Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
Not open for further replies.
Please everyone do your search before... The Nehalem Xeon got the same QPI/memory architecture than Core i7 right ? So ...

QPI:
From a technical point of view, a QPI link is bidirectional and has two 20-bit links—one in each direction—of which 16 are reserved for data; the four others are used for error detection codes or protocol functions. This works out to a maximum of 6.4 GT/s (billion transfers per second), or a usable bandwidth of 12.8 GB/s, both read and write.

FSB:
The FSB on the most recent Intel processors operates at a maximum clock frequency of 400 MHz, and address transfers need two clock cycles (200 MT/s) whereas data transfers operate in QDR mode, with a bandwidth of 1.6 GT/s. With its 64-bit width, the FSB also has a total bandwidth of 12.8 GB/s, but it’s usable for writing or reading.

So a QPI link has a theoretical bandwidth that’s up to twice as high, provided reads and writes are well balanced. In a theoretical case consisting of reads only or writes only, the bandwidth would be identical to that of the FSB. However, you have to keep in mind that the FSB was used both for memory access and for all transfers of data to peripherals or between processors. With Nehalem, a QPI link will be exclusively dedicated to transfers of data to peripherals, with memory transfers handled by the integrated controller and inter-CPU communications in multi-socket configurations by another QPI link.



The advantage of an integrated memory controller isn’t just a matter of bandwidth. It also substantially lowers memory access latency.

More here.

Thanks for clearing that up between the hardware differences. I just didnt have the technical stand point to compare each hardware side by side as you have.

Finally a reasonable voice.
 
jjahshik32 is an evil Apple shill? Hehehe...

Nah, he's just confused by Apple's bait & switch. Hey that shiz works on some people you know... :D

I just dont understand how I'm misinforming people?? Nehalem IS faster than the previous models that they are replacing.

If I were in the market for a new Mac Pro (which I am) I would DEFINITELY get the Nehalem Models, no questions asked. Why settle for old models? Future proof by buying the new architecture now.
 
Just as an update... I didn't make a purchase today because I have some vendors competing on price and I'm awaiting qoutes. The configuration I'm looking at is the 2.26GHz with 8GB RAM (4x2GB modules) in preparation to upgrade to 12GB RAM (6x2GB modules) 'sweet spot' that BareFeats discovered.
The 64-bit Geekbench results are very interesting, but if I remember correctly, the 3.2GHz 08 did beat the 2.26GHz in other tests. That doesn't deter me. I think the 2.26GHz is more affordable (I can't even locate a 3.2GHz 08 octad), and I think it will prove faster as future apps emerge. Also, in a couple years I'm going to begin stalking for a pair of upgrade CPUs, so there's a bit more headroom the 2.26GHz can avail.
Maybe I'll list the lowest vendor qoute here? Would that violate the forum rules, I wonder? :eek:

my 2.26 outrenders a 3.2 octad on maxwell render benchmarks (benchwell in particular) 6m30s benchmark of 1620 vs 7m21s bench of 1459:D
so yes in real world RENDERING situations, which will use all the cores, the 2.26 will outperform a 3.2Ghz Harpertown Octad
 
I just dont understand how I'm misinforming people?? Nehalem IS faster than the previous models that they are replacing.

If I were in the market for a new Mac Pro (which I am) I would DEFINITELY get the Nehalem Models, no questions asked. Why settle for old models? Future proof by buying the new architecture now.


I don't think you're misinforming. I think Apple was successful in pulling a fast one over on it's customers by confusing the comparisons they chart.

Here's a just one of the machines more intelligently compared and charted that I wrote in another post:

1998 Apple Releases a 1 core G3 266MHz (AV) $2500 ... Speed increase from previous = more than 2X
2000 Apple Releases a 2 core G4 450MHz $2500 ... Speed increase from previous = more than 2X
2002 Apple Releases a 2 core G4 1.00GHz $2500 ... Speed increase from previous = more than 2X
2004 Apple Releases a 2 core G5 2.00 GHz $2500 ... Speed increase from previous = more than 2X
2006 Apple Releases a 4 core 2.66 GHz $2500 ... Speed increase from previous = more than 2X
2008 Apple Releases a 8 core 2.8 GHz $2800 ... Speed increase from previous = more than 2X

See this is about how it's supposed to go. Since about 1980 this has been pretty much the case with every vendor. Every two years we get 2X the speed for about the same price. But what about this year?

