Xeon X3210, X3220 vs C2Q Q8400

[arniks8]

This is vastly wrong and you should feel bad for thinking as such... Fx8150 is on par with nahalem i7s at the same clock rate which is a generation newer than those xeons and its a relatively large generational gap... Fx 8350s are just below sandy bridge i7s and just above westmere i7s so roughly 2.5 generations better than those xeons Ltt even did a video of those Xeons vs 8350s and the 8350 beat them in any CPU load easily

Overclocked nahalem is faster than FX8. Nahalem has low clock speed, so pushing it up makes it quite faster. Sandy is faster at stock and way faster overclocked. X5450 is about the same @4.4Ghz. Keep in mind that I meant single core performance, because with a real multithread scenario Xeon will be slower due to less cores.

On LTT video Xeons weren't clocked high.

[GloriousPain]

Overclocked nahalem is faster than FX8. Nahalem has low clock speed, so pushing it up makes it quite faster. Sandy is faster at stock and way faster overclocked. X5450 is about the same @4.4Ghz. Keep in mind that I meant single core performance, because with a real multithread scenario Xeon will be slower due to less cores.

On LTT video Xeons weren't clocked high.

X5450 and such xeons are not nahalem though they are older than nahalem/westmere with a decent overclock they might match fx8120/50 at stock... The nahalem stock is about on par too but yes nahalem oc beats 8150 oced and even a lot of 8350s westmere same thing... Sandy stock beats all but some of the best OCS of Anything predating it and it can go even further beyond... But I think to say that an old 775 setup beats all fx8s is just an exaggeration yes oced they'll beat a stock or mild oced 81 doesn't just make them better but yeah the ltt video didn't have them oced it also didn't of the fx and the tests were at 4k where CPU bottlenecks are far less present because of GPUs hitting their limits

[arniks8]

I know that X5450 is harpertown. I don't understand where I said that LGA 771 xeons are faster, because I said:"X5450 is about the same @4.4Ghz" ( single core performance of course ), I did say that nahalem is faster when overclocked.

[Dabombinable]

X5450 and such xeons are not nahalem though they are older than nahalem/westmere with a decent overclock they might match fx8120/50 at stock... The nahalem stock is about on par too but yes nahalem oc beats 8150 oced and even a lot of 8350s westmere same thing... Sandy stock beats all but some of the best OCS of Anything predating it and it can go even further beyond... But I think to say that an old 775 setup beats all fx8s is just an exaggeration yes oced they'll beat a stock or mild oced 81 doesn't just make them better but yeah the ltt video didn't have them oced it also didn't of the fx and the tests were at 4k where CPU bottlenecks are far less present because of GPUs hitting their limits

Overclock a Xeon X5450 to 4.4GHz=i5 4440 performance. Overclock 2x Xeon X5470 through BSEL mod to 4GHz=close to an i5 4440 single threaded and flogging an FX 8350 multi threaded. BTW I know that from my experience with my X5450 and i5 4440.

[djdelarosa25]

Would a Q9550 work on this board?

[arniks8]

Would a Q9550 work on this board?

If it supports other core 2 quad CPU's, then it would support q9550

[djdelarosa25]

If it supports other core 2 quad CPU's, then it would support q9550

Got the Q9550 for 30 bucks. Was it worth it?

[Dabombinable]

Got the Q9550 for 30 bucks. Was it worth it?

Yep. Its pretty much just a Xeon except with lower binning.

Q9550: http://ark.intel.com/products/33924/Intel-Core2-Quad-Processor-Q9550-12M-Cache-2_83-GHz-1333-MHz-FSB

E5440: http://ark.intel.com/products/33082/Intel-Xeon-Processor-E5440-12M-Cache-2_83-GHz-1333-MHz-FSB

 

Spot the difference in specs (apart from the Xeon consuming about 15W less).

[arniks8]

Got the Q9550 for 30 bucks. Was it worth it?

Not bad, but as Dabombinable said, xeons are binned and therefore overclock better.

[djdelarosa25]

Not bad, but as Dabombinable said, xeons are binned and therefore overclock better.

 

My mobo sucks a OC'ing. Won't even let me change voltages. Besides, I live in the Philippines and I can't find any mod stickers. And I'm just a HS student, and the money I used was my mom's (lol). She would rarely spend money on these types of things (that's why my PC is crap) and she only agreed that I need an upgrade since we use Adobe Premiere at school. Spending more money than just this product would give her a headache. Thanks for the help

[arniks8]

My mobo sucks a OC'ing. Won't even let me change voltages. Besides, I live in the Philippines and I can't find any mod stickers. And I'm just a HS student, and the money I used was my mom's (lol). She would rarely spend money on these types of things (that's why my PC is crap) and she only agreed that I need an upgrade since we use Adobe Premiere at school. Spending more money than just this product would give her a headache. Thanks for the help

Just because your a student doesn't mean that you can't earn or save money. My parents didn't give me a penny for my PC. It took me two years to get the cash, but I got it and surprised myself about my capability, but now my CPU died and I can't buy a new one , so q6600 it is for a while ( hardcore money problems ).

