Well, it looks like sb is in for a world of hurt. :D
Printable View
Well, it looks like sb is in for a world of hurt. :D
Factor in the fact that the comparison was to a part running at 22.5% higher clockspeed on the games that are single-threaded. ;)Quote:
Originally Posted by Loque
The retail parts will have turbo active, and will slightly outclock the older parts, in addition to their significant IPC advantage.
Looks like AMD is about to get Conroed, again. ;)
Fixed. :welcome:Quote:
Originally Posted by flippin_waffles
So ~35-40% single thread perfomance advantage over Thuban at the same clocks? How's AMD going to close that gap? :eek:
Obviously with more cores.. but for singel threaded performance... :shrug:Quote:
Originally Posted by STaRGaZeR
So maybe informal's guesses about BD single thread performace increase over Deneb wheren't that farfechted at all... or better they are neccecery to keep the same performance gap as they have now when you compare deneb/nehalem.
http://img.photobucket.com/albums/v2...ozer-44583.jpgQuote:
Originally Posted by STaRGaZeR
Yes, it ain't gonna be pretty.Quote:
Originally Posted by xVeinx
you must really love intel:p:Quote:
Originally Posted by terrace215
Anyway i hope AMD will release competitive (enough) products
Well looks so-so for overclocking enthusiasts, it much depends like anandtech pointed out several times on the pricing of the K SKUs and OCability. SB looks a lot more attractive to non-overclockers tho if those prices will be true cuz u get a decent boost in clock frequency for the same price as previous gen. :p
Really wonder how far the unlocked chips can be clocked. On the paper with signicantly higher stock speeds and less power consumption and 32nm it sounds like it could have a decent OCability if you simply look back and compare overclockability of say Conroe - todays CPUs. Imagine if 4.7 ~ 4.8GHz was possible on high vcore (1.3~1.35v) and 4.3~4.6GHz on say 1.2~1.25v or sth like that, then paying the price premium for the K SKU wouldn't be an issue for overclockers either I think. :p Pure speculation though and I may be very far from the reality but 1.0v stock vcore (prolly just energy saving state so could actually be sth like 1.10~1.15v set in bios though), 3.4GHz clock speeds from launch, lower power consumption and 32nm sounds like a good recipé from an OC standpoint...
AMD doesn't need to give anything to a review site. They can simply release the information on a slide at some future event. Let's hope for the sake of sane CPU prices that AMD can get something out the door before September 2011. I'm personally going to SB myself as I'm not going to wait for AMD. I had a spark of enthusiasm in the other AMD thread, but after seeing some SB numbers, that spark is extinguished.Quote:
Originally Posted by terrace215
I would wait for the S2011 SB and by that also BD should be close around the corner.Quote:
Originally Posted by freeloader
funniest post in the thread :rofl:Quote:
Originally Posted by flippin_waffles
i do want amd to drive bulldozer all over intel's lineup, shake up the market a bit, bring prices down, and drive some more inovation. however, amd has had some real problems delivering on the cpu side of the business and i'm worried they might be trying to pull an nvidia (show world + dog what they're working on without having any working samples or knowing where performance will be).
i guess we'll know for sure q2 2011...
http://host.trivialbeing.org/up/tran...-bulldozer.JPG
My god, I am almost ashamed to own intel with the collective-patting-on-the-back of terrance, stargazer, and co.
FPS up--> IQ down ! ---> Intel Graphics HD !
Would be real nice with benches with turbo off on all chips, or better yet, with turbo on on all chips.
Even if you can do some calculations, you still can't do any accurate perdictions. :(
You can of course never get to much processor speed. But now it seems like we are getting spoiled here. :D I remember when processor speed mattered more than graphics cards, and when you actully felt a big difference from going up a notch. Of course a speed bump then was more like 15-30% higher frequency, against 3-5% today.
Haha well, it looks like you're in for a world of hurt also. ;) btw, how are those 32nm issues at intel coming along? Surely they've got it all worked out by now. No? Oh. :DQuote:
Originally Posted by terrace215
where are those bulldozer benches?Quote:
Originally Posted by flippin_waffles
hopefully BD will have a smaller gap than today's Nehalem vs Deneb...
