• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

TCCD vs. UTT results...have a look!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

NinjaZX6R

RAM Junkie
Joined
Nov 19, 2002
Location
In slots 2 & 4!
Hey everyone. I have been seeing much controversy over which ram is superior on an A64 platform. I still cannot choose, but I would like to post my results which each one so others can see. I did a few basic tests that lots of us run on our machines. I did not go into extensive comparison, but hopefully my results will show something productive. BTW, I am not taking sids here :)

TEST SETUP

Both setups were run at 275x9, so the cpu frequency did not change. The UTT, however, was run on a divider of 166/200. The TCCD was run at 1:1

Patriot XBL 2x512mb @275mhz 2.5-4-4-7-1t (2.6v)

TwinMos Speed Premium 2x512mb @ 225mhz 2-2-2-10-1t (3.3v)

Super PI

TCCD - 34 seconds
UTT - 34 seconds

Sandra Memory Bandwidth

TCCD - 7262/7160
UTT - 6728/6654

3d mark 2001se

TCCD - 24,727
UTT - 25,121

AquaMark

TCCD - 70,888
UTT - 71,697

PC Mark 2002

TCCD - CPU:8,034
Memory: 13,774
HDD: 697
UTT - CPU: 8124
Memory: 12,436
HDD: 689



Those are all the tests I ran. If anyone would like other tests run, I wouldn't mind doing them. I think it is pretty interesting that at 50mhz less, even on a memory divider, the UTT was able to come out on top in both 3d mark benchmarks. Granted, it took way more voltage via a ddr booster, but it is interesting none-the-less. Thanks for reading :)

-Collin-
 
Last edited:
High memory bandwidth for the TCCD on Sandra only because it's running 1:1. I personally like the "gaming" performance benchmarks.
 
Could you do a PCMark02 CPU test comparison, and a PCMark02 memory bandwidth test comparison?

They are primitive tests, but a free benchmark, and one that is a fairly good performance indication.

Also, would it be possible for you to do an Everest comparison?


Thanks for posting this :).
 
felinusz said:
Could you do a PCMark02 CPU test comparison, and a PCMark02 memory bandwidth test comparison?

They are primitive tests, but a free benchmark, and one that is a fairly good performance indication.

Also, would it be possible for you to do an Everest comparison?


Thanks for posting this :).

Give me a couple hours, but sure. Might be later tonight. I have pc mark 2002, so that'll be easy.

-Collin-
 
Can you use a 183 divider for the UTT so as to get your memory up to 248MHz?

Even at 225 we see it laying the smack. Once you get into the 240-260 range (which isn't hard) UTT/BH5/BH6/CH5 is virtually unstoppable, to this day.
 
Gautam said:
Can you use a 183 divider for the UTT so as to get your memory up to 248MHz?

Even at 225 we see it laying the smack. Once you get into the 240-260 range (which isn't hard) UTT/BH5/BH6/CH5 is virtually unstoppable, to this day.

My gigabyte board has no 183 divider, which is truly a shame. I wish i could run the HTT around 275 and the ram around 250, that would be sweet. Or, maybe I can try for around 260 1:1, but that will make my clock speed around 2.3ghz. Can't have my cake and eat it too! I will try a few more combinations, but I can see that UTT will end up being my ultimate choice :)

Gautam....Sen tells me you have a somewhat better than usual X800 to say the least, how is it going with that?

-Collin-
 
NinjaZX6R said:
My gigabyte board has no 183 divider, which is truly a shame. I wish i could run the HTT around 275 and the ram around 250, that would be sweet. Or, maybe I can try for around 260 1:1, but that will make my clock speed around 2.3ghz. Can't have my cake and eat it too! I will try a few more combinations, but I can see that UTT will end up being my ultimate choice :)

Gautam....Sen tells me you have a somewhat better than usual X800 to say the least, how is it going with that?

-Collin-
Try A64 tweaker. It may work, it may not. Other option to get CPU/10 is to run the 8x multi with 166, but then you'd need a 310MHz HTT; not sure if that's possible or not, but I've seen it done.

As for the X800...I think the sig speaks for itself. :D
 
Man, with 3.3v you should be able to run that UTT a lot faster than that. Have you tried? This is the same stuff as VX correct?

10 x 250fsb would be nice.
 
mattspalace said:
Man, with 3.3v you should be able to run that UTT a lot faster than that. Have you tried? This is the same stuff as VX correct?

Yes, its the same stuff. But, I cannot go any higher on this board. Well, I could run it 1:1, but cpu frequency will be too low. Oh, and 3.3v is nothing, I had it up to 3.6v just fine. Maybe I should see if it can do 270!!!!!!

-Collin-
 
Wow, I didn't expect the UTT to beat TCCD at just 225FSB. Nice post ninja, now I'm tempted to pick up some Twinmos.

Your timings seem a little loose on the TCCD for just 275FSB. Does your XBL respond to voltages? With 3.2v on mine I can get 235-240FSB 2-2-2-5 .
 
Those are not the same ones that he's talking about. The ones he has are Speed Premium, seen here . The ones in your link do respond well to voltages too though. From what I've read the Speed Premium are 44D and the Non SP ones are 50D. Results in both cases have been similar nonetheless.
 
Very nice test there. Very much appreciated. Looks like my mind is made up on TCCD versus OCZ Gold Vx. I'll take the Vx's. Too bad the TwinMOS can't be had in Canada yet or that would probably be the way I'd go.
 
kiljaden5 said:
is this the memory in question??

http://www.newegg.com/app/viewProductDesc.asp?description=20-218-062&depa=0

if so i'm picking that up asap

ctringali my cousin has just picked up a set of those. They ARE UTT chips and act just like mine do. I could not tell a difference.

In the morning (cause im just too freakin tired right now) I'll post results in pcmark 2002. I don't have the full version, so I cannot just narrow it down to the memory, but I will give scores with each. Stay tuned :)

-Collin-
 
felinusz said:
Could you do a PCMark02 CPU test comparison, and a PCMark02 memory bandwidth test comparison?

They are primitive tests, but a free benchmark, and one that is a fairly good performance indication.

Also, would it be possible for you to do an Everest comparison?


Thanks for posting this :).

Updated with the PC Mark 2002 results. Looks like we have some competition!

-Collin-
 
Very worthwhile comparison and very timely, too, with all these discussions about TCCD being discontinued.

Your results are consistent with others that I've seen, which is basically that while TCCD will give you more bandwidth, that does not always relate to better 3DMark scores. The lower timings are what have always made the Winbond ram so attractive. That has not changed with the UTT.

There are situations where the TCCD is a better match for your rig than others (seeking high fsb or no high-voltage option, for example), but if you have the means to give them juice, the Winbond ICs respond very nicely. I've stated many times that, in my opinion, both of those chips make for nice ram but if I could only keep one, well...I'll just say that I'm not selling my bh-5 anytime soon.
 
Back