Slashdot Mirror
Details of New Intel Dunnington and Nehalem Architectures Leaked
Daily Tech is reporting that details about Intel's new processor models were leaked over the weekend. Both the six core Dunnington and Nehalem architectures were featured in this leak. "Dunnington includes 16MB of L3 cache shared by all six processors. Each pair of cores can also access 3MB of local L2 cache. The end result is a design very similar to the AMD Barcelona quad-core processor; however, each Barcelona core contains 512KB L2 cache, whereas Dunnington cores share L2 cache in pairs. [...] Nehalem is everything Penryn is -- 45nm, SSE4, quad-core -- and then some. For starters, Intel will abandon the front-side bus model in favor of QuickPath Interconnect; a serial bus similar to HyperTransport."
Sounds like good names to be used in a D&D game!
Sir Dunnington against the evil lich lord Nehalem!
They could have gone to 3 cores, like the competition. That seems like the logical thing to do, but they said "Fuck it, we're going to six". What part of this don't you understand? If two cores is good, and four cores is better, obviously six cores would make them the best fucking CPU that ever existed.
http://www.theonion.com/content/node/33930
/I'm just waiting for the day Intel says "this one goes to 11"
[Fuck Beta]
o0t!
The Wikipedia page on QuickPath is very lacking in the realm of details. Does anyone know how it stacks up against HyperTransport? One of the most mouth-watering proposed uses for HT3 that I've heard of was the possibility for an external HT3 bus on a machine which could be used to link together multiple physical machines into one giant NUMA beast.
;)
Imagine a Beowulf of those
Still doesn't run Crysis.
QuickPath: because Intel doesn't adopt standards... it rewrites them.
See article.
The world is made by those who show up for the job.
The L3 cache is 16MB. Each pair of cores shares 3MB of L2 cache. They aren't the same thing at all.
Note: if you're tempted to mod this up, don't. I rehashed the summary.
You seem to have confuse the L2 and L3 caches. The L3 cache is 16MB, while each pair of processors have a shared 3MB L2 cache. So, it's 3pairs X 3MB = 9MB of L2 cache and 16MB of L3 cache.
Uh, no.
It seems that 16 MB of L3 cache is shared among all 6 processors. Then, each pair of cores has 3 MB between them.
So, 16MB L3 + 3 (pairs of 2 cores) * 3MB L2 = 25 MB total cache.
that means we have 9 MB of L2 cache (total) and an additional 16 MB of L3 cache.
now i need to RTFA
Don't call me back. Give me a call back. Bye. So yeah. But bye our, well, but alright we are on a shirt this chill.
Does it go to 11?
Karma Whoring for Fun and Profit.
your forgetting the L1 cache, maybe 128-256 KB per core... thats up to another meg and a half!
Don't call me back. Give me a call back. Bye. So yeah. But bye our, well, but alright we are on a shirt this chill.
So, 16MB L3 + 3 (pairs of 2 cores) * 3MB L2 = 25 MB total cache.
Which is more memory than RAM of my first computer and I'm only 21 years old.
This is gonna be sweet inside a MacPro! 12 cores, wow!
"Intel will abandon the front-side bus..."
I think I speak for us all when I say ABOUT FSCKING TIME!
Actually I think you could run windows 95 just in cache. Now say about bloat...
Extreme Programming - Redundant Array of Inexpensive Developers
I don't really know the situation surrounding the technology, but even if Intel could use it for free, they would lose a huge battle in the PR War. I can see it now, "Remember that interconnect AMD has been using for years now? Well our design has finally caught up with theirs enough to use it." Remember that to the masses, the non-slashdot crowd, they have no idea what the techno-jargon spouted by Intel marketing means.
Intel currently has the superior technology, this is because of superior fabrication capabilities, not because of a superior architecture, if I've been following this correctly over the last few years. The general public is oblivious to the fact that internally the AMD architecture is cleaner and more elegant, the only thing they have to go on is marketing. If Intel were to adopt HyperTransport, which IIRC is trademarked by AMD, that would be a huge step backwards for Intel marketing, which is just recovering now that the Core 2 architecture has put them back on top.
I love how they intimate that their 3Mb L2 Cache is somehow better than the 512Mb Cache in the Barcelona. It very well may be better in their particular CPU, but functionality wise the two products may perform similarly.
