Slashdot Mirror
AMD's Barcelona to Outpace Intel by 50%
Gr8Apes writes "AMD is upping the performance numbers for Barcelona by stating that "Barcelona will have a 50% advantage over Clovertown in floating point applications and 20% in integer performance 'over the competition's highest-performing quad-core processor at the same frequency'". AMD also claims that the new 3.0 GHz Opterons beat comparable Intel Xeon 5100 series processors in three server-specific benchmarks (SPECint_rate_2006, SPECint_rate2006, SPECompM2001) by up to 24%."
I really hope this plays out. Not only do we benefit from better technology, but I get to read all the fanboy flamewars too!
Slashdot - where whining about luck is the new way to make the world you want.
It would be a shame if after what, 4 or 5 years? of being in the lead, AMD loses focus and stops making fast CPUs.
The last thing we need is for Intel to have no real competitors. Innovation would slow and prices would hike up.
Wow! The underscore makes all the difference!
The World Wide Web is dying. Soon, we shall have only the Internet.
has always bothered me.
"Up to" is sugar-coated for "You can't expect any better than this" with a implicit translation of "It can get a whole lot worse".
Ex: If CPU X get "up to" 100% more performance than CPU Y, but in all tests but one, actually has 1% of the performance, I'd rather have CPU Y.
"Up to" means nothing to me, except as an advertisement for the competator; whichever has the least unpleasant average and worst case performance is the one I'm interested in.
34486853790
Connection too slow for X forwarding? Try "ssh -CX user@host"
Well, quite.
But you know what they say about lies, damn lies, and benchmarketing.
Or, if they don't say it, they ruddy well should.
Yeah, and notice that they say "at the same frequency", when Intel currently has a frequency advantage (not as big as P4, but then again Core 2 isn't an IPC dog like P4 was). Not that I expect any minor improvements Intel makes in the next 60 months to produce their own 50% leap in performance, this comparison still seems very suspect. As in pure marketing BS.
However AMD doesn't need to attempt become relevent again. They are currently very relevent. Did Intel become irrelevent when they were behind AMD on performance? No. In the past, AMD did lose more by not having the performance crown, and one could certainly imagine the momentum they were gaining in the K7 days fading quickly if Intel had come out with a superior chip. But today, AMD has both the marketshare and the OEM support to be merely competitive performance-wise and still be relevent. So they lose out at the top speed grades. If they can continue to match up their products to Intel's at lower speed grades, and they will, then they will continue to be a good choice for many people, and will definitely still be relevent.
The enemies of Democracy are
When the fastest Barcelona is ~2.5GHz and Clovertown is 3.0GHz, comparisons at the same frequency are pointless. What matters in reality is performance at the same price or performance at the same power or highest available performance at any price.
Chipset to cpu and cpu to cpu link with intel you have to use the chip set for one cpu to talk to another one.
Also If amd where to copy intel and put 2 dies on the same cpu they will have a better link for them that will not eat up chipset to cpu bandwith.
Over the past week we have heard about Intel's dominance and flashy new products, AMD's disastrous quarter, and now AMD's supposedly dominant new offering.
I read tech news daily and am getting sick of the media wars... It is no wonder casual users get fatigued trying to keep up. Casual opinions depend on which day (or week or month) a person chooses to research product offerings. It is no wonder I am always hitting a brick wall when trying to get my users to educate themselves so they can get more out of their tech. They don't know what to make of all the posturing.
This is not a function of the tech world developing *that* quickly. It is a result of the major players trying to out-strategize each other. I don't want to see anymore benchmarks (or hear about anymore promised software) until I am standing in front of a demo machine that is running the tech.
Guess I woke up on the wrong side of the bed.
Regards.
Obviously that was supposed to say "in the next 6 months", no "60". I'm betting Intel can produce at least a 50% speed improvement in 5 years. :P
The enemies of Democracy are
I can't wait to run Microsoft Word on these babies!
Note that AMD's claim is to be faster "at the same clock". When the P4 was pushing clock speeds into oblivion, AMD stressed the point that clock speed is irrelevant -- what really counts is how fast the system runs your software. How you get there is quite beside the point. How odd that AMD is now using clock speed as a key indicator.
Intel is already shipping 3GHz Clovertowns, and the article states that AMD has not released the Barcelona clock targets. It they ship substantially below 3GHz (2.4?), then Barcelona will probably still win on FP benchmarks (barely) but lose on everything else. This suggests it will be more competitive, but not compelling.
