earlier. The reason why they release them now... I think they're moving things strategy wise. Deneb X4's will be released with higher clocks, X3's might as well where X2's (either Agena or Deneb) will be replacing X3's.
Like if it matters where you live. If your local providers are charging you load of money for them that means nothing for the rest of the world. Blame your country/shops/whatever, not the manufacturer as those prices are not even close to what the chipset itself cost.Quote:
Originally Posted by Macadamia
In all that paragraph of yours change AMD with Intel and voilá! They're the same.Quote:
Originally Posted by Macadamia
No cheap P45? 100€ or less is not cheap for a kickass P45 mobo where you live? :shrug:
Point is, stop saying AMD mobos are cheaper because there are very good mobos in both sides for a very reduced price. Yes both means Intel and AMD. No I'm not smoking anything. It's just the real thing, believe it or not.
damn guess I have to find someone else to share with me:rotf:Quote:
Originally Posted by STaRGaZeR
honestly guys, calm down this thread is about the cpu's performance, if anything flame amd for tricking the average customer into thinking it offers better performance than the x2 6400, don't bother messing with people over the cost of motherboards
If you want to OC Phenom quads beyond 3GHz you pretty much have to get a board with very solid PWMs, and those board are priced very similarly to X38/48 here at least. So I'd say the motherboard cost is pretty much even across the two platforms. Glad to see the Phenom based chips overclocking pretty good, 3,2- 3,3 GHz pretty regularly on the quads, I'm quite optimistic about Deneb. I'd love to play with a 6500+ in the meantime.
In the UK the x48 is roughly 60%+ more expensive.
For example,
http://www.scan.co.uk/Product.aspx?WebProductId=787778 for 790fx
http://www.scan.co.uk/Product.aspx?WebProductId=831175 for x48
As of this point, buying an AMD cpu/motherboard does not make sense, it simply doesnt offer anything better than intel, simply nothing. No reason to buy AMD till they get their act together and come out with a good performing modern cpu. Old tech is not cutting it anymore, and K10 is simply an update of K8, nothing new really, K8 was 2002/2003 technology.
This kind of argument just doesn't work. They have price competitive products right now, their boards are great for low-power HTPCs/file servers. If K10 is "old" then what do you call the Pentium 3-based Core 2? It's all old, AMD just didn't out-engineer Intel. You and all the others asking for AMD to just come out with a new arch are ridiculous. They don't have the money to rush development into new projects, and besides, you don't just engineer a new arch overnight. Their chips sell in OEM systems, and they still have a place in the market.Quote:
Originally Posted by GAR
Of course they weren't. Yet, it's not like Intel desparately needed something better than Conroe even if there were Kumas on the market. Conroes did the job well enough.Quote:
Originally Posted by Karolis
QFT!Quote:
Originally Posted by ryboto
Yes, like if Core 2 equals P3, wouldn't that make K10 = K5? There are some great sub $80 Sc-775 P35, P38, P43 and P45. Others less than $120.
http://www.hardwarezone.com/articles...?cid=6&id=2288
Actual Cost was $77 shipped since I bought it during the rebate period.
http://www.zipzoomfly.com/jsp/Produc...tCode=10008667
I got it at Newegg for $87.99 shipped.
Two of my Bud's like this one!
http://www.zipzoomfly.com/jsp/Produc...tCode=10008670
So enough with the Cheap Boards excuses, they'll have to find something else.
Edits [I] so there's no doubt this is to show I agree, NOT disagree!
I got my dates mixed up, you are indeed correct, Phenom was released a couple of months earlier than Wolfdale.Quote:
Originally Posted by Karolis
Still, my point stands that at only 2.3GHz it wouldn't have made a lot of difference for AMD considering they had the 6400+ which would have been faster.
Quote:
Originally Posted by Macadamia
Macadamia... have you used an ASRock board ever... My Venice 3000+ used to overclock to 2.5 Ghz on air with a coupla sticks of Corsair 512 MB Value Select...
oh the board was a 939 DUal Sata with an upgrade path to AM2 and wats more it had AGP plus PCIe and both worked flawlessly :) so before u start busting a mobo, think..
oh and when this whole price/performance comes into the pic.. theres one cool video i Just HAVE to post :D ehehee..
http://www.youtube.com/watch?v=nq_XG411Lik
I see the difference, the planet in the AMD demo is FLAT! Look again at 0:35.Quote:
Originally Posted by saveus222
Meaning they're at least 500 years behind Intel . . .:shrug::D
And the song is just a Streets ripoff.