2009 Apple releases a 8 core (w/HT) 2.66 GHz $4700 ... Speed difference = between 0.9X ~ 1.68X

Yeah, between slower and only about 70% increase when heavily multitheading - but almost TWO THOUSAND dollars more???

See what I mean? What the future holds I dunno. Hopefully an Apple remedy! But this kind of jump in price for little or no benefit... Shrug! I dunno what to make of it. Other vendors have not and likely will not be doing anything like this. So it's only Apple? If so then I think it's time to get OS X running on non-Apple Hardware where we will be getting our 2X speed increase for the same $2500 to $2800 and not $4700. Either that or wait till 2010 and see what happens.​

So you can see what I and others are talking about.
 
I don't think you're misinforming. I think Apple was successful in pulling a fast one over on it's customers by confusing the comparisons they chart.

Here's a just one of the machines more intelligently compared and charted that I wrote in another post:

1998 Apple Releases a 1 core G3 266MHz (AV) $2500 ... Speed increase from previous = more than 2X
2000 Apple Releases a 2 core G4 450MHz $2500 ... Speed increase from previous = more than 2X
2002 Apple Releases a 2 core G4 1.00GHz $2500 ... Speed increase from previous = more than 2X
2004 Apple Releases a 2 core G5 2.00 GHz $2500 ... Speed increase from previous = more than 2X
2006 Apple Releases a 4 core 2.66 GHz $2500 ... Speed increase from previous = more than 2X
2008 Apple Releases a 8 core 2.8 GHz $2800 ... Speed increase from previous = more than 2X

See this is about how it's supposed to go. Since about 1980 this has been pretty much the case with every vendor. Every two years we get 2X the speed for about the same price. But what about this year?

2009 Apple releases a 8 core (w/HT) 2.66 GHz $4700 ... Speed difference = between 0.9X ~ 1.68X

Yeah, between slower and only about 70% increase when heavily multitheading - but almost TWO THOUSAND dollars more???

See what I mean? What the future holds I dunno. Hopefully an Apple remedy! But this kind of jump in price for little or no benefit... Shrug! I dunno what to make of it. Other vendors have not and likely will not be doing anything like this. So it's only Apple? If so then I think it's time to get OS X running on non-Apple Hardware where we will be getting our 2X speed increase for the same $2500 to $2800 and not $4700. Either that or wait till 2010 and see what happens.​

So you can see what I and others are talking about.

You can also make the argument that the new base model 2.26GHz Nehalem Mac Pro out renders and out scores a 3.2GHz 8 core model for cheaper.

$3299 for the 2.26GHz model and the 3.2GHz model I believe was in around $4799 a whopping $1500 difference.
 
You can also make the argument that the new base model 2.26GHz Nehalem Mac Pro out renders and out scores a 3.2GHz 8 core model for cheaper.

$3299 for the 2.26GHz model and the 3.2GHz model I believe was in around $4799 a whopping $1500 difference.

That's the confusion that Apple is trying very hard to pull off, yes. But throughout history you could always make that argument. Because they're 2x faster then the older higher speced one which carried a premium price tag now is not as fast - yet cost more. That's not how it works though and this is the change-up that Apple itself is trying to get over on us with.

Well, I for one am not that dumb. I'm not going for it. It's $2000 more for the same machine a year more into it's progression.

And actually the 2.26 isn't even close to the 3.0 let alone the the 3.2!

The new 2009 2.26 is about the same speed as the MacBook Pro 2.66. :D
https://forums.macrumors.com/posts/7270035/
It's the green bar that makes the difference - not the yellow one. The Yellow bar is if you're doing nothing on the machine yourself - it's just rendering, off by itself, in a corner somewhere.
 
That's the confusion that Apple is trying very hard to pull off, yes. But throughout history you could always make that argument. Because they're 2x faster then the older higher speced one which carried a premium price tag now is not as fast - yet cost more. That's not how it works though and this is the change-up that Apple itself is trying to get over on us with.

Well, I for one am not that dumb. I'm not going for it. It's $2000 more for the same machine a year more into it's progression.

And actually the 2.26 isn't even close to the 3.0 let alone the the 3.2!

