Help me understand this please.

sindows

Golden Member
Dec 11, 2005
1,193
0
0
Well considering no there is no support for Conroe in any motherboard, I wouldn't be surprised if Conreo is performing like its suppose to. However once people start making bios which support Conroe, I suspect that we can see speed increases across the board.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
He has a good point on the cache and has some real value when you look at superpi as well from 1m to 32m....However he is as guilty as most of the test....To little testing has actually been done and previews with unsupported boards is really just not a fair basis to come up with conclusions, as of yet...

Remember Conroe did well in gaming apps in ATs benches as well...NOt surprising games seem to show some of the biggest gains with larger l2 cache....Intels gimmicks of adding more cache to its early EE models as well as prescott chips in its waning years to try to close gaps with A64....Dont believe me then look at reviews showing venice A64s versus winchesters....

4mb of unified cache is quite large....It could have some large gains...

Now programs like video encoding as described in a past article by AT does not take advanatge of L2 cache and that would be a nice test to really see...
 

BrownTown

Diamond Member
Dec 1, 2005
5,314
1
0
there are a million benchmarks is you look for them, all the synthetic benchmarks have been done (Sandra etc..), all the simulation type benchmarks have been done (3DMark, PCMark, Cinebench, ScienceMark, SuperPi 1M-32M, many others), and even real appl (games, winRAR, encoding). So, the vast majority of generally accepted benchmarks have been done of Conroe, however some will doubt the Intel provided benchmarks thinking maybe they are too high, and others will question the XS benchmarks saying they too low because only 512 RAM and beta board/BIOS. However, of all the benchmarks done this person has choosen the ones that show Conroe in the worest light. Of all the benchmarks Conroe wins 90% of the time (and remember this is the 316$ Conroe, not the 1200$ one). It is true that the number of sources is limited, but it is VERY deceptive to post the 1 benchmark that the A64 wins the most. Not like I can really understand why all this matters, you can't buy a Conroe, so whether or not its gonna blow everyones socks off makes no difference until you can actually buy one.

Also, the crazy claims about 4G AMD64s etc are just comming from nowhere. He is getting mad that people with REAL benchmarks of a REAL processesor are posting about how good it is, then right after that posting that a 4G A64 with Krazy Kool Z-Ram is gonna be way better when he has absolutely no basis for that claim.
 

Greenman

Lifer
Oct 15, 1999
22,196
6,420
136
It never occured to me that the cpu wasn't supported by the board they were testing on. Is the part about not allowing Anand to look at any of the settings on Intels test setup true? I don't remember much about the preview (need to read it again), but if thats true, Anand should have refused to print the results. If it turns out that Intel crippled the A64, Anand will be in a very poor position indeed.

You posted as I was typing BrownTown. I'm not concerned about which is faster, but I do wonder about how those speeds are tested. Intel not allowing Anand to inspect the system means they had something to hide. That something might have been information they didn't want made public for valid reasons, but it would have caused me to refuse to go public with the benchmarks produced.

Anyway, thanks for the input. I hope Conroe does well.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: Greenman
It never occured to me that the cpu wasn't supported by the board they were testing on. Is the part about not allowing Anand to look at any of the settings on Intels test setup true? I don't remember much about the preview (need to read it again), but if thats true, Anand should have refused to print the results. If it turns out that Intel crippled the A64, Anand will be in a very poor position indeed.



Because of those last statements many have decided to take that preview with a grain of salt....

Wait for real reviews is all we have....

We dont know if INtel did or did not...Good argument for Intel is if they did mess with the cpu they would only look bad once the truth got out....

That being said a few tweaks one way or another to get a few extra percentage points may have happened....I dont think they would lie about 20%...

especially since their lieing crooked ways already has their arses in court...

 

Markbnj

Elite Member <br>Moderator Emeritus
Moderator
Sep 16, 2005
15,682
14
81
www.markbetz.net
He does have a good point about the working set. More tests would be needed. If the performance consistently falls off above 4m then I don't see how a motherboard update would improve things. I'll come to some conclusions when I see more real world results.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,222
16,101
136
Originally posted by: Seer
Why would this guy know more than the rest of us about what Anand did with the box?

