Actual deneb review/comparison to Intel

Page 7 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
here is what i compare
Agena vs Kentsfield = 65nm
Deneb vs Yorkfield = 45nm
AMD vs Nehalem = well nehalem is a new cpu, not a shirnk like kentsfield to yorkfield

AMD is doing a replay of the RV770, if you cant beat them, offer good performance and then undercut them 😛

Given the fact reviews of deneb uses beta bios or unsupported BIOSes, then we got drivers that arnt optimized for deneb, even the applications they test with isnt optimized for deneb yet, heck Agena got a upto 20% booast on x264 encoding when the x264 developers updated it

Nehalem & Yorkfield has been out for ages now

can any of the intel fanboys and girls remember the Pentium 4 fiasco?

i currencly got a Phenom 9850 in my system and it runs great its fast in games, wont be upgrading cpu for some months atm, meybe in AM3 deneb

more memory meybe 4GB more, currently got 4GB, and a faster graphics card (got 4870 atm in the upcoming months)
 


Un Kentsfield was not a die shrink. It was a new CPU, Core 2 Quad, based on the old Coppermine Pentium III just with much better IPC and much higher clocks (Coppermine topped out around 1.3GHz then came Pentium 4) and the 65nm tech from the old Ceader Mill Pentium Ds.

So a beta BIOS is reason for results? Meh. Not too sure. The chipsets already support Deneb so I doubt that the BIOS would undercut the performance that much. And when people used multiple GPUS Nehalem got 35%+ in some games better performance per clock vs older CPUs. Encoding got better for Yorkfield when they added support for SSE 4.1. But we shall see.

The Pentium 4 fiasco? You mean the Prescott fiasco right? Cuz the Pentium 4 Northwood (130nm) was a good CPU. I have one that at the time I bought it was the best performer for the price. Then Athlon XP took Prescott to school. Then you had Pentium D vs Athlon X2, we all know how that ended up since Prescott was hot running. But Ceader Mill based Pentium Ds were decent as they had 3GHz CPus at 65w TDP.

Thats great you have a Phenom. That was your choice. Problem is that we have no concrete performance numbers. Someone says it will compete with Yorkfeild on a even level. Other sites suggest that the highest end barely out paces the 2 year old low end Q6600. Mixed reviews tend to bode bad.

That and all they have been focusing on (AMD that is) is how well it OCs on LN2. We have yet to see any big review sites post well..... just that. A full review with not only OCing results but real world apps and gaming results.

Oh and Nehalem just came out in November 2008...... thats not ages ago. Yorkfield has been out a little over a year, depending on which CPU and Kentsfield is over 2 years old now and still the Q6600 is one of the best bang for buck quad cores out there period.
 
I would surely like to know how you figure that a Q6600 is better than a Phenom2 when a Phenom 1 perfores similarly in games perhaps behind by 100mhz or so. Since most of the time when I see the reviews the phenom is shown using ddr2 while the q6600 is using ddr3 as well as other advantages usually do to phenoms being shown at stock speed vs a overclocked q6600 and somehow this proves a q6600 is going to be better than a phenom 2!? Next of all if the q6600 is such a great performer why is it that intel is cutting the price on them so much? A few months ago they were running for a lot more than now. Could it be that the amd's were actually doing quite well with the performance/price on the 9950be's?

http://www.tomshardware.com/charts/desktop-cpu-charts-q3-2008/3DMark-Vantage-CPU,817.html
http://www.techwarelabs.com/reviews/processors/amd-phenom-9950/

Yes a q6600 will oc higher i'm sure everyone here knows that but for a long time a q6600 cost more than an AMD for similar performance. The tests though fail to mention what mobo chipsets/gpu/ram/etc was used to get each result so since it's difficult to compare a phenom1 against a q6600 directly when they have both been on the market for so long i'd like to know how you figure a phenom 2 which hasn't even been released yet is not as good?

And to let you know as a matter of fact I AM pro-amd.