The AMD HT transport product has also been on the market for years, and in real computing (non desktop) applications has serious advantages over the Core2 and other Intel CPUs. Intel's solution/redesign for a similar feature isn't even on the market yet, while AMD's is a mature product.
I also like how they talk up power consumption on their CPUs, while leaving out the 100W hair dryer that is typically their memory controller, whereas the AMD memory controller is included in the CPU and in their CPU wattage.
The only way to test these things is at the wall socket, complete systems.
yea if they used hypertransport they would have to pay amd for it
Unless they negotiated a cross-licensing deal on patents, of course.
Karma Whoring for Fun and Profit.
Which kind of puts in perspective just how long Duke Nukem Forever has been in development. It's almost getting to the point where the CPU alone meets the minimum requirements for RAM.
When can we expect this new architecture? 2011? Ok, maybe I can deal with that.
What will I need to upgrade? A new motherboard, since surely the northbridge-less model will not appreciate the northbridge currently present. Ok, a new morbo is like, 90 euros, I can handle that.
Can I upgrade the motherboard only, so that I could use my old Core2 chip in it? No, there's no northbridge. And the socket is likely incompatible since the processor is now directly connected (well with some mild glue for voltages etc.) to the memory chips. Guess I'll have to suck it up.
With a new motherboard comes new memory modules. Let's be realistic and assume that the minimum amount of memory in a new computer in 2011 is going to be at least 4 gigabytes. Anyway, regardless of the year, a new computer's new memory modules cost some 200 euros total.
This is starting to look kind of expensive! Guess I salivated early. I suppose my next upgrade will be to an AM3 motherboard and then to a Barcelona chip. I mean, those are here this year, not in three years, right?
Jeez, Intel, please provide something more quickly than just promises of Nehalem (I mean, we've been hearing about it for _years_ now) and son-of-Nehalem. Pony the fuck up! I mean, at least AMD's got _something_ out besides vapour.
But imagine what will happen when DNF finally arrives. It could be start of a new epoch (it certainly be the end of one).
Extreme Programming - Redundant Array of Inexpensive Developers
Royalty free membership must be a bad thing?
see - http://www.hypertransport.org/consortium/index.cfm
Get off my lawn. My first computer came with 1k of RAM. I still have it.
Stick Men
This is a sever chip and the FSB may get in the way and be a big slow down with you need to go to a other socket or need to load a lot of data to the cpu all of that L3 and L2 helps as well the 24meg buffer in the chipset that needs FB-dimms but all of that pushes the cost up once intel drops the FSB the need for all of that L2 and L3 will go down as well as moveing to DDR3 that gives off less heat and needs less power.
also there needs to be quick path / HTX slots not sockets for add on 3rd party chips on the cpu bus and they need to let you any chipset like how there is a number of them for amd systems on the desktop and the sever side.
Skulltrail with a just a nvidia chip set without having to use a intel one like how the amd 4x4 system is setup and may desktop DDR2 / DDR3 will be better system with lower cost and less heat if they are able to go this way.
and intel needs to have on board video with it's own ram ATI and nvidia are working on this.
Yeah, because EM64T was novel.
The base component of this is the Core 2 Duo. That is a dual-core unit, joined by a common L2 cache. What they are then doing is putting 3 of these together, and joining them with L3 cache. Hence, 6 cores. My guess is they figure that 8 cores would be too expensive, too hot, whatever to do at this point.
Piker. My first computer had three bits of memory, and it clocked at 1Hz if I was fast. No friggin kilo- prefixes for me.
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
Please check your facts, AMD doesn't _own_ HyperTransport, so why would Intel have to pay them anything? HyperTransport can be used royalty-free by anyone joining the HT consortium. Yes, AMD is a member of the consortium, just like a lot of other tech companies such as NVIDIA, one of AMD/ATi's biggest competitors. AMD are not the owners of the technology nor are they in control of the HT consortium. They are simply one of the most visible tech companies that has strongly embraced HT in their products.
