Please register or login. There are 0 registered and 964 anonymous users currently online. Current bandwidth usage: 326.30 kbit/s December 16 - 09:34am EST 
Hardware Analysis
      
Forums Product Prices
  Contents 
 
 

  Latest Topics 
 

More >>
 

    
 
 

  You Are Here: 
 
/ Forums / Intel's Conroe, it is good, but don't hold your br...
 

  Intel's Conroe, it's good, but don't hold your breath? 
 
 Author 
 Date Written 
 Tools 
Continue Reading on Page: 1, 2, 3, Next >>
Sander Sassen Mar 16, 2006, 07:10am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List Replies: 56 - Views: 3828
It'll be interesting to see how this develops, TechReport has a similar article up as the one from Anandtech, with a set of benchmarks being run on both systems.

Best regards,


Sander Sassen
Editor in Chief - Hardware Analysis
ssassen@hardwareanalysis.com
Want to enjoy fewer advertisements and more features? Click here to become a Hardware Analysis registered user.
Bungle Mar 16, 2006, 07:36am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
I agree, even if Intel's new chip is superior to what AMD is offering (and it certainly looks like it is) thats a good 6 months out, and in computers if you always wait for the next best thing you will be waiting forever, because in 6 months when conroe launches both AMD and Intel will be demoing their next chips. You might as well buy the best you can now and worry about future products when you upgrade in the future.

Core i7 4770K @ 4.4GHz | Corsair H110 | Gigabyte GA-Z87X-UD5H | 32GB 1866MHz Corsair Vengeance
2x EVGA GTX 780 SLI | 256GB OCZ Vector SSD | 4TB Hitachi 7K4000
Corsair AX1200 PSU | Corsair 650D | Windows 7 Ultimate x64
Blood Brother Mar 16, 2006, 07:37am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
i've read that it outperforms AMD by 40%. but reconsidering intel's old crap i seroiusly doubt it

--------------------------------------------------------
s754 Clawhamma 3000+
g'byte K8Ns 250
1 Gb kingston ddr400
mSi 6600gt AGP

Bungle Mar 16, 2006, 07:42am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
if you're refering to the anandtech benchmarks, they updated them the next day. They made a mistake in the FEAR test. The AMD based computer was running one resoloution mode higher then the Conroe computer. They re-ran the tests and the AMD was 20% behind.

Core i7 4770K @ 4.4GHz | Corsair H110 | Gigabyte GA-Z87X-UD5H | 32GB 1866MHz Corsair Vengeance
2x EVGA GTX 780 SLI | 256GB OCZ Vector SSD | 4TB Hitachi 7K4000
Corsair AX1200 PSU | Corsair 650D | Windows 7 Ultimate x64
A_Pickle Mar 16, 2006, 07:50am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
There's a lot of Conroe benchmarks. HotHardware, Bit-Tech, PC Perspective and X-Bit Labs all have some Conroe benchmarks, too. Anandtech ran the first set of Conroe benchmarks a little bit iffy on both sides, like you mentioned they had the resolution higher on the AMD FX-60 in the FEAR benchmark, and they shortchanged the Intel by not enabling SMP in Quake 4. They remedied all of this in round 2, which also addressed some criticisms made by Rahul Sood in his blog.

-Pikl

Radomir Jordanovic Mar 16, 2006, 08:24am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
I remember a while back, two Hypersonic PCs were reviewed by PC Gamer. One had an AthlonXP 2100, and the other had a 2.53GHz P4. Otherwise, the two PCs had identical ram, video cards, etc. The AMD system was behind by no more than 30 fps in most tests despite the 730MHz clock speed difference and other performance advantages the Intel processor had.

In my mind, AMD has always been the better choice. The processors were cheaper and pretty fast practically all throughout the Socket A line. Do you think AMD will still hold a large market share after Intel starts massively and carelessly advertising their new product line?