Probably not going to show up for a while now if this is all they have to compete with. Thanks a lot intel...:p:Quote:
Originally Posted by 570091D
Not feeling the ~$500 for a 4c 8t overclockable 2600K
Don't tell me you're one of those people that gets carried away with boilerplate risk disclaimers? :rolleyes: BTW, I don't work for Intel.Quote:
Originally Posted by flippin_waffles
:ROTF::ROTF::ROTF: The first stage of grief is denial...Quote:
Originally Posted by flippin_waffles
We never would have guessed, looking at your posting history. :rolleyes:Quote:
Originally Posted by cegras
You mean all the times in the past few months where my posts were those asking you to stop being arrogant in requesting info, making overblown assumptions about AMD, and generally spreading FUD? I mean, I haven't contributed anything more than that other than calling you and your company out for the collective-back-patting.Quote:
Originally Posted by terrace215
Do I have to go through this thread and quote all of you to show you how terrible it's getting? Let's try:
Considering you can play games at 60 fps on 15" resolution with the lowest details, I really do see why you're so excited. I really do. (I really don't)Quote:
Originally Posted by STaRGaZeR
Quote:
Originally Posted by terrace215
Quote:
Originally Posted by terrace215
Quote:
Originally Posted by terrace215
Quote:
Originally Posted by STaRGaZeR
Quote:
Originally Posted by terrace215
Quote:
Originally Posted by STaRGaZeR
Quote:
Originally Posted by terrace215
Quote:
Originally Posted by terrace215
^ My god, you're so excited about SB you had to double post.Quote:
Originally Posted by terrace215
Quote:
Originally Posted by STaRGaZeR
http://www.8thcivic.com/forums/image...s/facepalm.gifQuote:
Originally Posted by terrace215
Ironically displayed by yourself.Quote:
Originally Posted by terrace215
Sandy Bridge does not appear to be even worth the platform upgrade for anyone owning a nehalem.
Poor show.
Yeah, 20% IPC increases just fall from the trees everyday... totally worthless! Meanwhile, Phenom II owners will... what? wait until Q4 of next year to buy a new BD platform for a Zambezi that will be significantly behind the competition? That's right, BD is officially NOT Am3 compatible:Quote:
Originally Posted by Jowy Atreides
http://www.xbitlabs.com/news/cpu/dis...forms_AMD.html
Copy pasta, since I couldn't have said it better myself:
Quote:
So the 3.1 GHz Sandy Bridge is 23% faster than a 2.8 GHz Nehalem. So roughly 10-15% faster clock for clock as Anand says in the teaser intro on the main page. Also, you get 10% less power. Not bad. However, that means that Bulldozer only has to slightly perform better than Nehalem to match Sandy Bridge clock-for-clock and live up to the 10-100W power envelope.
I change my vote on the weekly poll to match Sandy Bridge performance.
Quote:
Be careful when comparing clock-for-clock. In the comments to the article, there was this exchange:
------------------------------
"I take it turbo was also disabled on the rest of the parts used to compare, right?"
"Turbo was enabled on everything else - SB performance should be higher for final parts.
Take care,
Anand"
------------------------------
The Nehalems are tricky: given half-way decent cooling (which I'm sure AnandTech has), they will Turbo even when all 4 cores are active. Thus, the supposedly 2.8 GHz Nehalem actually runs at a minimum of 2.93 GHz and goes up to 3.33 GHz in single or dual threaded tasks. A lot depends on what Turbo scheme Intel has chosen for Sandy Bridge.
but there is no 2.8 GHZ Nehalem in this test..Quote:
So the 3.1 GHz Sandy Bridge is 23% faster than a 2.8 GHz Nehalem.
760 is 3.06-3.33GHZ
880 is 3.2-3.7GHZ
What a shocking revelation. That I posted about last week.Quote:
Originally Posted by terrace215
BD will bring a bigger jump in perf than nehalem to sb.
Intel is about to get athlon'd
880 is 3,33ghz (2 speedbins ;) )
And if you look at Singelthreaded scores, the increase in performance is quite respectable... a 3,1ghz cpu beats a 3,73ghz cpu (ST)? Or in MT apps a 3,1ghz cpu beats a 3,33ghz cpu, and most of the time we are in the double digit range...
