Slashdot Mirror
45nm Opteron Performance, Power Efficiency Tested
An anonymous reader writes "Now that Intel has unleashed its next-generation Core i7 processors, all eyes are turned to AMD and its incoming wave of 45nm CPUs. To get a feel for AMD's future competitiveness, The Tech Report has taken a pair of 2.7GHz 45nm Opterons (with 75W power envelopes) and put them through the paces against Intel Xeons and older, 65nm Opterons in an extensive suite of performance and power efficiency tests — from Cinema 4D and SPECjbb to computational fluid dynamics and a custom XML handling benchmark. The verdict: AMD's new 45nm quad-core design is a notable improvement over the 65nm iteration, and it proves to be a remarkably power-efficient competitor to Intel's Xeons. However, 45nm AMD chips likely don't have what it takes to best Intel's Core i7 and future Nehalem-based Xeons."
...but have since really lost momentum and competitiveness. They truly awakened the sleeping giant when they were kicking Intel's ass a few years ago.
I'm not wise to all the marketing names that chip vendors use these days: will this 'Opteron' chip be priced competitively as an alternative to the Core i7, or will it just be an expensive server processor? I know that having the fastest top-end chip has a halo effect on the rest of the range but with Intel's mid-range processors being good and cheap, that's where AMD most needs to make an improvement.
-- Ed Avis ed@membled.com
--- Or at least I can't reach it. I guess their servers doesn't feature any of the cpu's reviewed ;)
Mirror anyone?
It's too bad AMD rested on their laurels after destroying Intel's itanium. Too bad they destroyed Itanium by the simple expedient of backwards compatibility, as opposed to superior architecture.
Because that's what's happened to them now--Intel has them dead beat on core architecture, and no amount of size reduction or megahertzing can save them now.
Sure hope they're hard at work on some kickass new architecture in their basement, because we desperately need Intel to have a strong competitor.
expandfairuse.org
I find it disappointing that the test of the supposed server-oriented processors does not include web server tests - after all it's probably the largest market for such processors.
I mean, does anyone really care about Folding@Home number these processors can crunch? Or "VRAD map build benchmark"? WTF?
Again, I wonder if the benchmarks used AMD optimized code (they have to use the proper GCC backend). It seems that most of the time, the benchmarks for non-Intel processors are based on Intel optimized code. I have never seen mentionned in the benchmarks if the tools were using the best machine code for the targetted processor... yeah... that smells bad.
AMD didn't really destroy Itanium and then rest on their laurels. Although you have to give them some credit for coming up with reasonably good chips that the market wanted, it was more that Itanium was the reason AMD was competitive with Intel in the x86 space for a few years in the first place.
Intel has orders of magnitude more R&D budget and especially capital for fab construction than AMD does. So AMD is perpetually at least a half-generation behind Intel on the tech curve: they keep coming up with chips that could beat Intel... if they had come out a year ago. Now when Intel effectively skips a generation, as they did when they sunk all their resources into Itanium and mostly ignored x86 for a year or two, this is enough to give AMD the lead. But once Intel shifted fully back into x86, they crushed AMD again.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
..and benchmarks. If AMD apply the name "Silverstone" to something that runs like a dog, I'll not be happy. Of course, if it were water cooled, it would probably be quite apt.
Right now, the core should probably be called "Monaco" since the chances of them overtaking the competition are pretty slim unless they get really lucky.
They might and as far as bang for the buck AMD is still there. The main thing that burns me with AMD is the stunt they pulled with the short-lived 939 socket. AMD left those people out to hang, high and dry.
Shai Schticks:"You don't make peace with friends, you make peace with enemies"
i've often wondered if microsoft has a deal with intel to slow amd processors.
Yes, sort of.
It's called NUMA - Non Uniform Memory Architecture.
Up until recently Intel platforms had the memory controlled by the northbridge, with all CPUs and all cores having the same access to the memory.
Newest Intel platform and all 64bits AMD had the memory controller on the processor package. In a multi-socket configuration, each processor controls it's own chunk of memory, so for some range, the access will be faster because the processor is directly accessing it, and for other the latency will be increased because the processor has to ask its neighbour over HyperTransport / QuickPath.
To be able to function in a such configuration, an OS should pay some attention when scheduling process and threads to cores : it should be best that all threads from some process are all scheduled to cores having all direct access to the resources used by said process. (While at the same time scheduling two threads at a physical core and it's corresponding hyperthreading virtual core if there's a another physical core sitting idle)
Windows has always deeply sucked at this. Opensource OS, on the other hand, have much more work applied to them for that. (That's why they are much more popular on super computers).
