• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Nocona Reviews Missing?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
What else was new to make it worth reviewing?

Besides it's ability to run x86-64?
Well, if you include Tumwater (chipset)
1. DDR2
2. Demand-based speed-stepping (as on their mobile uP)
3. 800MHz FSB
4. 3.6 GHz is supposed to be out
5. HT

However, if it can't run 64bit, it's just an overpriced normal Xeon with a few nifty additions...
 
Oh ok, I thought they were skipping over 533 and 800 and going straight to 1000 whatever it was. Apparently it was just 533. Either way, the bus increase will certainly help Xeon SMP setups that were struggling with bandwidth. The link doesn't work.
 
Originally posted by: Viditor
However, if it can't run 64bit, it's just an overpriced normal Xeon with a few nifty additions...
To me (and to most users right now, 64-bit is useless). I couldn't care less if Intel skipped 64-bit altogether for now. But the real thing that is great is that all of this comes without a price increase. Thus they aren't overpriced (compared to the other Xeons).
Originally posted by: PariahThe link doesn't work.
That is odd, I tried it with two different computers. Basically go to Dell, click on small business, click on workstations, click on Choose Dell Precision, see the new Precision Workstation 670 with 800 MHz fsb. Edit: Now they have also added a Precision Workstation 470. $300 cheaper. 😉
 
I'm definately interested in the 64 bit capabilities of Opteron/Xeon.
Redhat EL supports AMD64, as does IBM with db2, that's enough to make me and a few other fellas at work interested.
HPaq's DL585 is a very interesting box, most likely better performance than a Sun v440(which would be the logical competitor), and cheaper than Itanium.
If Intel's AMD64 implementation is shaky, that DL585 will keep looking like the best buy to me.

Of course I guess not too many AT'ers care about RH EL/db2 🙂
 
Intel finally has a press release. That certainly isn't a review, but finally almost one benchmark is available. Nearly a 30% boost in SPECfp_rate_base2000 when comparing the previous dual 3.2 GHz/533 MHz/1 MB chips to the dual 3.6 GHz/800 MHz/1 MB chips. Processor speed should account for a 10% boost, as should fsb increases. Where the other 10% came from is anyone's guess (maybe fsb gave more than 10%?).
 
Common server applications are almost EXCLUSIVELY ALU operations, memory transfer operations, and comparisons/conditional branches.

I am expecting a HUGE difference on web and SQL benchmarks.
 
The problem is that Nocona is basically Prescott, meaning it inherits Prescott's uber-long pipeline - not so good for integer performance. Intel does have a few improvements though to minimize the impact of that deep pipe: new ALUs, better branch prediction and of course larger caches. The increase in FSB will help server applications, we will take a look at server performance as soon as server (read: not workstation) platforms are available.

Take care,
Anand
 
How about Linux benchies eh Boss? 😉
AMD64 compatible Linux distros have been available for some time now 🙂
 
Will you compare the old 7505 based Xeons to the new 7525 platform ? That way we can see if Intel has made an improvement in their new platform.

Regards,
Jose
 
Thinking about this some more, the default operand size ( sizeof(int) ) is 32-bit, even when in 64-bit long mode, so web servers probably won't be affected. DB will still be hurting though.
 
So according to the article, the new Xeons are useless as far as 64bit addressing is concerned.. 🙁
I saw some articles/post that said Prescott's will be the first dual-core processors from Intel, if both the new Xeons & Prescott's are based on the same design, then it's time to flush Intel down the crapper ! . Who want's a software 64bit emulation for a processor ?? 😱

Regards,
Jose

BTW I hope Anand's review brings this info to light for everyone.
 
First real reveiw that I've come across (Game PC: 3.4 GHz Xeon vs Opteron 250). Yes there were the odd benchmark or two floating around, but here is finally a full review covering many different aspects.

Again, we have to say that this is a smidge unfair as we?re comparing AMD?s top of the line chip against the second-to-the-top from Intel. We tried our hardest to get our mitts on the 3.6 GHz ?Nocona? Xeon processors, but they?re near impossible to get in the US at this time. Our estimation is that within the next two to four weeks, we should see these chips become available, but as of now, the 3.4 GHz Xeon is the fastest which is actually available.
While Intel may be the choice for content creation types, the Opteron still wallops the Xeon in terms of memory bandwidth/latency, number crunching, webserver performance, and media encoding.
we were able to load our latest build of Windows XP 64-bit Edition on both systems and were able to run some quick 64-bit SiSoft Sandra benchmarks
 
Originally posted by: dullard
we were able to load our latest build of Windows XP 64-bit Edition on both systems and were able to run some quick 64-bit SiSoft Sandra benchmarks
64-bit SiSoft Sandra do not use the extended registers of x86-64, so "64-bit" or not, it only shows off 32-bit performance. I have this slight suspicion that this is so because Intel wants it that way.
 
So, it's been two weeks and still no 64-bit benchmarks.

Remember when the Opteron came out? People were benching that left and right, comparing to the top P4s and Xeons to try to get an idea of what the A64 would do on release. Nocona's 64-bit performance will be directly indicative of any consumer-grade IA-32e CPU, so why aren't any reviewers as interested in Intel's first 64-bit offering as they were with AMD? Even if it's just some Linux server benches I'd love to see a 64-bit comparison.
 
Originally posted by: Alkaline5
So, it's been two weeks and still no 64-bit benchmarks.

Remember when the Opteron came out? People were benching that left and right, comparing to the top P4s and Xeons to try to get an idea of what the A64 would do on release. Nocona's 64-bit performance will be directly indicative of any consumer-grade IA-32e CPU, so why aren't any reviewers as interested in Intel's first 64-bit offering as they were with AMD? Even if it's just some Linux server benches I'd love to see a 64-bit comparison.
There are a few reasons:
1) Official Win 64 Beta available in another ~2 weeks, so why not wait.
2) Near complete lack of 64-bit programs. With almost nothing to test, how are they going to test it?
3) Lack of 64-bit drivers. The Xeon release came with many changes, all of which need the proper drivers. Plus for things with drivers, they still aren't at a mature enough level. Thus many tests with 64-bit is far slower than 32-bit. This will change somewhat given time.
4) Once #1-#3 are taken care of there still is a lack of benefit from 64-bit. Most programs don't need it and won't benefit from it. So it makes it silly to do tons of tests showing that yep nothing happens by using 64-bit.

Yes there are a few rare programs out there that are 64-bit, and do benefit from 64-bit (64+ bit encryption for example). Some reviews do things like test Fhourstones 2.0 connect-4. WTF is Fhourstones? They had to resort to specialty programs that none of their readers care about. Specialty programs specifically written to take advantage of 64-bit features that nearly all programs don't need.

It just is silly to do a full review until #1 is ready since the deadline is so close. Look back at all the original Opteron tests, almost exclusively it showed worse or even performance with 64-bit due to the same reasons listed above (with the exception of the few specialty programs mentioned above).
 
Back
Top