The new 2009 2.26 is about the same speed as the MacBook Pro 2.66. :D
https://forums.macrumors.com/posts/7270035/
It's the green bar that makes the difference - not the yellow one. The Yellow bar is if you're doing nothing on the machine yourself - it's just rendering, off by itself, in a corner somewhere.

Are you sure your not that dumb? 2.26GHz XEON 8 core is the same speed as a macbook pro? Uhhh huh. :rolleyes:

Dont forget to look at the geekbench results along with the maxwell rendering benches, which both indicates the 2.26GHz model is faster than the last gen 3.2GHz model.

neh03_g64.gif


And the 2.93GHz spanking the 3.2GHz.
 
Are you sure your not that dumb? 2.26GHz XEON 8 core is the same speed as a macbook pro? Uhhh huh. :rolleyes:

Look at the chart I posted. It's right there.


Dont forget to look at the geekbench results along with the maxwell rendering benches, which both indicates the 2.26GHz model is faster than the last gen 3.2GHz model.

GeekBench is extremely unreliable. I wouldn't trust anything it said. My machine can score up to 600 points different just depending on air temperature and the phase of the moon. All conditions otherwise the same. GeekBench should be banished into the darkest recesses of the universe. Hey, I'm being serious! :)
 
I just find it funny that the 08 mac pro owners always dismiss the maxwell renderings and the geekbench overall benches.

I understand always looking at the cinebench 10 which does make the 8 cores look good (probably due to the nehalem not fully optimized in those tests).
 
Look at the chart I posted. It's right there.




GeekBench is extremely unreliable. I wouldn't trust anything it said. My machine can score up to 600 points different just depending on air temperature and the phase of the moon. All conditions otherwise the same. GeekBench should be banished into the darkest recesses of the universe. Hey, I'm being serious! :)

Sure you can keep believing that its unreliable. I do understand that it will make your old hardware still look good. :)
 
Using bad numbers to prove a point is kind of disingenuous. I guess you didn't know how bad GeekBench was (is) so i won't hold it against you. Here's a chart with more accurate results:

Cinebench10_Numbers.jpg
 
Using bad numbers to prove a point is kind of disingenuous. I guess you didn't know how bad GeekBench was (is) so i won't hold it against you. Here's a chart with more accurate results:

And why is cinebench 10 the most outdated crap more reliable??

Funny how most tech sites always use geekbench more so than cinebench.
 
Because it shows single threaded AND multithreaded performance.

I'm not an 08 owner. I don't dismiss Maxwell render marks - but for maxwell be warned that it only measures the rendering speed of exceptionally well written multithreaded code. No single thread performance right? GeekBench I dismiss. As you can see here: http://browse.geekbench.ca/user/Tesselator Lightwave/geekbench2 it's just not even close to reliable.



.
 
I'm not an 08 owner. I don't dismiss Maxwell render marks - but for maxwell be warned that it only measures the rendering speed of exceptionally well threaded code. No single thread performance right? GeekBench I dismiss. As you can see here: http://browse.geekbench.ca/user/Tesselator Lightwave/geekbench2 it's just not even close to reliable.

Also there is an error to your graph. Cinebench 10 measures your CPU along with open GL of your graphics cards. So if say the 3.2GHz model had a superior card (nvidia 8800GT) as to the 2.26GHz model (GT120) in these tests it can cause a huge difference in these scores. Also it isnt as memory intensive as what geekbench offers.

Geekbench purely stresses the CPU + Memory for its overall tests which is by far more accurate assessment. That bootleg graph you made means nothing.

Also for that geekbench link you provided, nice job trying to make a 1st gen mac pro look close in points to the 8 core machines.

If I remember correctly and I know that I'm right, there were NO 8 core 2.66GHz model of the MacPro1,1 models (first gen). Either you upgraded it by dropping in 4 more cores which would make sense for the geekbench to score that high, a Hackint0sh or you just made that $*it up.


Also now that I think about it, the jump from clovertown to harpertown was negligible at all. At least the Nehalem even though priced a bit more is still much faster than the previous models it replaced. 2.26GHz being faster than the previous 3.2GHz and the 2.93GHz being 2.0x faster than the 3.2GHz which is clocked more.
 