This also introduces technical difficulties (like keeping the cache coherent). That's also why heavily multi-socketed (4 and up) motherboard won't be coming during the first year of Core i7's life. They probably have to fix all the fine details before that. As usual expect a change in socket format and a new iteration of Core i7 not quite exactly compatible with the previous one.
On AMD's side, currently sold Opteron are already adapted for 4 and more sockets configuration. (As explained by other /.ers, the 8000 series has a coherency protocol running on 3 HT interconnects, which should be enough to help on 4 and more sockets).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Hilarious no! Profoundly disturbing yes!
Andy Warhol got it right / Everybody gets the limelight
Andy Warhol got it wrong / Fifteen minutes is too long.
On our benchmarks we can't get 4 cores' worth of performance out of an Intel CPU, but we can get nearly 8 cores' worth out of AMD. AMD's memory bus architecture is simply better.
Amd has lower cost chipset and good on board video.
Also you can use any chipset in a 2P and 4P+ system unlike intel where you are stuck with nvidia chipsets. Back when the intel mac pro first came out at the time amd 2p systems had more pci-e lanes and a better I/O setup for them. Also intel does not have low to mid range corei7 chipsets / MB they only have high - mid and up.
The new phenom 2 will work with to days amd chipsets / boards unlike intel. Core i7 is fast but $250 - $300 + for a MB that is over kill for some people we all don't need a high end 3-4 way sli / crossfire fire board + $300 for the cpu.
Amd has good $80 to $140 boards some even have on board video with side port ram. High end dual full x16 or more boards go for $150 - $200.
Intel needs to have a low end - mid range corei7 cpu + lower end chipset / board with the lines for 1 full x16 port can be split to x8 x8 and left over ones for the other ports maybe a x4 and few x1 lanes for on board stuff / 2-3 pci-e x1 slots.
the codenames AMD is using for the opteron line are all cities that hosts, or used to host, formula 1 grand prixes.
maranelo, sao paulo, magny cours...
nice to see my own city (sao paulo) mentioned in the road map. gotta start saving $$$ to buy me one of those "sao paulo" chips when they get released.
What ? Me, worry ?
AMD was kicking Intel's ass only during the time period where Intel was making what turned out to be a losing gamble on the end of x86, putting most of their resources into Itanium instead. They ran their x86 line during that period largely on autopilot, which AMD took advantage of to catch up and surpass Intel in the x86 space, betting (correctly) that x86 would remain the platform of choice, and Itanium would go nowhere.
When Intel eventually realized this was the case, and shifted most of their R&D back to x86, the result was the Core architecture, and we were back to the usual scenario of Intel dominating the market.
Basically I don't think AMD has the resources to take on Intel head-on, unless Intel is putting a significant proportion of their resources elsewhere, as was the case in the Itanium years.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Unfortunately, Ballmer's current cunning plan, Vista, hasn't worked out :-) The next cunning plan, V7, probably won't do worse, and maybe will do better, but by around XP time the CPU people had pretty much outrun MS, and the RAM people have also gotten ahead of it even with DDR or certainly by DDR2. One thing I'm missing by not doing Vista is the cache-on-flash feature, which is too bad; it'd let my disks spin down most of the time.
The other traditional driver for the CPU makers has been gamers, who can use all the power they're given, but the last few rounds that's mostly been in the graphics board camp, not the CPU camp; we'll see if the CPU side starts integrating more graphics functions or if the game companies increasingly leave PCs for consoles. (I'm not the target audience for that; I'll probably continue using mobo-builtin graphic unless my next LCD flatscreen needs a different interface, in which case I'll buy a =$49 board, and Nethack never did need all that DirectX anyway :-)
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
My impression was that AMD's lead over Intel only really lasted until the introduction of the Core architecture around 2006. Some of AMD's offerings were arguably competitive with the Core, but it seemed to start taking back market share, whereas before that AMD was generally always the chip of choice on a price/performance basis.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
I myself unfortunately bought a Intel Core2Quad a little over a year ago, and regret big time I did not wait for the AMD quad instead. The reason is that Intel only allow real Virtualization in crippled protected mode!!! That essentially destroyed my objective for buying a Quad core... For virtualization, AMD has a much better design that avoids these problems. If you intend to run separate systems such as Windows and Linux sharing resources such as the network card and so on without loosing power, you better choose AMD over Intel! I know - it is supposedly a solution in Linux kernel 2.6.27 to bypass some of the inherit problems with the Intel chips with virtualization, but that is yet not supported by many distributions such as CentOS, which still is my OS of choice...
It appear many people have ran into similar problems when incorrectly believing Intel would be best:
http://forum.pfsense.org/index.php?topic=3294.msg44574
I certainly trade a little speed for real virtualization any day!