Seriously though, I get the point. I personally prefer AMD even if they're not the fastest, but I just can't do it when they're such power hogs AND slower.
I said wait for Deneb if you already haven't got a K10, and the reason is that K10 isn't that good.
AMD can and will do better, even if they won't become the best with Deneb.
I mean, if there's any truth in the only ES preview I've seen then it's worth it just for the temp drop.
I'd never have said the same just before the first 65 nm K8 showed up because the 90 nm's were really good.
Quote:
Originally Posted by Mats
ur probably the only peson who summed it up perfectly man :) respekt..
oh.. and isnt the earth flat ? ooops.. umm... then ? err.. hmm.. seriously tho.. u actually noticed.. :D lols..
Did you even read my post? this looks like a response to the others who were arguing about motherboards, read mine again, and this response doesn't really fit.Quote:
Originally Posted by Donnie27
Quote:
Originally Posted by ryboto
It was an agreement post LOL! Like yepp, can't make that excuse anymore! Now did I say you're wrong or etc.....? NO! I meant to add to what you're saying.:up: Maybe I'll go back make that BOLDLY CLEAR!Quote:
Originally Posted by ryboto
I don't care if a CPU is based on a beer can, as long as it performs good. But that's just my opinion.
K8 was based on K7, and C2 was based on CD/PM/P3, and that's not a bad thing! Both were successful!
Making a new CPU from scratch is extremely expensive, takes a lot more time and in most cases it's just plain stupid if you're asking me.
Edit: Maybe we'll have benchmarks sooner than we thought.
Going from 2.6 GHz, 140/125 W to 2.7 GHz, 75 W is just sweet!
I've helped folks pick AMD based system for bang for the buck rigs and even a couple of mid range ones. Just because Intel holds several advantages doesn't mean AMD sucks. IMHO, I think it is a great time to be a Geek! Most computers are overkill for the average user.Quote:
Originally Posted by Mats
Here's an example. I built my Daughter a Computer. If I wouldn't have already had an Intel Motherboard, it would have been AMD based. Why? It was cheap enough, more than fast enough, not as hot and power hungry since it wasn't going to be overclocked. In fact, the E2180 I installed ($64 shipped) isn't overclocked either.
Most of the disagreements or arguing going on here is nit-picking. It is easily not worth personal attacks, hurt feelings and flames in general.
well, alright then, that's good we agree. All this squabbling is silly. I agree with all the points in your other post about it being a great time to be a geek, or just anyone in general who needs a computer. Even some of the cheapest machines can do just about any non-gaming task.Quote:
Originally Posted by Donnie27
You got it!Quote:
Originally Posted by ryboto
Donnie: Totally agree. Most buyers doesn't even understand what their needs are, so they end up buying overspeced computers that cost way too much for them.
And the reason why? "Well since I'm finally buying a new computer, let's pay some extra bucks just to get something really good"
That "extra" they get is a HD 4850 (probably much worse) that will never run a 3D game.
Yup! I got that Intel motherboard for some labor. I did a motherboard upgrade and took the old one for pay or services. I was shopping for an AMD based sytem. If I hadn't gotten that mobo, I would have bought it.Quote:
Originally Posted by Mats
Who are you and what have you done to donnie27? Could you pass the potion to Shintai as well?Quote:
Originally Posted by Donnie27
http://i129.photobucket.com/albums/p...larm/PEACE.jpg
:D :D :D :D
I like this new and improved Donnie. At first, I wasn't even sure it was the same person.Quote:
Originally Posted by justthefax
Exactly.Quote:
Originally Posted by Donnie27
Anyone have an idea what happened to it168.com review of the Kuma 6500? It was posted at English hardspell website and pulled later on.Also it was pulled off from the original website too.