So offering more performance then a 880 and consuming 20% less energy while doing it.
I want to see 3dmark 01 & 05 on SB :D
You know that amd has deliver a performance boost of 35-40% over deneb (per core) to be in the range of SB, and thats with the current numbers, without turbo.Quote:
Originally Posted by Jowy Atreides
this is great news and a good performing chip + gpu, interesting times ahead. i'm still excited on what ontario/llano brings to the table though
In about 14 months. 'nuff said.Quote:
Originally Posted by Jowy Atreides
I am predicting 1q 2012 release. Seriously. :yepp:
not 1 ?Quote:
Originally Posted by Hornet331
1/1/4/5
http://en.wikipedia.org/wiki/List_of...icroprocessors
2 failures i must fix.:rolleyes:Quote:
Originally Posted by flippin_waffles
have you even read natarajan's paper? no. do you even understand the first thing about what makes a process good?
for one thing it is quite concerning when at iedm 2009 there were many papers on new 32nm processes and none of them were from global foundries. intel has the best gate and metal pitch, the fastest drive current, almost 2x faster pmos performance, and the fastest SRAM. seeing that they recently posted record profits they must be doing something right with their god awful manufacturing and architecture.
fail. IPC is a misleading performance metric. you must have never multithreaded code or understand how it works i take it.Quote:
Originally Posted by terrace215
This may be a ridiculous, but could someone please explain what "HT on" means on an i5 part stated to have 4C/4T?
Quote:
Originally Posted by meanmoe
On page : The testQuote:
In order to help Intel’s partners test HT functionality however, the i5 2400s being sampled right now have Hyper Threading enabled. For the purposes of our test I’ve run with HT both enabled (to give you an idea of higher end SB parts) and disabled (to give you an idea of i5 2400 performance).
thank you. somehow I missed that.Quote:
Originally Posted by Olivon
Yep, or how about "KO'd". :DQuote:
Originally Posted by Jowy Atreides
I am not sure how this is gonna happen, other than just brute force more cores (modules).Quote:
Originally Posted by flippin_waffles
AMD's presentation was 33% more cores for 50% more throughput over MC, amortizing 50% over the core count does not give them the IPC improvement you guys are thinking it will.... the math just doesn't work.
Now, if AMD matched module count to Intel physical core count, it may turn out they are in good shape... but that would be one massive CPU in terms of die size.
That's easy to do when your starting point is lower. But what does that mean for BD vs SB? In Anand's WoW test SB was 50%+ faster than a higher clocked Thuban and if Turbo was in fact disabled then it's only going to get worse. So are you saying that we should expect 50-60% performance improvements from BD too?Quote:
Originally Posted by Jowy Atreides
I'm loving the performance per watt; not so much loving the new chipset locked 100bclk scheme although the K chips help... but it still sucks. Mid-end chips are Meh! and integrated graphics and transcoding silicone on a desktop can suck it! save that crap for laptops. Wish they would go ahead and step up their game with a Hexa/octo core launch of Sandy Bridge instead of 2c/4c launch. :down:
You're forgetting that AMD's plan isn't to compete core per core with Intel, but thread per thread. One Bulldozer module competes with one hyperthreaded core. Not saying that it isn't a long shot for AMD to equal/beat SB, but BD is a new kind of chip that challenges the way we typically think about performance.Quote:
Originally Posted by Hornet331
If I remember correctly, an Intel-FTC settlement forces intel to provide an interface for its cpus in next 6 years. That would be a big opportunity for nvidia to design chipset especially for overclockers.
A PCIe interface... nVidia is pretty much out of the chipset business.Quote:
Originally Posted by kl0012
Might share with us your insights about Intel's 32nm problems ? I'm sure Intel's management would love to have such problems ( which lead to 60%+ gross margins ) all the time.Quote:
Originally Posted by flippin_waffles
sig'd :DQuote:
Originally Posted by Jowy Atreides
the clock generator is now on-die with SB so intel can lock it down and no-one can touch it. :down: the free performance upgrade was fun while it lasted. :shrug:Quote:
Originally Posted by kl0012
Considering how far behind Nehalem Phenom II single-thread IPC is, it now looks doubtful that they will even narrow the gap by much.Quote:
Originally Posted by Jowy Atreides
Prediction: As a client processor, BD is basically dead. The core performance that the turboless SB sample showed at Anandtech is just too strong. Server & HPC it should do ok, modulo GloFo's 32nm process continuing to flounder.