[GloriousPain]

Overclock a Xeon X5450 to 4.4GHz=i5 4440 performance. Overclock 2x Xeon X5470 through BSEL mod to 4GHz=close to an i5 4440 single threaded and flogging an FX 8350 multi threaded. BTW I know that from my experience with my X5450 and i5 4440.

http://www.anandtech.com/bench/product/53?vs=1368

They're the same architecture so I'ma just assume IPC is the same (which other small variance it is) the 2.4Ghz Xeon X54-- just with less binning, so yeah in Cinebench it scores 2,778 let's assume you can get 4.8Ghz out of it? Which doubles it's score (even though clock rates usually have slightly diminishing returns) Then it's still only at 5,556... Versus the i5 4570S which is a power efficiency based model, clocked at 2.9 Ghz, 200Mhz LESS than i5 4440 and it's a whopping 6,994, Roughly 26% LESS than an even lower clocked model than the i5 4440  with an overclock on the Xeon higher than even your 4.4Ghz number... 

TLDR on Maths; X5450 @4.8Ghz, Still worst than an underclocked i5 4440 by 25~26% in single core performance. 

Iff 4.8Ghz X5450 or C2Q Q6600 in cinebench R10 single thread= 5,556 (In theory)

Stock FX 8320 and 8150 (3.5 and 3.6Ghz Respectively) are roughly= 4,000

(Source: http://www.anandtech.com/bench/product/698?vs=434 )

Stock 8350 and OC'ed 8320E (4Ghz and 4.8Ghz Respectively) are roughly=4,338 and 5,107... 

(Source: http://www.anandtech.com/bench/product/1403?vs=697)

Also I'd like to point out that all the FX 83XX should have identical IPC (instructions per clock) but if you do the math they don't... 

FX 8320   Score 3,987 / Clock 3.5Ghz = IPC 1,140

FX 8350   Score 4,338 / Clock 4.0Ghz = IPC 1,084

FX 8320E Score 5,107 / Clock 4.8Ghz = IPC 1,063

FX 8320E Score 4,056 / Clock 3.2Ghz = IPC 1,267

http://www.anandtech.com/bench/product/1402?vs=1403 and http://www.anandtech.com/bench/product/698?vs=697 is my sources plus a calculator with small rounding for the math portion

Also notice IPC seemingly goes down as clock rate goes up, infact the "IPC" difference between the 8320Es (which IS the exact same chip anandtech just benched it at stock and OC'ed)  is ~20%   with a 50% clock increase so there's no way the doubling clock of the Q6600 I did earlier would double the Score...  clock rates the higher they go just have diminishing returns which is why FX9590 doesn't beat 8320 or 50 by much... taking that into account an insanely overclocked Q6600 or Xeon x5450 would be about even slightly better in single core performance with 8350 at stock, and you'd be missing some very consumer centric features like possibly DDR3, Sata 6Gb/s possibly even 3Gb/s, and PCIe 2.0 possibly even 16x Pcie, lots of variables based on motherboards and chipsets  SLI support ect ect... Also do remember the difference in core count 4 vs 8... But in the end FX vs Xeon X54-- with a decent OC is sadly fairly close...

 

All that aside, I think FX vs Old Xeons x54-- being OC'ed (which on it's own presents some challenges) being fairly close in terms of sheer single core strength and in multicore FX should win,  and in feature set also should win... and Nahalem faired better than I thought after doing some quick searches which in my opinion only further proves my earlier statements in this thread that OP should just try to go with a nahalem or westmere setup because they're the best solution out of the 3 choices here (FX, Xeon X54--, or i7 8/9--/Xeon X55/56--) in terms of price to performance and overall performance, and at the moment in terms of the used PC market I really hope this means in another year or 2 Sandy bridge gets the pricing/availibility Nahalem/Westmeres showing now considering Sandy bridge was where IPC had it's big jump

[Dabombinable]

http://www.anandtech.com/bench/product/53?vs=1368

They're the same architecture so I'ma just assume IPC is the same (which other small variance it is) the 2.4Ghz Xeon X54-- just with less binning, so yeah in Cinebench it scores 2,778 let's assume you can get 4.8Ghz out of it? Which doubles it's score (even though clock rates usually have slightly diminishing returns) Then it's still only at 5,556... Versus the i5 4570S which is a power efficiency based model, clocked at 2.9 Ghz, 200Mhz LESS than i5 4440 and it's a whopping 6,994, Roughly 26% LESS than an even lower clocked model than the i5 4440  with an overclock on the Xeon higher than even your 4.4Ghz number... 

TLDR on Maths; X5450 @4.8Ghz, Still worst than an underclocked i5 4440 by 25~26% in single core performance. 