:rofl::ROTF::clap:Quote:
Originally Posted by justthefax
someobody quick touch wood.. :D ey is this some weird intel undercover agent.. lol.. kiddin :)
My point was AMD doesn't need to use misleading names for their processors. In this case I'm Pro Consumer just as I've always been.:up:Quote:
Originally Posted by saveus222
Yes and no actually, I understand your point but you've to be more carefull with what you say.Quote:
Originally Posted by GAR
Buying an AMD based PC right now is a bit of cannon fodder since you jump right into a hole. K10 is finally performing better and better due to slightly improving production, but mainly due to SB750. However, Deneb ES previews thus far shown a very nice improvement and with that in mind, it's not worth it a lot to buy another Agena unless you really cant wait. Athlon X2 6500 is nice to play around with of course, but just talking about Phenom X4, it's better to wait a little more.
However, regarding 'old tech', no. In the end most cores or any other product can be re-directed to older technology. Besides that, K10 has a huge improvement technology wise over K8 and Yorkfield. And that's also why K10 is worth buying, you've a lot of things to play with. Of course Intel is about to release their i7 Core soon which might fade this argument a bit although it depends a little bit on how tweakable they actually will be since there are quite some mixed stories going around. And well, I could ask any Yorkfield owned the very same question, why invest in older technology? And yeah, that's a pretty stupid question I guess:p:
Dunno what happened, world turned upside down, but I like it:up: Now only keep it up:clap:Quote:
Originally Posted by User9498
yea.. kuma overclocks to 3.2 Ghz on air.. so i think its a pretty decent buy
Well, depends of it is price. It needs to be at least 15% cheaper than a E7200 in order to offer a similar bang for the buck and when it comes to OC it needs to be further cheaper because the E7200 does 4GHz.Quote:
Originally Posted by saveus222
I am gonna pick on ya' a little bit here.... I do agree, K10 was an improvement over K8 ... i have argued in other venues that AMD did a good job on K10 when you look at the relative differences and improvements at the core level and defended their designers vehemently while others ridiculed the K10 microarch .. but I am not so sure it is a 'huge improvement' over Yorksfield, given that Yorksfield clocks higher, has higher IPC, and consumes less power I don't really think an argument that K10 is superior technologically is applicable.... could you explain in more detail the rationale behind your statement...Quote:
Originally Posted by Rammsteiner
Jack
But gOJDO is right on one part. 6500+ will have to compete with some stiff competition from other AMD processors. I don't think AMD used this name for without a reason. Why would they name it 6500+ and list it for a lower price than a 6400+? I hope I'm wrong here. In the OEM these will mean one thing that might be fine for AMD. But in this market that might be the case. Think about it? If folks are complaining about Nehalem overclocks, 3.2GHz will not be much when even my old Wolfdale will do 4GHz on air:up:Quote:
Originally Posted by saveus222
I wanted to reply to the same post also, but I let it go because it was OT and because Rammsteiner is an AMD lover. He wants Phenom to be better than Yorkfield and Kentsfield and he believes in that even he has no proof to support his argument.Quote:
Originally Posted by JumpingJack
Phenom + BS = "smoother" (copyright by FUGGER):rofl:
Kuma's NEVER going to OEM- at least not competitively, or EE/BE only. It's a channel product targeted at upgrades or new overclocking systems.Quote:
Originally Posted by Donnie27
Regor, the new 45nm native dual design coming out in 09, will be replacing Brisbane.
And really, you think we so desperately need how high your Wolfdale can clock, seeing you didn't even read the thread? :ROTF:
How much is architecture? How much is process?Quote:
Originally Posted by JumpingJack
I think a lot of people forgot that high-k is actually quite amazing in terms of process. AMD's not going to get it till 32nm, but I don't think it's at any aspect fair for those guys to get slugged that hard because they're less radical on processes.
lets get one thing straight here.. amd processors cannot be compared this generation to intel processors.. suggested price..109 USDQuote:
Originally Posted by gOJDO
http://www.shopblt.com/cgi-bin/shop/...html&order_id=
Godjo was comparing it to an e7200 not another AMD processors, but I see your point, though, im sure once this is released, the price of the 6400+ will automatically fall down ??Quote:
Originally Posted by Donnie27
(approx price: 109 USD)
http://www.shopblt.com/cgi-bin/shop/...html&order_id=
Thanks to the ODMC(s) and IMC, the Athlon64 and Phenom have much better Macro-architecture then Core2, but the micro-architecture of Core2 (Conroe) core is superior to K10 and K8 in any regard. Thanks to the much more advanced Core2 micro-architecture, although with crippled FSB and no ODMC, Core2 is still faster at same clock than both K10 and K8.Quote:
Originally Posted by Macadamia
Well, both IBM/AMD SOI3 and Intel bulk 65nm processes have advantages and disadvantages. The SOI3 has very low leakage and at idle the CPUs produced with this process are wasting less energy then CPUs made with Intel's bulk 65nm. But CPUs produced with the Intel's 65nm bulk process are wasting less energy when on load.Quote:
How much is process?