That doesn't change the fact that a synthetic benchmark is "interesting" but it doesn't mean that it will perform as well in a real world app.

check out http://www.tomshardware.com/forum/page-256560_28_100.html#t1891767 for my post on the performance with multi gpu core i7's and c2extremes and you can see than a phenom 1 even is able to do passingly well while competeing against products costing 6x more $$$ and then come back and tell me how well a q6600 performs.

edit: Anyway my bad I was reading the second page and posting and now I see i'm on the 4th page where everyone is discussing other things my bad.
 
It's one thing to say that the Q6600 is aging, and will likely fall behind the brand new Phenom II CPUs when they show up. That's probably true. However, to claim that a current 65nm Phenom can keep up with a Q6600 in almost anything is basically laughable. The 6600 is significantly faster, and the reason that is is currently so cheap is simply because AMD cut their prices even lower (you can get a 9950 for $170), and because Intel has plenty of higher end CPUs in the $300 range that would take sales from the 6600 if it were too expensive.

As for that scaling article? While it is true that the i7 is the point, you can also easily see in that article that the QX9770 beats the phenom in almost every test, and it uses the same basic architecture as the 6600. Keep in mind that a typical 6600 can easily be overclocked to 9770 or greater speeds (I would say 3.4-3.6GHz is fairly common, and 3.2 is easy). Because of this, I can say with reasonable certainty that a 6600 beats the Phenom I in almost every way (with the exception of extremely memory intensive stuff), although if you don't overclock, the 9950 is pretty close to the 6600.
 
Yes that is true but this doesn't prove that a q6600 is going to outperform a phenom2 cpu. Since I doubt AMD would release a phenom2 that performs worse than a q6600 and charge more for it. Also as you can see in the article a QX9770 is working at a similar clock speed as many ppl push their q6600's too while the phenom1 is at 2.6ghz since there isn't any mention that any cpu was oced at all and was using ddr2 ram ram etc while the qx was using what ddr3 1600 or something? Obviouly the focus of the article isn't the cpu's but people saying that a q6600 is gonna perform at the same level as a phenom2 sounds a bit far fetched.

qx9770 is somewhat different from a q6600 but either way the fact that a 2.6ghz phenom1 can post those numbers against a 3.2ghz qx9770 is not really that bad. But say you did take a q6600 and oc it to 3.4-3.6ghz the phenom can be oced as well to around 3.2-3.4ghz so what does that prove? Obviously I expect the qx9770 to perform better but at $1400 I mean...seriously? I did some half hearted attempts to find any direct comparisions between similarly prices intel quads vs amd quads but usually once you read through them intels have either 1) A mhz advantage 2) a ram advantage 3) a price advantage i.e. using a significantly more expensive cpu vs a much cheaper amd cpu or 4)some combination of all of these.
 
I didn't say that the 6600 is going to outperform a phenom II (actually, I would be quite stunned if the Phenom II wasn't faster than the 6600). I was saying that the current gen phenoms can't match the 6600 all around yet. As for the QX using fast DDR3, it shouldn't matter. I'd bet the actual memory bandwidth was somewhat similar (as the QX is limited by the FSB).
 
However it does cause concern as to validity of results since it not following a meathod which singles out the cpu and there are many other factors that then need to be considered as to whether it's the cpu that is lagging behind or the ram/chipset/etc causing the problems. That is why I would rather look to see how it does in a video game enviro using similar ram chipsets and gpu's etc to see how a system performs as a whole eince it's not like I'm buying a cpu that can do everything I need a system that will work to give me the best performance/$$$
 


Well the MHz advantge is dependant. I have seen stock Q6600 vs Phenom 9950 reviews where the Q6600 keeps up and sometimes beats it and the 9950 has a 200MHz advantage. The memory advantage does not exist. Yes it can have higher end memory but wasn't that the point of a IMC? I mean a Phenom should be able to take more advantage of the memory bandwidth available to it than a C2Q can. The price advantage normally works in the cheaper CPUs advantage really. Thats why all the AMD ads on Youtube pit the 9950 against a QX9650 instead of a Q6600 or Q9450. Because they want to seem cheaper.