...then shalt thou count to three, no more, no less. Three shall be the number thou shalt count, and the number of the counting shall be three. Four shalt thou not count, neither count thou two, excepting that thou then proceed to three. Five is right out. Once the number three, being the third number, be reached, then l...Note that Intel did adopt AMD's 64-bit extensions to the x86 instruction set. I regard that as far more significant than, hypothetically, licensing HyperTransport. For example see this article on Wikipedia or any other history of AMD64/Intel64 or "x86-64" or whatever everyone is calling it these days.
This was a PR blow to Intel, but still made good business sense at the time, and seems to have been good for Intel and for AMD (bad for Itanium though).
This is great for many computing environments, but my home system is not one of them. Honestly there isn't much software I use on a regular basis that really taxes the second core, let alone six of them.
Some people run windows, and they have to have a virus checker running all the time. Loads of activity every so often, which makes another core nice. And the window manager hangs sometimes and does these bizarre full-desktop refreshes every time you look at it crossways. It's good to have your program keep running at full tilt when that happens.
Multiple cores is the way to go, if that's your lot.
Weaselmancer
rediculous.
Though we are talking about an Intel design, where the cache isn't exclusive. Thus not including it in total usable cache would be correct.
I don't even own a dual core machine and now there's going to be Windows Craving 6-core machines that run Vista Ultra Quantum Home Edition!
I feel like they would do us all a favor if they just told us the date that none of the software we'll need to run will stop operating on 'old' hardware. I can hardly wait for my HS Jr. to go off to college and they tell me I need yet another $2400 laptop as a requirement.
When you are watching WoW traffic, have you ever been tempted to analyse the packets? I thought it might be useful to make a program that extracts English messages from a WoW bitstream, e.g. whispers and other chat. These are sent in plain text. But I discovered that the packets appeared to have no obvious structure that would allow chat messages to be distinguished from other data. The nearest thing to a packet header is a 32 bit word that appears every so often. Its position suggests a packet header, since it is at a consistent offset from chat message text, but it looks random (when passed through a frequency analysis, all bits have equally high entropy). It is as if the packet headers are being encrypted using an algorithm and key shared by both the server and client with the specific intention of making the protocol more difficult to reverse engineer.
But why just the header? Can you shed any light on this? (I claim that this is slightly on topic because it is a very nerdy discussion, and, erm.. a possible use for additional CPU cores...)
>north
You're an immobile computer, remember?
"The general public is oblivious to the fact that internally the AMD architecture is cleaner and more elegant, the only thing they have to go on is marketing."
It doesn't help that in most benchmarks, AMD has been trounced by Intel this past year.
http://www23.tomshardware.com/cpu_2007.html
I drive the coast from California to Washington state 4-5 times a year, and each direction, it is a must to stop at the Tillamook factory and drop $100 or so on cheese and whatever. Best Sour Cream in the world, too. My office has 10 or so coffee mugs from the place. The city also has a decent Air and Space museum too.
Exactly. Its DEC technology which AMD adopted. It was developed under the brand "Alpha Processor Inc" who called it LightingDataTransport.
I made the disclaimer in my post I had no idea what the licensing was for HyperTransport. However, I think it would be bad PR for Intel to adopt now what AMD has been doing for years, even if it is the right thing to do technologically. I also qualified PR to mean Public Relations with the unenlightened masses, those who know nothing of Open Source Software or Open Standards.
Intel has always been about the marketing, first it was clock speed, now its cores. Bear in mind marketing usually has very little to do with reality, yet will often drive engineering decisions. Brands are important, it's easy for us techies to forget that.
Too bad, bitches!
P.S., next time you don't want something modded to +5 Informative, try posting a link to goatse.
If a BIOS option was available to disable the DIMM slots and run purely off the L3 cache, it would be kinda cool run an OS inside it.
Life is not for the lazy.
> http://www.realworldtech.com/page.cfm?ArticleID=RWT082807020032
Can RWT please hire a typographer? Trying to read even the first page caused my eyes to cross.
The reality isn't what matters. What matters is public opinion.
It's already bad enough that Intel's own 64bit successor, the Itanium, is widely called "Itanic" and that they ended up adopting AMD's 64bit instruction set.