On the Llano side most likely yes, on the BullDozer side its pointless to speculate at this point. BD is completely different from Phenom II as you know. Its anyones guess what and when it will bring to the show.Quote:
Originally Posted by terrace215
guys, its no corect good full test of SB, but "pro Intel" marketing preview...No atack to SB, it will good CPU, but now its only Intels reaction at Bulldozer technology dates from Hotchips
These mainstream SB seams to be meant (the preparation) for great mobile CPUs.
The clock-4-clock performance increase is impressive for these CPUs, but this IGP would make a lot more sense in a laptop than desktop, this combination will make a big splash on laptops, soon :)
These mainstream SB is not too bad for desktops, but the high-end SB would decide the real impact on desktop. I think we will see a much greater impact on laptops than desktops in coming months.
lol ignore lists are more useful everyday! anand isn't always straight when it comes to Intel. ill wait till the cpu comes out...not that i would buy it...just curious.
:yepp:Quote:
Originally Posted by god_43
SandyBridge is an evolutionary step for Intel Nehalem / Westmere architecture.
~10-20% IPC gain is not that much.
Bulldozer was a tech slideshow with no performance numbers, this is a performance preview bordering on a full review. This is a sign of :banana::banana::banana::banana:iness more than anything from intel. They are showing AMD they are not scared and might as well reveal its cards now because they are confident AMD will not show them up.Quote:
Originally Posted by FlanK3r
When anand said about the K series he did not mention the AMD's unlocked processors as one of the reasons why he K series exist.
Interesting on K version
Is it? I wonder if ppl are serious saing such things.Quote:
Originally Posted by qcmadness
Still no info about OC options for 2011 socket CPUs :mad:
Of course not, they want you to purchase their current goods now, instead of playing the waiting game.Quote:
Originally Posted by Rebel44
At least for now it seems overclockability through bclk (dmiclk) remains unchanged on S2011 platform.Quote:
Originally Posted by Rebel44
http://www.bit-tech.net/news/hardwar...overclocking/1
http://images.bit-tech.net/news_imag...el-p67pch2.png
http://images.bit-tech.net/news_imag...intel-2011.png
Single thread performance is chaotic also expect exorbitant prices for K editions, will be worth? lol
I'm seeing a light in the end of a tunnel, is it AMD? lol
http://images.anandtech.com/reviews/...roadmap_sm.jpg
Price to segment would be same/similar.
i5-2500K in the price range of i5-760 looks interesting.
2500k version for only £150. It is too good to be true. I will believe when I see it.
http://www.scan.co.uk/Products/Intel...21x-95W-Retail
Doesn't Fusion make all of this irrelevant?
2500k for £150 wouldn't surprise me, because this is the segment where AMD can compete. It is too early to speculate on prices, but my bet would be on really low prices for the whole range of mainstream SB.Quote:
Originally Posted by Metroid
So will X58 / I7 970 remain high end and a valid upgrade path instead of Sandy Bridge?Quote:
Originally Posted by ajaidev
Dont think so, anand simply doesnt have any info on the high-end roadmap. Notice how they added ">" in front of the high end i7's noting either price drops or Sandy Bridge replacements. ;)
with the exception of Video encloding Sandy bridge Corei5 almost matches hexcore 980X and with this power envelope ...... holy :banana::banana::banana::banana:e and that too from Mainstream 661 equivalent.!
http://images.anandtech.com/graphs/s...2107/24416.png
LOL.. do people actually understand what they are saying?Quote:
Originally Posted by qcmadness
10-20% increase in IPC (on x86) not much?