Anand couldn't do crap with the box, it was provided by Intel. As Duvie said, WAIT UNTIL ALL THE REAL REVIEWS COME IN ON HARDWARE YOU CAN BUY !
 

Elfear

Diamond Member
May 30, 2004
7,165
824
126
Originally posted by: Greenman
It never occured to me that the cpu wasn't supported by the board they were testing on. Is the part about not allowing Anand to look at any of the settings on Intels test setup true? I don't remember much about the preview (need to read it again), but if thats true, Anand should have refused to print the results. If it turns out that Intel crippled the A64, Anand will be in a very poor position indeed.

You posted as I was typing BrownTown. I'm not concerned about which is faster, but I do wonder about how those speeds are tested. Intel not allowing Anand to inspect the system means they had something to hide. That something might have been information they didn't want made public for valid reasons, but it would have caused me to refuse to go public with the benchmarks produced.

Anyway, thanks for the input. I hope Conroe does well.


Well, here is a direct quote from the Anandtech preview:

"The AMD system used 1GB of DDR400 running at 2-2-2/1T timings, while the Intel system used 1GB of DDR2-667 running at 4-4-4. Both systems had a pair of Radeon X1900 XTs running in CrossFire and as far as we could tell, the drivers and the rest of the system setup was identical. They had a handful of benchmarks preloaded that we ran ourselves, the results of those benchmarks are on the following pages. Tomorrow we'll be able to go into great depth on the architecture of Conroe, but for now enjoy the benchmarks.

As far as we could tell, there was nothing fishy going on with the benchmarks or the install. Both systems were clean and used the latest versions of all of the drivers (the ATI graphics driver was modified to recognize the Conroe CPU but that driver was loaded on both AMD and Intel systems).

Intel told us to expect an average performance advantage of around 20% across all benchmarks, some will obviously be higher and some will be lower. Honestly it doesn't make sense for Intel to rig anything here since we'll be able to test it ourselves in a handful of months. We won't say it's impossible as anything can happen, but we couldn't find anything suspicious about the setups."

It doesn't look like Intel tried to hide much from the Anandtech staff. If they could verify that the drivers looked the same and that the rest of the system setup was the same than it looks like they had some freedom in looking the rig over.

I'd like to see some other review sites test Intel's new chip out (wishful thinking for now) to get a larger sampling, but I believe that the current results that we've seen are probably right. It would be a PR nightmare for Intel to fool with results to make the Conroe look better. They aren't going to sell anymore chips since Conroe isn't even available for sale yet and, when they do become available, there will be plent of reviews out there to confirm or dispute these early findings.
 

BrownTown

Diamond Member
Dec 1, 2005
5,314
1
0
as for the cache issue, if you look at the scaling of processors based on increased cache size you will find that it often does little good. Just look at Presler vs Smithfield, almost no benchmarks show a difference of more than 1% for Preslers 4MB of cache. Conroe cache gives it a large advantage because it is very fast, not because it is so huge. Obviously the size helps, esecially in a few limited cases, but we all know that by the time you get to the MB cache sizes you really arent reducing miss rates by very much.
 

coldpower27

Golden Member
Jul 18, 2004
1,676
0
76
Originally posted by: BrownTown
as for the cache issue, if you look at the scaling of processors based on increased cache size you will find that it often does little good. Just look at Presler vs Smithfield, almost no benchmarks show a difference of more than 1% for Preslers 4MB of cache. Conroe cache gives it a large advantage because it is very fast, not because it is so huge. Obviously the size helps, esecially in a few limited cases, but we all know that by the time you get to the MB cache sizes you really arent reducing miss rates by very much.

Once were at release we will have the Allendale cores to compare it to, so just take a 2.4GHZ Conroe and put the multiplier down to x8 and walla a 2.13GHZ Conroe vs the 2.13GHZ Allendale and we will see how much the cache actually helps....
 

BrownTown

Diamond Member
Dec 1, 2005
5,314
1
0
well, unfortunately we don't have any allendale cores currently so it aint worth much to talk about them. Especially since by the time they come out we will also ahve 20 different reviews of Conroe running all kinds of tests and we will have a very good picture of what makes Conroe so good. Also, should be noted that Yonah has the exact same L2 cache as allendale, not that that tells us anything, but still gonna throw it out there as a point of interest.