Now if once again they use the same technology as AMD instead of building their own that the marketing department will call "better", the public will start to think that Intel isn't able to come up with new ideas and relies on AMD to make revelant advances in CPU technology.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Intel did have a hell of a time confusing people before the concrete samples were available as to whether it was the same thing as AMD's 64-bit. They avoided using any term AMD associated with it for a time, instead tossing around ia32e and em64t and bs like that. I know some projects even baked into plans how to cope with yet another processor architecture for lack of a commitment from Intel that their 64-bit x86 compatible stuff would be the same.
Intel's hand was effectively forced because they learned their lesson from Itanium, don't screw with an incumbent variation of your flagship instruction set. With AMD's lead they would've risked yet another Itanium fiasco, so they picked the safe path and tried to PR dance around the existence of AMD's 64-bit stuff.
Itanium was an odd path in the history of Intel proving they truly thought they alone dictated the course of x86 technology. It stands in stark contrast to the history of supporting legacy all the way back to the 8086 days.
In this case, it's not the end-user or software developers being impacted, just hardware implementors who already have to do whatever the processor architecture dictates. Despite that freedom, Intel's unable to offer something that isn't obviously similar to the competing offering since it just is such a damn good idea. AMD has led some revolutionary changes in x86 architecture, while Intel has been able to follow up with evolutionary advances, fabrication, and marketing to continue eating the more significant profit margin space.
XML is like violence. If it doesn't solve the problem, use more.
Skulltrail by definition is an Intel design. There are a host of multi-socket chipsets that could do it theoretically, but Skulltrail is little more than a concerted marketing strategy around such capable parts. One interesting aspect of HT and QPI is that the planar chipsets become socket-count agnostic, just the processor you buy must support the right number of links for a multi-socket config. That said other than the '1337' factor, Skulltrail is a ludicrous waste of money. They explicitly declare it not suitable for server-class operation, and yet the only thing that currently meaningfully pushes beyond 4 cores is not to be found in the desktop space (programmers just aren't threading the intensive stuff that fine grained yet).
XML is like violence. If it doesn't solve the problem, use more.
AMD does however own the cache coherency protocol that they use over HyperTransport for processor to processor communication so Intel would either have to design their own or license AMD's implementation.
sexcores?
For that, a dual core is an excellent idea, a quad core is not. A quad (or sex?) core is only useful when your workload can be divided into 4 roughly equal parts. This is true for servers, which are running dozens of threads of the same application at the same time, this may be true for some workstations running specialist applications that are sufficiently multithreaded to make use of multiple cores, and no doubt in the future there will be games that make good use of multiple cores, but for normal desktop use, more than 2 cores is useless.
Most of the time, there's one application that you're actually using and that's actually doing something. Everything else uses maybe a few percent of the CPU. In 99% of the cases, one process will be using at least 50% of the CPU. A second core will certainly improve the responsiveness of the system, giving the less demanding processes a chance for CPU time without interrupting the main process, but more than two cores will add nothing.
By all means, buy a dual core, but before you waste money on a quad core or bigger, please check if you'll actually be able to make use of all the extra cores.
So, what you're saying is that iNTEL's ability to flood the market is the only thing keeping them on top.
Right?
I find that large jumps in technology often leave the rabbit behind, but I'm not sure I think it's a good thing.
Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
They use pasteurized milk from cattle that are fed on artificial feed, and they aren't aged nearly long enough. You've not had good cheese at all if all you've had is cheese from America. Sorry, not dissing on the US, there's plenty of things we do well. Cheese just isn't one of them.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
Because it's demonstratably better?
.
And they can show that they moved beyond the executive pigheadedness that lead to netburst...
I mean I thought about buying a CPU at the end of 2007 and I was like why would I get a core 2 when I could get a P4 (It's 2 better!) at 3.06 Ghz (Wow!) of course it did support that 3Dnow BS that AMD made (Never sell out, never compromise!)
You're talking about a tiny minority who: read CPU spec sheets, know what hypertransport is, and don't know that needing a flexible serial solution to interconnects is one of the main bottlenecks of multicore processors.
While Intel does amazing cpu design if they liscenced or incorperated some alternate technologies their processors would be even better: on die memory controllers, serial interconnects, 64 bit instruction sets etc... in the marketplace those = expensive motherboards with lower memory access, more cpu overhead and reduced performance in high stress apps which take advantage of n cores, and incompatibility wars.
Intel has the best processors, and their $200+ motherboards don't scare me away... but they could be doing better with a little flexibility.