Intel's 65nm is faster than IBM/AMD 65nm. IBM/AMD are slowly closing the frequency gap betwen their and Intel's 65nm, but it is irrelevant since Intel are already abandoning their 65nm and are focused on their 45nm high-k.
AMD doesn't have enough resources to fight on all fronts with Intel and they are struggling to compete in production process, although IBM is doing most of the AMD's homeworks on that field.Quote:
I think a lot of people forgot that high-k is actually quite amazing in terms of process. AMD's not going to get it till 32nm, but I don't think it's at any aspect fair for those guys to get slugged that hard because they're less radical on processes.
I was also comparing it to 6400+.Quote:
Originally Posted by saveus222
Quote:
Originally Posted by gOJDO
I think that the 6400+ won't be cheaper than the 6500.Quote:
Originally Posted by saveus222
Got it! He said, he was comparing it to the 6400+! AMD would loose money selling only in the channel as those small amounts aren't even worth packing IMHO!Quote:
Originally Posted by saveus222
i agree. i dont know how theyre even managing to sell their current processors at such low prices.. but i guess thats why theyre doing the spin off and all that.. i guess thats the only way tey could go.. lets see what happens :) but i think im gonna get myself one of these :DQuote:
Originally Posted by Donnie27
I see your point and maybe, who knows.Quote:
Originally Posted by saveus222
I still think they'll sell well anywhere near that listed price. I'd think about $89 would be better or settled on later. I might upgrade the wife's 3500 and old A8N 16 with one as well. I work on computers on the side and need the practice:up:
I'm pretty sure this is an X2 6500, not 6500+... its a model number, not the PR rating used on the old K8 X2s.
Anyhow, the X2 6400+ is irrelevant since it is EOL, the fastest X2 currently is the 6000+. Its priced at around $90 and is faster yet cheaper than the X2 6500. Yes I know it'll be faster than K8 X2s at max overclock but when has a chip been priced based on its overclocking abilities? From an enthusiasts/overclockers perspective this is by far the fastest dual core CPU from AMD but its still way behind any 45nm C2D in this regard.
Finally someone points out the right model number.There is no + in the model number and it has nothing to do with older 6000+/6400+ Athlon X2s.Quote:
Originally Posted by Epsilon84
I was looking for a BIOS upgrade for my Asus M2N-E board and I found out on the Asus CPU support page that the new 1504 BIOS support this processor ------> Athlon 64 X2 6500+ ,2.3GHz,1MB,rev.B3,AD6500WCJ2BGH,95W,SocketAM2Quote:
Originally Posted by informal
The plus is still there.
But do you think everyone will understand it this way?Quote:
Originally Posted by informal
This is a much harder question to answer ... the quick answer is a bit of both.Quote:
Originally Posted by Macadamia
A somewhat simplified analogy is to think of the CPU as a progression of simple to complex. Transistor level, circuit level, then architectural level in that order much like cell, organ, organism. On the transistor level, the performance is centered around process and design of the physical transistor, and within the device, there are a multitude of different transistors used for different desired electrical properties... example, transistors making the bits in SRAM are different than the transistors making the logic circuits ... SRAM transistors are much smaller and also require higher minimum voltage to operate (by this I mean a higher minimum limit say 1.1 v as opposed to 1.0 volt minimum to work without error, just an example).
Today's transistors operate with a switching time of 1.0 -1.5 picoseconds, to simplify this lets assume 1.0 picoseconds (if you don't believe me, read for example IBM's process tech papers and they will quote you a td or gate delay time of about 1.2 ps for thier PMOS transistor on 65 nm). This is the time required for the transistor to charge the channel and turn on, so it is the limit at which it switch on the fastest. Let's double that number just for sake of engineering error, you have a transistor today that can switch as fast as 5x10^-12 seconds, or in frequency (1/td) = 200,000,000,000 or 200 GHz ... so why don't we have 200 GHz processors???