Silverstone 750W, Asus P5KC, C2 Q6600, 2GB OCZ DDR3 1600
ATI HD 4850 512MB, 1TB Caviar. 1TB Seagate
someone else Mar 16, 2006, 10:40am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Let me preface by saying that I believe the conroe they benched has a legit lead over the athlon used for testing. Just not 20%, and possibly not much at all outside of the few benches they were allowed to run. I disagree that intel's got no incentive to mislead with this round of benches, they have every reason. Regardless of what theoretical hit they might take later for this, they are currently in full-scale open wound mode....and the bleeding is accelerating. Their sales were slowing anyway, so if people hold off now to wait for conroe, this will hurt amd as much/more than intel (that's their idea...we'll see if it happens that way).

I believe that this, above all else, is the reason for these uncharacteristic "sneak peeks". They need to stop the bleeding now. Here's some of the more glaring issues with these benches...

1. No one was allowed to look inside the case. My take? The 2.6ghz conroe they used was cherry picked, and they aren't getting good yields at that speed yet (water/peltier cooling used possibly?). Does no good to have superior performance per clock if you can't clock high enough to beat the competition. We're still 6 months from launch, so this might get straightened out. But notice people were saying yonah was going to hit 2.5ghz....has that happened yet? Not even close.

2. No one was allowed to look in Device Manager. This just screams shenanigans. They admittedly used a "special" ATI Crossfire driver that was specially made to "recognize conroe". Don't know about you, but I've never had a video card that needed special video driver to use a particular cpu. Even better, in the interest of "fairness", they used this same Conroe-specific driver on the athlon machine. There is virtually no chance that this was the optimal setup for the athlon. This is also only mentioning the video driver. There are limitless things you could hide in device manager that could artifically inflate conroe, or cripple the athlon.

3. Why would you bench a cpu using games at low/moderate resolutions, and then use two vid-cards in crossfire mode (and why wouldn't you use SLI instead, which is widely known to be superior on Athlon64). Along this same track, they didn't even use a particularly good crossfire chipset. It's the original rev (480), known to be subpar, and newer less broken versions (580) have started showing up. Intel's excuse? The newer ones weren't available enough two weeks ago for us to get one to use. Riiiiight, intel, the king of the x86 world can't get one, but little review websites can.

4. Every website that got to "review" these machines were only allowed to bench the same 5-6 things. They couldn't possibly be cherry picking benchmarks that favor them, could they? No, intel, the pargon of truth in advertising would never possibly do that. Remember years ago when they said P4 (and later prescott) were gonna set the world on fire? Well, they sorta did, but in more of a literal way than intel intended.

So what do we know for sure? I will concede that it's likely Conroe holds a < 20% lead over current athlons in the the 5-6 benchmarks that they have displayed. That's about it. AM2 is not going to be as doom'n'gloom as has been predicted here. Expect parity with conroe at least. If amd's got more tricks up their sleeve than they've been saying, maybe a small lead over conroe (5-10%). Given that am2 is going to be out well before conroe, that should be enough to hold them over until early '07, when K8L debuts. This will be a more significant reworking of the K8 core, and will almost certainly leave behind the conroe we've seen so far.

Mike L. Mar 16, 2006, 10:40am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Mar 16, 2006, 10:44am EST

 
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Radomir Jordanovic,

You can't base something off of a company now, that happened to them so many years back that's why it's called "improvement". I will agree that Intel screwed up this time and let AMD take the crown.. but as it seems, that's all changed now. Once again, it'll be a good 6 or so months before we even see Intel's new chip in stores, for purchase. IF indeed, these are actual, true figures we're seeing now.. then I think AMD seriously needs a miracle; because it looks like they've just screwed themselves on their AM2 platform.

___________________________________________
I am what you would call depressingly comfortable...
Radomir Jordanovic Mar 16, 2006, 11:32am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Mike,

What I meant is that AMD's slower processor, which was that much slower, was only 30fps behind in benchmarks. Therefore, CPU speed shouldn't matter all that much. The reason I say this is that I believe that video cards don't require all that much more power than what an FX-60 can provide so that they can pretty much reach close to their max fill rates.