Historically, company come out with something 'unexpected and amazing' after a really disasterious quarter.
I would be prudent.
The Kruger Dunning explains most post on
I think one of the major reasons why AMD did so poorly last quarter was its silly marketing campaign. Towering signs on billboards and large airport ads tout AMD as "smarter choice", since it uses less power.
Marketing a chip as using less power is the same as having Toyota make an exclusive advertising campaign toward wheel-chair bound people: the group you're targeting has few people in it and they're going to research any product they buy. The server market is important, but when I buy my shiny new server, power consumption isn't my first consideration, nor is that the only thing AMD offers.
With this announcement, I'm hoping AMD starts a new slogan touting, say speed. That's what I buy a processor for primarily. AMD's always been fast for the cost and it's high time they market themselves as being faster and better rather than being "as good as" Intel. My new pick for a marketing slogan? "Upgrade to AMD" AMD should position its chips to be slightly more expensive at every pricing tier, but in doing so, blow them away in performance. (In the present economy, businesses have money and will gladly spend more money on products they feel are superior. Ford spends more money on marketing than BMW (but which would you rather own?). AMD should be trying to make Intel look like Ford, rather than being the "Ford alternative".)
AMD is marketing to a minor concern of a niche audience, while they ought to be using their superior performance (at a given price point) to sell hardware. Would you rather be a "power saver" or "upgrade to AMD".
-- Political fascism requires a Fuhrer.
Note that AMD has 3GHz Opterons out now. The frequency advantage at the moment is slim to non-existant in shipped CPUs.
In any case, what matters is what Barcelona will ship at, which has not yet been specified. In any case, if Barcelona lives up to AMD's stated expectations on performance, it will be a killer CPU. Your statement about Intel's potential improvement leaps are spot on, and fall into Inforworld's Tom Yager's statements about Intel which are essentially phrased as "Core 2 is Intel's last hurrah". Why? Because Intel is essentially running on 10 year old technology and is rushing to catch up, despite some of the nifty architectural things they did recently to speed up C2D (integrated L2 cache for example).
I also believe that Intel is now following AMD's lead by leaving extra headroom for those that prefer to OC their CPUs and concentrating more on TDP and stability. I've noticed that Intel's chips since P4 are certainly more stable, while my rather severely OC'd AMD CPU occassionally (twice this year) shuts down, most likely due to heat or a RAM instability (since the shutdowns happen during low usage periods at night, I'll bet the 20% OC'd RAM is probably to blame).
Basically, right now Intel owns the crown, but they own it while comparing to AMD's last gen CPUs which are 3+ years old.
The cesspool just got a check and balance.
Indeed. I work in a law office as a graphic designer/web designer/video editor. That's what I do all day (when I'm not reading slashdot).
2 of our attorneys just got quad-core Mac Pros with Studio displays. For writing documents on. Maybe the occasional slide show. I'm stuck on this 3-year-old Dell with dual CRT monitors. Old ones.
Sorry, just had to bitch a little. Your comment is more real-world than you may have realized.
Sony ha
I'll do you 50% faster and 20% harder than your date last week, and promise not to cost you more.
But marry me soon baby, I need the money
SSE4? Please, don't get distracted over little things like whether or not I can cook!
Note that AMD's claim is to be faster "at the same clock". When the P4 was pushing clock speeds into oblivion, AMD stressed the point that clock speed is irrelevant -- what really counts is how fast the system runs your software. How you get there is quite beside the point. How odd that AMD is now using clock speed as a key indicator.
Ummm... they're not. If they were using clock speed as a metric, they would be saying "Look! We're running at 3.5GHz and Intel is only running at 3GHz!" while completely ignoring the actual performance -- exactly what Intel did all those years. They are instead talking about performance-per-clock-cycle, which (according to this) means that a 2.66GHz AMD chip would be considerably faster than a 3GHz Intel chip. We can expect them to continue touting the overall performance rather than raw clock speed, since they look better from a performance standpoint and worse from a raw clock speed standpoint.
How is that different than what they've been saying all along?
ZFS: because love is never having to say fsck
What's really relevant to me is the performance per dollar ... not just dollar of CPU cost, but also dollar of whole system cost (including software, if that goes above zero), and dollar of energy cost (including the cost of shoving waste energy out the back door in seasons I does me no good to keep it indoors).
now we need to go OSS in diesel cars
It would be more relevant to know how does it perform real life tasks, eg kernel compilation time comparison...