The answer lies within understanding that the transistors are strung together to make the circuit and the total gate delay of the circuit determines the speed of a circuit. That is, the total capacitance (which determines the delay in an oscillating circuit) is additive or an inverse of inverse law when summing the circuit. Circuits are built into logical blocks and stages (i.e. pipeline stages -- more on this later) that ultimately also add up to limit the total capable clock of the circuit. Circuit designers will measure this total delay in a metric called FO4 or fan out of four (which is a simplified measure of 4 inverters chained up). The total FO4 delays are a limitation by design and how they target their work to hit a target clock (read more about this here on IBM's power6: http://www.research.ibm.com/journal/rd/516/tocpdf.html the limited, for example, critical FPU delays to 13 FO4's).
In the end, two things in general affect the clockability of a processor -- the strength or switching speed of the individual transistor (process driven) and the depth of the circuit (design driven). Generally speaking, the more complex (i.e. higher transistor count) that goes into the functional block making the task work, the higher the FO4 delay, the slower the ultimate clock speed for a given process type.
This is why long pipeline CPUs are said to be designed to clock higher -- here is how it works. An OoOe superscalar processor does 5 general things when running code -- fetch, decode, reorder, execute, retire (unreorder). The first three are complex and what is sometimes referenced as the pipeline. The work done in the fetch, decode, reorder phase can be broken down in to stages, and but the total work through the three is always the same. Thus if I break it down into 10 stages (say 3 for fetch, 3 for decode, 4 for reorder as an example) I can get there but designers have broken it down even further into say 30 stages to do the same amount of work.
The complexity and transistor spent per stage in a 10 stage design is much higher, FO4 delay is much longer, and clockability is not as good as a 30 stage design where each stage has fewer transistors, lower F04 delays, and hence higher clockability. This is mother natures cruel little joke ... to extract high IPC and better per clock efficiency one must add transistors to the equation, but adding transistors also makes clocking the device harder.
It gets even more complex than this... beyond my understanding (I am always trying to learn more), but in a nut shell this is how both process and design play in.
AMD has generated a elegant native, monolithic quad core CPU -- but elegant does not make it technologically superior. AMD has great technology going into the processor but it also has some baggage holding it back... Hector called it the most complex x86 CPU to date, and his is absolutely right and high complexity is harder to clock -- for reasons I gave above and if you read the IBM paper reasons you will better understand. The inability to clock higher on 65 nm is a combination of a slower overall transistor in the 65 nm process combined with a tremendously complex design.
Intel is advantaged over AMD on both fronts, which is what is making the Intel products so potent in the battle of the big two... Intel wins on IPC and they win on process which drives the clock equation (to a large extent). Intel also uses 3 simple decoders and 1 complex, AMD uses 3 complex. Frankly, I am impressed that Conroe could clock as high as it could considering Intel more than halved the number of stages in the pipeline. The one area AMD really still holds the advantage is in aggregate BW, but the advantage doesn't show up until 2 socket high BW server loads, and pretty much most 4P setups (except it appears Dunnington is changing the game in that area as well).
Jack
In junction leakage this is true, SOI will always have lower leakage, but this is third on the rung of potential leakage paths ... the 3 dominant paths are sub-threshold (source to drain in the off state, or sometimes denoted Ioff), gate leakage (which we all know is addressed with Hi-K), and junction leakage.Quote:
Originally Posted by gOJDO
AMD's problem is gate leakage at 65 nm, and this can be seen by the rapid take off in dynamic power as the clock speed goes up even 200 MHz. Gate leakage is not solved by SOI and only appears when the transistor is in the on state. AMD can show great idle power, but the dynamic power is just horrendous at this point. They will be fighting this again with 45 nm to an extent I suspect.
Jack
lol i dint know your wife was heavily into overclocking :D :clap::up:Quote:
Originally Posted by Donnie27
how many times am i gonna repeat this.. we cant compare c2d to amd processors now.. period.. no one is.. everyone knows it cant touch c2d....Quote:
Originally Posted by Epsilon84
That's like saying one can't compare a GTX280 to a 4870X2 because 'everyone' knows the 4870X2 is faster'... yeah, how ridiculous does that sound? :rolleyes:Quote:
Originally Posted by saveus222
Its on the market, so its fair game and open to comparison... if you don't like em, don't read em, no harm done.