Iff 4.8Ghz X5450 or C2Q Q6600 in cinebench R10 single thread= 5,556 (In theory)

Stock FX 8320 and 8150 (3.5 and 3.6Ghz Respectively) are roughly= 4,000

(Source: http://www.anandtech.com/bench/product/698?vs=434 )

Stock 8350 and OC'ed 8320E (4Ghz and 4.8Ghz Respectively) are roughly=4,338 and 5,107... 

(Source: http://www.anandtech.com/bench/product/1403?vs=697)

Also I'd like to point out that all the FX 83XX should have identical IPC (instructions per clock) but if you do the math they don't... 

FX 8320   Score 3,987 / Clock 3.5Ghz = IPC 1,140

FX 8350   Score 4,338 / Clock 4.0Ghz = IPC 1,084

FX 8320E Score 5,107 / Clock 4.8Ghz = IPC 1,063

FX 8320E Score 4,056 / Clock 3.2Ghz = IPC 1,267

http://www.anandtech.com/bench/product/1402?vs=1403 and http://www.anandtech.com/bench/product/698?vs=697 is my sources plus a calculator with small rounding for the math portion

Also notice IPC seemingly goes down as clock rate goes up, infact the "IPC" difference between the 8320Es (which IS the exact same chip anandtech just benched it at stock and OC'ed)  is ~20%   with a 50% clock increase so there's no way the doubling clock of the Q6600 I did earlier would double the Score...  clock rates the higher they go just have diminishing returns which is why FX9590 doesn't beat 8320 or 50 by much... taking that into account an insanely overclocked Q6600 or Xeon x5450 would be about even slightly better in single core performance with 8350 at stock, and you'd be missing some very consumer centric features like possibly DDR3, Sata 6Gb/s possibly even 3Gb/s, and PCIe 2.0 possibly even 16x Pcie, lots of variables based on motherboards and chipsets  SLI support ect ect... Also do remember the difference in core count 4 vs 8... But in the end FX vs Xeon X54-- with a decent OC is sadly fairly close...

 

All that aside, I think FX vs Old Xeons x54-- being OC'ed (which on it's own presents some challenges) being fairly close in terms of sheer single core strength and in multicore FX should win,  and in feature set also should win... and Nahalem faired better than I thought after doing some quick searches which in my opinion only further proves my earlier statements in this thread that OP should just try to go with a nahalem or westmere setup because they're the best solution out of the 3 choices here (FX, Xeon X54--, or i7 8/9--/Xeon X55/56--) in terms of price to performance and overall performance, and at the moment in terms of the used PC market I really hope this means in another year or 2 Sandy bridge gets the pricing/availibility Nahalem/Westmeres showing now considering Sandy bridge was where IPC had it's big jump

The Xeons I'm talking about are old ones from 2008. And they can not safely hit 4.8GHz without some serious cooling and luck in the silicon lotterey. They stop at about 4.4GHz which is at about the performance of an i5 4440. I know that a Xeon X5450 at 4.4GHz (and indeed any 45nm quad core from Intel) comes very close to matching an i5 4440 in performance as I actually own both of them and have run the benchmarks. BTW for gaming, a 4GHz FX 8350 is worse than a 4GHz Intel 45nm Core 2 Quad (and their large L2 Cache helps with that).

[GloriousPain]

The Xeons I'm talking about are old ones from 2008. And they can not safely hit 4.8GHz without some serious cooling and luck in the silicon lotterey. They stop at about 4.4GHz which is at about the performance of an i5 4440. I know that a Xeon X5450 at 4.4GHz (and indeed any 45nm quad core from Intel) comes very close to matching an i5 4440 in performance as I actually own both of them and have run the benchmarks. BTW for gaming, a 4GHz FX 8350 is worse than a 4GHz Intel 45nm Core 2 Quad (and their large L2 Cache helps with that).

Can you do some benchmarks real quick? I'd love to see them like cinebench, in specific, especially since as I said mathematically even a theoretically perfect scaled 4.8Ghz Xeon X5450 shouldn't be close to a 2.9Ghz i5 4570S (unless you see 26% slower as close) and yeah I did say in the end the FX is slower, but not by as much as I feel you were and are stating, considering you are stating 4.4Ghz x5450 to be close to a 3.1Ghz i5 4440, when based on theoretical math a 4.8Ghz Xeon X5450 is ~26% slower than an 2.9Ghz i5 4570S... which between a clock for clock example with the FX there's a ~9% difference (based on the 4.8Ghz FX 8320E vs the theoretical 4.8Ghz Xeon X5450 in cinebench R10 single threaded benchmarks by anandtech) meaning the OC'ed  X5450 is almost 3 times closer to the FX (also OC'ed but to the same clockrate) than the UNDERCLOCKED haswell i5 4570S... and that's without stating the advantages the FX has over the Xeon...  

Basically my point is you're pissing me off and I feel wording badly the facts, by saying a ~26% difference in a thing STACKED in it's favor is "very close" when something that's ~9% off in a fair comparison is simply "worse"