Actually if Intel is running on 10 year old chip design technology (I presume this is primarily a remark about the FSB) then this suggests they have the potential to *radically* increase performance over AMD. If Intel's process advantages and chip design teams can actually gain a performance advantage while using 10 year old technology they can sit back and pick the low hanging fruit (changing to modern methods) and gain huge performance boosts while AMD has to do truly innovative things to gain any performance increases.
Frankly this sounds more like fanboi talk than a serious analysis. If your goal is to diss Intel and give AMD props then saying they are using 10 year old technology makes sense. If you are actually trying to argue that AMD's future is much brighter than Intel's it's totally non-sensical. If Intel can gain huge performance benefits just copying stuff AMD is doing now while AMD has to make huge advances just to stay competitive I know who I would put my money on.
If you liked this thought maybe you would find my blog nice too:
This post is very confused.
First of all you can't put two dies on the same cpu, or at least it would be a horribly bad idea. You can put 2 cpu's on a die. Now I thought AMD already did this but they could just package several chips together and I'm feeling too lazy to look it up.
Anyway, yes for Intel chips they must communicate over the FSB. However, as I've recently been finding out they don't do that much communicating. For instance most cache state info is generated just by listening on the FSB. Though sometimes one CPU needs to invalidate a read.
However, not having an FSB AMD's chips don't have a set total system bandwidth they 'use up.' Each chip has it's own memory controller and HT lanes. Perhaps putting the chips in the same package will allow AMD to speed up hypertransport or indirect memory lookup but since AMD doesn't use just an FSB it seems they actually have less to gain than intel by putting many cpus on one chip.
If you liked this thought maybe you would find my blog nice too:
It's not purely about FSB, although that is a major short coming for multi-core designs. Intel won't radically improve their performance by going with a hypertransport design, but they'll improve their scalability (the quads are already beyond the maximum FSB - see the Xeon benchmarks for Mac Pros that's available). The 10 year old design refers to their falling back to the P-III core and then adding in a bunch of cache logic, fab processes and other more minor tweaks to improve that core's performance to what it is today in the C2D. The next major improvement for Intel's performance scalability is with Nehamal (I think it is) when they're supposed to finally implement a HTT type solution for memory communication, although I was unclear whether it also addresses multi-CPU/core communication. I would think it would, as anything less would be moronic.
AMD has other items up its sleeve. Not all are pure performance improvements, but interesting things like hardware virtualization. This will become more improtant in the future, and may even filter down into the consumer space. Wouldn't it be great for every app to essentially have its own virtual world? That would indicate a much lower potential for harm inherent to the OS architecture. (Note, odds are MS won't go this way because it breaks forced dependencies on the OS, but it certainly opens up new worlds for others)
Intel can't copy AMD's advances. They'd have to license some tech to do so, and that's not very likely (I believe HTT for instance belongs to AMD). Intel's current big performance leap occurred because they dumped P4 and followed AMD's approach. In any case, Intel did have a lot of smart people, but with the 10,000 personnel layoffs, how many of those smart peopel won't look for better jobs elsewhere? (Layoffs don't exactly inspire confidence for a company's future) Of course, AMD announced they may have to do the same after Q1's results.
The cesspool just got a check and balance.
Well, I highly doubt that AMD's integrated memory controller's ability to hide the latency can scale any faster than Intel's cache tech. After all they basically control latency in the same manner, smartly guessing what memory the chip is going to need and making sure they keep the cache filled with that data. Of course the on board memory controller means that the latency starts at a lower place in the first place. Though it could be that the extra latency of going across the northbridge gets worse as clock speeds increase but I kinda expect it will remain proportional to bus speed.
The real issue in the next generation is simply the raw bandwidth. If AMD keeps moving in the direction it has it's next systems will be NUMA architectures like the Quad FX. This means that each chip (maybe next gen it will be core) has a constant memory bandwidth while Intel has to divide one FSB by the total number of chips in the system. Also AMD no doubt will have copied some of Intel's smart cache techniques.
Still if Intel can just keep parity (and they have the advantage of just now moving to 45nm) for another year or so (and not fuck up CSI again) that's pretty much good enough. At that point Intel no loner suffers a fundamental architectural deficit compared to AMD and their superior process technology gives them a serious advantage.