Silverstone 750W, Asus P5KC, C2 Q6600, 2GB OCZ DDR3 1600
ATI HD 4850 512MB, 1TB Caviar. 1TB Seagate
Yousuf Khan Mar 16, 2006, 12:22pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
All of the testing was done on a closed system set up by Intel itself. This is a red flag, and immediately should be an alert that you need to wait for actual production systems to show up before anything concrete can be said. When production systems come out, then all kinds of scenarios can be tested, such as motherboards and chipsets combinations, and production drivers. The motherboard in Conroe is non-production, it's a highly optimized engineering testbed from Intel itself, and it's being tested against a production mobo/chipset that is normally never used by FX buyers -- the particular chipset in question (ATI Rx480) is used more by laptop computers running Sempron or Turion, or integrated graphics value desktops running Sempron or low-end Athlon 64.

CrAsHnBuRnXp Mar 16, 2006, 12:51pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
I wont believe any benches between the Conroe and an AMD equivalent until it comes out officially. Right now, these benches are only preliminary.

Josh Mar 16, 2006, 01:19pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Someone else - you have some Very good points that I didn't think about. I guess AMD has a better chance to stay on top then what I was thinking. Plus, if the AM2 does come out a bit earlier then conroe, it will be able to have some upgrades and updates before conroe is released.

Never be afraid to try something new. Remember, amateurs built the ark. Professionals built the Titanic.
AMD Phenom 9850 || DFI Lanparty 790FX-M2RS || 8GB OCZ Platinum || XFX 7900GS XT ||
Check eBay computerman_89
CrAsHnBuRnXp Mar 16, 2006, 01:22pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
I read it was coming out on the June 6th.

Josh Mar 16, 2006, 01:28pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Which one - Conroe or AM2?

Never be afraid to try something new. Remember, amateurs built the ark. Professionals built the Titanic.
AMD Phenom 9850 || DFI Lanparty 790FX-M2RS || 8GB OCZ Platinum || XFX 7900GS XT ||
Check eBay computerman_89
Mike L. Mar 16, 2006, 01:29pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Mar 16, 2006, 01:30pm EST

 
>> Re: Intel's Conroe, it's good, but don't hold your breath?
I agree that a video card can only rely on so much help from a CPU. A matter a fact, buying the strongest CPU doesn't mean anything anyway, all the work is left up to the video card. That's the whole point of buying a top of the line video card, because the video card itself is supposed to do all the work anyway. The only time it starts tapping into physical RAM and help from the CPU is when you know it's getting flooded. So I see your point now..

___________________________________________
I am what you would call depressingly comfortable...
Josh Mar 16, 2006, 01:30pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
That's true with games, but with other non-3d apps, cpu is the muscle.

Never be afraid to try something new. Remember, amateurs built the ark. Professionals built the Titanic.
AMD Phenom 9850 || DFI Lanparty 790FX-M2RS || 8GB OCZ Platinum || XFX 7900GS XT ||
Check eBay computerman_89
A_Pickle Mar 16, 2006, 02:38pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
CPU is muscle for everything but games. The only area where a video card honestly matters (other than gaming) is 3D animation, and even then you CAN do 3D animation without a video card, just... not very well and obsessively difficult.

-Pikl

Mike L. Mar 16, 2006, 04:12pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Agreed.

___________________________________________
I am what you would call depressingly comfortable...
Gerritt Mar 16, 2006, 06:59pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
It seems that I read a article in PCWorld that compared the new Intel vs. Present AMD offerings.
If I recall properly, the Intel won in 4 out of 7 gaming benchmarks, by a factor of less than double digit FPS.
In the other benchmarks the top off the line AMD won by a factor of less than double digit FPS.
All seven of the benchmarks were games.
I believe the AMD equivilent is the FX-60, and given the pricing available the Intel Conroe may be the better price point.

Gerritt

Ad Astra Per Aspera
(A rough road leads to the Stars)
We all know what we know, and everyone else knows we are wrong.
System Specifications in BIO
CrAsHnBuRnXp Mar 16, 2006, 07:20pm EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
"Which one - Conroe or AM2?"