10-20% better IPC is really good :up:
Looking for LGA 2011 iterations ... ;)
No, for Intel, it's not much. If it was AMD, however, it would be a MASSIVE increase. I wonder why :rolleyes:
On topic: I find it hard to believe that an LGA 1155 part is matching the 980X in so many cases AND consumes less.. and the IGP performance with only half the shaders enabled is really promising. Makes me wonder what the high end would do to today's parts. :rolleyes:
Quote:
Originally Posted by RaV[666]
There is no point rationalizing Intel's poor decision to deliberately crippled SB with regards to overclocking.Quote:
Originally Posted by Movieman
It not like Intel's sales were harmed with C2D and i Series overclocks. IMO these things happen with monopolies. Look at how intel brought back overclocking big time with C2D. when there was good competition from AMD. The original reason to overclock is to get as much performance as possible out of your system, specially useful if you are short on money Collage, new job etc. Businesses wont overclock their office systems anyway nor will pro server/rendering farms
This is just hurting the enthusiast crowd who want more bang for their buck.
On the other hand there's no point rationalizing people's feeling of entitlement either.Quote:
Originally Posted by geo
Isn't K-series targets espetially the "enthusiast" crowd? After all it seems that mainstream K-series is going to be pretty affordable and makes the "enthusiast" life much easier.Quote:
Originally Posted by geo
Also theres nothing new in SB limited overclocking. I mean this is not because the overclocking is locked on the chip level, but because the lack of appropriate PCI-e multipliers inside the cpu. There was chipset in the past without appropriate AGP multipliers and this didn't stop a real enthusiast from an overclocking, it just was a bit harder job. I mean overclocking these days is not what it was some time ago.
well, since people pay for the cpu's it isn't difficult to imagine people getting the feeling of entitlement :rolleyes:Quote:
Originally Posted by trinibwoy
I didn't limit "enthusiast" to LN2 guys. if you read my reply i mentioned people in Collage, new jobs etc who aren't going to buy $1000 cpu's.Quote:
Originally Posted by kl0012
yeah. K series is fine but i doubt the limited multipliers can just be used top push the cpu more than 400-600mhz .DrWho said, SB overclocking isn't limited but has a different way of overclocking . well, now Anand is saying something else.
i
i agree, this along with amd's solution will hurt nvidia something fierce. and if the amd part will crossfirex with a regular videocard, there will be big trouble for nvidia.Quote:
Originally Posted by informal
SourceQuote:
Today's statement also says of the risk factors, "Intel is in the process of transitioning to its next generation of products on 32nm, and there could be execution issues ... including product defects and ... lower than anticipated manufacturing yields".
Yes exactly, you pay for the CPU and its advertised speeds and features. You don't pay for the right to overclock it.Quote:
Originally Posted by geo
Wasn't going to believe the inquirer, but I clicked on their source link and indeed it comes from Intel:Quote:
Originally Posted by slaveondope
http://newsroom.intel.com/community/...w-expectations
I highly doubt they'll have those issues just because they already have current 32nm hex-cores.
I'm sure such things are always listed in the "Risk Factors" sections of various reports.Quote:
Originally Posted by richierich
You want better OCing, PAY MORE.Quote:
Originally Posted by geo
:rofl:
Seriously, WHERE does it say Intel has to sell you a 4GHz CPU at a 2.6GHz price point?
because green side do it :up:Quote:
Originally Posted by mstp2009
And we all know why they do it...Quote:
Originally Posted by madcho
They didn't when they were on top.Quote:
Originally Posted by madcho
Remember the FX? AMD only does this b/c they HAVE to. If they were in Intel's position, as they have shown in the past, they would not.
You should realize Intel's business model is not sustained by high-performance, high profit margin components, but by decades old legacy support in hardware.Quote:
Originally Posted by tajoh111
Releasing information on Sandy-bridge is not some brilliant tactical/strategic maneuver. Sandy Bridge is here. Release is just a few months away. Bulldozer release is probably at least a year.
Have you had a real conversation with JF-AMD? Read his posts? "Can't talk about it/Don't know about it/Customers care only about performance&wattage." Trying to coax information from this man is harder than getting Glenn Beck to admit he needs round the clock psychiatry.
any one know what happened to the rumor that there was specific silicon for "transcoding" ?
Can't say these numbers are convincing me to upgrade.