Also there is an error to your graph. Cinebench 10 measures your CPU along with open GL of your graphics cards. So if say the 3.2GHz model had a superior card (nvidia 8800GT) as to the 2.26GHz model (GT120) in these tests it can cause a huge difference in these scores. Also it isnt as memory intensive as what geekbench offers.

No, wrong.


.
 
Are you sure your not that dumb? 2.26GHz XEON 8 core is the same speed as a macbook pro? Uhhh huh. :rolleyes:

Dont forget to look at the geekbench results along with the maxwell rendering benches, which both indicates the 2.26GHz model is faster than the last gen 3.2GHz model.

neh03_g64.gif


And the 2.93GHz spanking the 3.2GHz.

Hahahahaha according this my ancient 2.66 from 2006 beats or is about the same as the new 2.66 and last year's 2.8. Bahahahahaaa!!! I wish!

http://browse.geekbench.ca/geekbench2/view/115008 <-- Me! 2006 2.66 hehehe

And this guy with an identical machine as mine snuffs BOTH of them and competes with the new 2009 2.93GHz box... Tee Hee....
http://browse.geekbench.ca/geekbench2/view/113749
 
No, wrong.

"CINEBENCH 10 is a benchmarking tool based on the powerful 3D software CINEMA 4D. Test results conducted using CINEBENCH 10 carry significant weight when analyzing a computer's performance in everyday use. Especially a system's CPU and the OpenGL capabilities of its graphics card are put through their paces (even multiprocessor systems with up to 16 dedicated CPUs or processor cores). During the testing procedure, all relevant data is ascertained with which the performance of different computers can subsequently be compared, regardless of operating system. CINEBENCH 10 is available for Windows 32-bit and 64-bit operating systems and as a "Universal Binary" version for Apple Macintosh computers with PowerPC or Intel Core Duo processors. The database functionality and simple export allow for easy statistical data collection."

http://www.benchmarkhq.ru/english.html?/be_video.html
 
I really think I found why the 08 mac pro owners are feeling hostility about the Nehalems as to their harpertowns. They all know that its a good amount faster but deep down they do want to update to the new precious but its too expensive to do so.

So what would they do?

First step: Seeing the benchmarks of the Nehalems and not yet knowing of the real world speeds and especially the release of Snow Leopard of the Nehalems compared to the 08 models. And also seeing that the 2.26GHz Nehalem is faster than the 3.2GHz 08 model (Denial).

Second step: Would be looking at the prices of the Nehalem Mac Pros (anger).

Third step: Would be trying to convince themselves if the price difference they'd pay after selling the 08 model is worth it (bargaining).

Fourth step: Finding out when Snow Leopard hits that Nehalem is infact faster and benefits greatly (depression).

Fifth step: Putting up the 08 Mac Pro on eBay to get half the money for the new Nehalem Mac Pros, because its undeniably faster and better (Acceptance).

Sixth step: Seeing the new Westmere Mac Pros in 2010, repeat steps 1-5.
 
i DONT want to update because 08 octo has enough power to suit my needs,
and i dont care if it is cheaper now, because i have used it successfully for 1 year and it has in fact earned me more than the 500$ difference is :)

if i would be buying a computer NOW i would think twice before buying either model due to price/performance ratio.
15% peformance difference isnt even enough to skip a generation!!

hostility is not about Nehalem computers, it is about you, because you sound like a cheap commercial.


you dont even know what QPI is, nor what FSB was last 20 years, because you obviously have no idea how computer is structured

look ill make a comparison as if i was you when 08 came out
"just wait till the new OS SL comes out, youll see what 1600 fsb will do when it will be unleashed compared to 1333mhz fsb! FTW!111"

look, some of us actually USE computer and unlike you would have a hard time being without one for such period of time, as you clearly can afford that, its a question if you really need a mac pro, and if you dont, dont give advice to people who do.

you, regarding your reasonable voice, and completely dismantling your "SL will boost nehalem" statements.
Finally a reasonable voice.
Fourth step: Finding out when Snow Leopard hits that Nehalem is infact faster and benefits greatly (depression).
I don't think that snowleo will make advantage of 0,00001% of Mac Pro '09 users. It will loose some fat but it will be for all Macintosh. Maybe in 3 or 5 years we'll see the results of GC and OpenCL. Yes I'm talking to you Adobe. :D
 
OK so you have decided to spread this crappy arguement across the forums.

Give it up already.
 
Status
Not open for further replies.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.