AM2

angryhippy Mar 17, 2006, 02:18am EST Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> Re: Intel's Conroe, it's good, but don't hold your breath?
Some basic details of the thing would be nice.

One tidbit shared in one of the keynote sessions revolved around power. Since power usage goes up as the square of the voltage, if you cut the power in half, the actual performance drops by less than 15%. This allows the designers to build a dual-core processor that uses roughly the same power as the previous top single core CPU. In essence, you get two processors, each with 85% of the horsepower of the previous, high frequency single-core CPU.

But Conroe isn't just a Dothan for the desktop.

For one thing, Conroe is a four-wide architecture. The CPU can issue four instructions at the beginning of the pipeline and retire four at the end. This increased the instruction bandwidth, since earlier Pentium 4 and Pentium M CPUs were three wide. So on a per-clock basis, more work is being done

Each core of Conroe has a dedicated 64KB of L1 cache, split into 32KB for instructions and 32KB for data. Each L1 cache has two prefetchers for data and one for instructions. The 4MB L2 cache is shared—and sharing is dynamic. That is, each core can allocate what it needs, up to the full cache size. An item of data in the L2 cache that needs to be shared by the two processors is, well, shared—the data is not replicated in L2 for each core, as it would need to be if each core had its own, exclusive L2 cache. The L2 cache has two prefetchers of its own, which are flexibly allocated between cores as needed.

Conroe (desktop), Merom (mobile), and Woodcrest (server) support something Intel calls "micro-op fusion." An x86 instruction is decoded into micro-ops. But in certain cases, two adjacent, dependent micro ups can be combined into a single op and executed in a single cycle. This was something that was inherited from the Pentium M. What's new is "Macro-op fusion." In this case, certain x86 instructions may also be paired into a single instruction, then executed in a single cycle.

So in certain cases, five instructions can actually be read from the instruction queue, then execute as if only four instructions were issued. These cases, as it turns out, are fairly common, such as loads and compares, so any performance benefit will occur across a wide array of applications.

Substantial enhancements have been made to the SSE engine. SSE registers are 128 bits wide, but in the past, data paths within the CPU were 64 bits wide. That meant that SSE instructions would take two cycles. In Conroe, the entire SSE engine is 128 bits, not just the registers. So SSE instructions can now execute in a single cycle.

This can make a huge difference in applications that use SSE. For example, in our past benchmarks using Windows Media Encoder, an Athlon 64 FX-60 running at a stock 2.6GHz would encode a file faster than a Pentium Extreme Edition 955. But we saw a Conroe system, running at 2.66GHz, finish a Windows Media Encode session of a relatively small file two minutes ahead of an Athlon 64 FX-60 clocked up to 2.8GHz (roughly 3:13 on the FX-60 versus 1:13 on the Conroe system.)

All the CPU horsepower that Conroe will deliver will exist in a thermal design power rated at 65W. This is half that of the current Pentium Extreme Edition 955, and less than most AMD dual core CPUs. Using techniques developed for the Pentium M, extensive clock gating is built into the CPU. As Bob Valentine, one of the architects on the project noted, unless a particular part of the CPU actually needs to be used, it's not turned on. An example was given of a tight loop that could be executed while the instruction fetch and pre-decode section wasn't active.

http://snipurl.com/npw3

Basically this is a whole new ball game. Release is scheduled for Q3 of '06

Get Hippied out!
http://www.angryhippy.net
Me at work: http://tinyurl.com/3nvncb3
My rig! A Blah blah.With a blah blah! SWEET! http://tinyurl.com/4yujmff
Da Beast! http://tinyurl.com/3sapr2b
i5 3570K 4.6GHz http://snipurl.com/26r3cot
Win7-8 Pro 64bit

Write a Reply >>

Continue Reading on Page: 1, 2, 3, Next >>

 

    
 
 

  Topic Tools 
 
RSS UpdatesRSS Updates
 

  Related Articles 
 
 

  Newsletter 
 
A weekly newsletter featuring an editorial and a roundup of the latest articles, news and other interesting topics.

Please enter your email address below and click Subscribe.