The three two questions are these:
1) How much market share can AMD gain capitalizing on Intel's memory starvation in the next year or two? Mostly this is going to depend on how well AMD can copy Intel's fancy cache and memory conflict checking hardware. If they can do a fair job of this then their underlying architectural advantages should let them overcome Intel's process advantages and make a fair bit of headway.
Unfortunately for AMD they could end up not being able to take full advantage of their superior (large scale) architecture if OS vendors don't provide better NUMA support. Eventually this is the way both AMD and Intel are going but better support now would be a huge boon for AMD.
2) Can AMD pull another rabbit out of it's hat that will give it a fundamental *architectural* advantage before Intel comes out with an acceptable version of CSI? Simply upping the GHz on HyperTransport isn't enough. AMD and Intel face the same fundamental obstacles to making high speed serial links and this area just isn't complex enough to let them totally out do Intel on transport speed. This is what AMD needs to overcome their process disadvantages and remain a serious contender for the performance and performance/watt crown (unless Intel just gives up the former to pursue the later). Given AMD's history they very well might pull this off but it's going to be tough.
Of course there is always the chance of a total shakeup in this industry if either Intel or IBM manages to patent some amazing process trick that isn't easily copied.
If you liked this thought maybe you would find my blog nice too:
Intel has already announced major price cuts for the Fall. I can't see how this does anything other than hurt Intel and current sales. For instance, I'm sure I'm not the only one that sees that and goes... hmm, I might have considered a new Intel CPU, but the prices are dropping significantly in a couple of months. And then there's the AMD roll out at the same time, maybe I'll wait and see what that brings to the table.
It's going to be very very interesting.
The cesspool just got a check and balance.
Really I find my current PC fast enough. What I want is lower power and heat for the entire system.
Now if AMD can produce a cheap and silent system with good graphics performance I am all for it. Say something as fast as an X24400 and an Nividia 7600 GT all for about $300 then you have a winner. You will sell millions.
A quad core system? I just don't need it yet.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Intel can't copy AMD's advances. They'd have to license some tech to do so, and that's not very likely (I believe HTT for instance belongs to AMD).
Actually Hypertransport is an open standard and anyone can implement it. AMD doesn't have the clout to force proprietary standards on the market, so their only hope to have a standard adopted is to make it open and royalty free.
Which is why Intel will (probably) never implement it. They aren't interested in standards which they don't control. They already don't like the fact that AMD is cross-licensed for all x86 tech, which was part of the motivation for creating the entirely separate IA-64 ISA. When IA-64 failed and Intel was forced to implement x86-64, the only reason they used AMD's spec was because Microsoft said that they would only support one x86-64 ISA, and AMD got their first. Basically it took MS to out-monopoly Intel. So unless they are forced to use HT, they won't, and I can't see any way they could be forced. They may implement something similar -- they will have to in order to address multi-socket scalability -- but it will not be compatible.
AMD would love for Intel to copy their tech. Every time they do, it makes AMD look like the leader and Intel the follower. You could practically hear the screams of orgasmic joy from AMD when Intel announced EMT64.
The enemies of Democracy are
What I want to know is - will this new AMD chip use AM2 socket, or a new version?
urd
For some reason, I thought HTT had to be licensed. That aside, your insight as to Intel's reluctance is spot on. They've already had to eat crow thrice (I just had to use it:) this decade, with the AMD-64 implementation, the failure of Itanium, and the failure of P4. A fourth stating their entire architectural approach was incorrect might just doom them in mindshare to a complete failure.
I don't think that anyone that's truly interested in the technical aspects of CPUs thinks Intel is a leader in this decade. Since the AMD-64 and Itanium launches, that position has been firmly held by AMD which continues to extend that lead from everything I'm seeing. (For you Intel fanbois, that's even despite the C2D, which really is irrelevant in the world of servers). AMD's technical lead is pretty much cemented in the server world, and I wouldn't be surprised if Barcelona puts that market firmly out of Intel's reach. The current Opterons match Intel's best in equivalent configurations, but scale far beyond what Intel is capable of, and apparently Intel won't have an answer for at least another 12 months.
Intel has a PTP alternative to HT in the works, but it won't be ready until late 2008 (Nehalem) supposedly. By then, Barcelona will have been out for at least a year.
The cesspool just got a check and balance.