This performance jump is good enough, actually more than enough to convince many of those who are using 45nm CPUs now. But it won't have the WOW-convincing-effect on performance before the high-end hits the market later, and it's natural of course.
These mainstream CPUs will be really convincing for those who want a cool and power efficient 32nm CPU, but don't need (don't want to spend 0n) high-end 32nm 6-core+.
Hmmm, not convinced enough to switch from the current crop of i7's but for the average user it's a convincing upgrade. And since were on xtremesystems, we'll just disregard the average user stuff hehe
OC crowd is very very tiny as a % of intel business. Intel traditionally allowed flexibility here because this crowd was vocal and it was useful to help drum up online buzz. AMD can't seem to get their act together, so dealing with issues associated with allowing people to muck around with ever more sensitive voltages and other parameters probably isn't as appealing to Intel.Quote:
Originally Posted by geo
We shouldn't disregard those enthusiasts that still are running on those hot and power hungry 45nm i7. Many of them are waiting, because 980x and 970 is not exactly cheap. You wouldn't call those as "average" and disregard yourself too, would you? hehe.Quote:
Originally Posted by Chaserjzx100
:rocker: Pentium 4 rulezzz :rocker:Quote:
Originally Posted by madcho
:ROTF:
I've been going through the numbers, and it's looking something like this..
The non-HT SB @ 3.1ghz is 25% faster than the 2.93Ghz core i5 760 in tedts that SHOULD be using all cores, or close to, and keep the 760 down at 2.93. If Freq-perf scaling was 100% that makes the tested part 18% faster clk/clk in these tests
Comparing the HT enabled chip to Anand's Bench results in the same manor using a Core i7 870, which should be running at either 3.06Ghz or 3.2 with these 4+thread benches is around 13%-17% faster clk /clk. This indicates HT performance scaling is either no better, or lower than Lynnfield, but it's hard to say when you're attempting to guess the frequency with each benchmark.
If these tests were really done with Turbo switched off, it's an impressive IPC increase. and some of the induvidual app increases are ridiculous. (Like Photoshop CS4) Especially on what we assume to be the same base architecture.
Very keen to hear the details of the Uarch.
Regarding HT.. Is it possible the efficiency increases achieved have left less room for HT?
Ahhhh, the Inq, back to their old ways....Quote:
Originally Posted by slaveondope
AMD says they can't make 32 nm processors and will hurt revenues:
SourcesQuote:
We rely on GF to manufacture some of our products, and if GF is unable to manufacture our products on a timely basis and on competitive process technologies or to meet our capacity requirements, our business could be materially adversely affected.
nVidia says they cannot transistion to new technologies and will hurt revenue:
SourceQuote:
The inability by us or our third-party manufacturers to effectively and efficiently transition to new manufacturing process technologies may adversely affect our operating results and our gross margin
Every quarterly report has risk statements, it is the legal part necessitated by the SEC for every company to inform shareholders of risks to the company that will affect their financial performance.
Noticed how they didn't include AMD X6 performance on Cinebench 11.5 multiCPU score? :down:
http://images.anandtech.com/graphs/s...2107/24409.png
Quote:
Originally Posted by slaveondope
Not hard to put it in yourself, CBR11.5 is pretty reproducible, my 1090T scores around 5.1 to 5.14 on this bench.Quote:
Originally Posted by Davey
They are all quad cores, and the X6 isn't really going to light that test up anyway as they'd have to put the 980X in too.Quote:
Originally Posted by Davey
they must be saving lots of money by locking the cheaper CPUs multiplier (and in this case killing the possibilities to modify the clock speed), if they want to charge for "auto OC" fine, but I like to have the possibilities to do a fine tuning on the cpu settings that I buy, and to work at the max speed this piece of hardware can (if this means no warranty, less durability, more power usage, untested regimes of operation, I have no problem in accepting these risks) and I find it convenient, and that's not the way it comes from the factory, for many reasons, by their point of view I understand, they want to maximize even more their profits, but if it's true that 95% of the users of cheap CPUs don't care, why bother locking the multiplier? they want this 5% to pay an extra just because they care and know how to make a better use of the CPU adjustment settings? if making OC possible had aQuote:
Originally Posted by mstp2009
great cost, in that specific part, I wouldn't see a problem,