Thankyou for the very informative post, I learnt a lot from it. I always love reading about low level stuff like this.Quote:
Originally Posted by JumpingJack
Opening the black box is so much fun :D
hahahahahha.... totally different context altogether.. for one.. nvidia didnt have the dominance that intel has over amd... nvidia vs ati is comparable... intel vs amd is not at this point..Quote:
Originally Posted by Epsilon84
Umm, Intel is AMD's direct competitor in the CPU market, nVidia is AMD's direct competitor in the GPU market...Quote:
Originally Posted by saveus222
But whatever, lets mix and match comparisons to your liking. Maybe we should inform all the major review sites not to review Deneb with Nehalem as comparison, as that would be 'unfair' to poor little AMD? :up:
Since the results that were linked via hardspell are gone(and the original review is gone too),here are the direct links to the pictures:
note ,while looking at the graphs,that Athlon 6500 works at 2.3Ghz by def.
Synthetic crap :) :
Sysmark2007
http://img.inpai.com.cn/article/2008...09dac672d2.jpg
Excel:
http://img.inpai.com.cn/article/2008...d67c10dd54.jpg
Cinebench10:
http://img.inpai.com.cn/article/2008...6afec339f4.jpg
1080i MPEG to AVI
http://img.inpai.com.cn/article/2008...d0fb6279d0.jpg
3DVantage
http://img.inpai.com.cn/article/2008...d81dd508ee.jpg
Quake4:
http://img.inpai.com.cn/article/2008...63107743df.jpg
HL2:
http://img.inpai.com.cn/article/2008...9503a0a30c.jpg
Crysis:
http://img.inpai.com.cn/article/2008...e114a0a76f.jpg
LostPlanetDX10:
http://img.inpai.com.cn/article/2008...4e4202e1f5.jpg
WiC:
http://img.inpai.com.cn/article/2008...81c8942019.jpg
All around pretty strong perf. even at lowly 2.3 core/1.8Ghz NB speed.
awww.. thats sweet of you.. why dont you try fud.. mebbe he'll listen atleast.. then ill be happy :D :up:Quote:
Originally Posted by Epsilon84
With OC about 3-3.5GHz x2 6500+ and NB 2400MHz it is killer AMD X2 Brisbane procesors :)...:welcome:
Post #190 shows why I said $89 for the 6500 Kuma:up: Big deal on it the + or not.:DQuote:
Originally Posted by saveus222
What is strong about it? All I am seeing is something roughly comparable to E5200 perf, which sells for $84 at newegg, while it will certainly consume more power and almost certainly not overclock as well as the Intel processor.Quote:
Originally Posted by informal
Anything more than $75 bucks for it is a rip-off.
Although i agree on the price and OC potential,my main point was the perf. is good when compared to ~200MHz higher clocked Penryn dual core and Brisbane @ 2.8-3GhzGhz.Quote:
Originally Posted by Periander6
also, higher NB speeds make a difference worth noting. Same goes for Intel's FSB, but that's mainly at more demanding 3D action, requiring more bandwidth.
Now we need 6mb L2. Is there going to be a dualcore Deneb spin-off?
No, but there is going to be that Tri-Core based off Deneb. 6MB cache across 3 cores (compare that to the current ones that share 2MB L3 per core).Quote:
Originally Posted by biohead
For dual core, it's a 0MB L3 + 2MB L2 solution. I wonder how it performs, but hopefully it's not a step back.
That's because they use thicker gates , 1.5nm vs 1.2nm for Intel.What is strange , in PDFs describing the 65nm process from 2005-2006 they mentioned 1.2nm IIRC.I guess tests showed they need thicker gates to control leakage.However , thicker gates = slower transistors => low clocks.A side effect is very good power consumption at low clocks which literally takes off a few hundreds MHz higher. ( I'd say their power shmoo curb takes a nasty increase over 2.5GHz )Quote:
Originally Posted by JumpingJack
Since they did nothing serious to address this at 45nm , these problems will haunt them.