I never said the Q6600 was better than Phenom II. We don't know yet. It may be. I have seen a lot of mixed reports. My point is that the Q6600 is still a damn good buy for anyone who wants a quad core. And still will be until they stop making it of something higher end comes out at the same price.

until then, the Q6600 is a great CPU. Its not really able to be argued. Much like the older Athlon X2 5000+ BEs are great CPUs or the current E8400 is. Mainly because it performs great, OCs great and is cheap.
 




I can live with that. Until the reviews are in we should a sit back pop some brews and play some games.
 
Well Intel has said the Q6600 will be discontinued sometime this month. So id expect stock on Newegg to dissapear sometime in March or April. Intel has already discontinued two Xeon quads, the X3350 (Q9450) and X3360(Q9550) when they released the X3370.
 
Even if it performs the same clock for clock, the P2 is already at 3Ghz, and looks like it ocees better as well, so the Q6600 isnt in the same league. Its the Yorkies AMD has to deal with, and at stock, itll do well, and clock for clock, it comes down to perf/costs, and platform life as well
 
now there really needs to be a flood of some high end boards when these are released. i already got the one i want but in looking one thing is apparent.. there are not a lot of high end boards for AMD by comparison. maybe that was because manufacturers didn't see any real demand to be had but i think that will change now.
 


Exactly my thoughts. Current boards suck.
 
i saw theirs. i also saw the one i bought (Foxconn destroyer) and a ASUS or two. now go look on the other side of the fence. not saying there are none just verrrry few. it is kind of important to have choices like that if that is indeed the market you want.
 
I dont know of any XFX AMD chipset board, but I do know they offically announced ATI/AMD support for 2009 on their website. But I have seen the XFX X58 motherboard.

For an AMD chipset, I would rather have Gigabyte or MSI though. Maybe Foxconn.
 
Well the issue I'm having is whether or not to wait for an AM3 ph2 or get a AM2+ ph2. On the roadmap the AM3 were supposed to have a lower tdp (i think i saw that, unless i was daydreaming it).

Question: When some more tests and comparisons come out what is a better testing method? To test against equal mhz or to test against price? Or would you rather see both?

When do you think we would get unlocked multi's? Or is it releasing unlocked?
 
i bet it will be released with the unlocked multi even though it will have a small bearing on overclocking anyways. it doesnt seem to be a roadblock on the i7's at all.
 



Actually, I think the RV670 would be a better analogy, and in more ways than simply referring to it being a die shrink.
 
The RV670 is not a better analogy than the RV770. Mainly because all it was was a RV600 that used less power. Same performance.

The RV770 on the other hand was able to keep up with the GTX280 in most areas, and beat the GTX260 in most areas as well all for less than half the price.

And ep, interesting chart. Will have to see more though. Hard to trust a forum really. I mean they have a guy talking about how a AMD chip is smoother than a Intel chip, very like some others here. I don't feel like registering to tell him I usually play a game, rip a movie while listening to WMP and surfing the web all smoothly but meh.

Anyways. We shall see more soon I hop.e
 


Clock for clock slower in what? Synthetic benchmarks? I don't care how many kilonodes per second a CPU can get in Fritz Chess, how well it does in Cinebench, or what score it gets in Vantage. The i7 gets 22k points in Vantage, and can't even beat Phenom II at 1680x1050 most of the time in the games that were tested (especially with 8xAA and 16xAF).

Most enthusiasts are going to go for gaming power, and the Phenom II will have it in spades. Sure, C2Q and i7 game better at 1280x1024. But are you really going to buy a more expensive platform to game better at lower resolutions rather than buy a less expensive platform to game better at higher resolutions and eye candy?
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom