• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Throughbred 1700+ from Newegg overclock, wow!

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Hi KIAman:

Is your CPU a retail or OEM version from newegg?
Or it does not matter which version. Thanks.

Thanks for answer my question.
So now I am going to order from newegg.
 
Nice numbers,redly1. It's too bad he didn't list typical currents on there though. Just for kicks, I crunched the numbers.

XP1700 Pali: Voltage - 1.75; Typical power dissapation - 57.4 W; Typical current - 32.8A;
XP1700 T-bred: Voltage - 1.5; Typical power dissapation - 44.9 W; Typical current - 29.93A;

So you're right about the current. Honestly though, isn't this whole exercise is pointless because AMD's data is collected from spec operation of this chip, and KIAman's chip is out of spec? Peace.
 
Originally posted by: JSSheridan
Nice numbers,redly1. It's too bad he didn't list typical currents on there though. Just for kicks, I crunched the numbers.

XP1700 Pali: Voltage - 1.75; Typical power dissapation - 57.4 W; Typical current - 32.8A;
XP1700 T-bred: Voltage - 1.5; Typical power dissapation - 44.9 W; Typical current - 29.93A;

So you're right about the current. Honestly though, isn't this whole exercise is pointless because AMD's data is collected from spec operation of this chip, and KIAman's chip is out of spec? Peace.

Not really pointless, though I can see how current consumption on the T-bred might cross over the Pali somewhere up the OC curve
 
My tbred was OEM from newegg. I dont know if it will make a difference between OEM or RETAIL besides the warranty.

My tbred came with all the L1 and L3 bridges intact and I have all multipliers from 6 to 12.5.

Also, I know for a fact that my tbred is revision "A" which is not as good as the revision "B."

I decided to do some work about the temps and they are still so strange. Im wondering if I applied my HSF wrong or something.
1)Both the pally and tbred show that a 50cfm and 84cfm fan on the slk-800 makes no difference in temps.
2)Tbred is consistantly 5c higher than the pally, even with different heatsinks and fans.
3)Upping the voltage doesnt seem to raise the temp as much as raising the speeds. 11x166=1826@1.5v increases temps about 4c while
11x133=1463mhz@1.7v raises the temps about 1c.

EIEIEIE anyways, overall I am very happy with this chip either way.
 
Originally posted by: pillage2001
Originally posted by: KIAman
I have the slk-800 on a vantec tornado 80mm fan at 84cfm. The thing is I tried 3 different fans.
The 30cfm generic fan was about 8c hotter, but the 50.1cfm AND my vantec 84cfm fan post the exact same results. Oh well, I kept my vantec in there since i dont mind the sound.


I didnt exactly mean "dissapointing" but it took a good .2v jump to make 100mhz more. Also keeping it at 2600+ speeds would have been awesome, but temps were so high just idling.

Another thing is, this chip seems hotter than my AROIA. It is about a consisten 5c higher idle/load at comparable speeds AND that includes the fact the throughbred is running at a much lower vcore. Wierd huh?

The core surface is relatively small. The heat could not dissipate from the core at a fast rate. Less surface contact means higher temps.

^ What he said. There's less heat being produced, but the reduction in surface area is proportionately greater than the reduction in heat generation, so the temperature is going to be higher. Just the way things are. 🙂
 
Schweet overclock! 🙂 Watch how high you push that Vcore though, @.13micron I don't expect that the T-breds will fair much better than the Northwoods when exposed to high voltage continuously, and you may damage your CPU.
 
From Kiaman:
My tbred was OEM from newegg. I dont know if it will make a difference between OEM or RETAIL besides the warranty.

Retail & Oem are no different. The only different is warranty, Oem is 30days & Retail is 3 years. There is no different for me, becasuse this is only less than $70.00, if this is broken & I will up grade to higher CPU like 2000 xp or 2100xp. I know Newegg has a good RMA policy but still have to up grade.🙂
 
My tbred came with all the L1 and L3 bridges intact and I have all multipliers from 6 to 12.5.

Ok, finally got my XP1700+, AIUGA 0245, from newegg today. As far as I can see, all bridges, L1 and L3, are seperated.
 
I am a newbie at this overclocking talk, anyway. What is the stepping of the processer, and what does Thoroughbred mean? What are the benifits of having a smaller die size? I also saw a XP 1800 Thoroughbred, wouldnt this be better than a XP 1700, even if not that much.
 
You questions are totally off track; next time better start a thread of your own.

stepping = the batch (date & location) that the processor was manufactured, since processors from the same batch tend to "behave" similarly, stepping will give people a general ideal of how good one processor is.

Thoroughbred is the code name AMD gave to their newest Athlon core; some older core names are Palomino and Thunderbird.

Smaller die size = can pack more transistors/ use less space, use less raw material (reduce cost), the end product will consume less power, since smaller, each transistor latacy will be smaller (run faster),... well of course you will need to take some basic courses like digital logic in order to understand what transistors/logical gates are about.

"Wouldnt this be better"-- now you tell me, what the hell does it mean "better". Is an orange better than an apple? Or is a $2 apple better than a $1 apple when all I need is an apple for target shotting?
 
Originally posted by: KIAman
I have the slk-800 on a vantec tornado 80mm fan at 84cfm. The thing is I tried 3 different fans.
The 30cfm generic fan was about 8c hotter, but the 50.1cfm AND my vantec 84cfm fan post the exact same results. Oh well, I kept my vantec in there since i dont mind the sound.


I didnt exactly mean "dissapointing" but it took a good .2v jump to make 100mhz more. Also keeping it at 2600+ speeds would have been awesome, but temps were so high just idling.

Another thing is, this chip seems hotter than my AROIA. It is about a consisten 5c higher idle/load at comparable speeds AND that includes the fact the throughbred is running at a much lower vcore. Wierd huh?

The heatsink itself is already $45 on newegg!!!!
My 1700+ TBred is only $58!!!
Can't afford it.
 
Did some testing...
My bad, the CPU is unlocked. I can adjust multiplier through the BIOS.

MB: Epox 8k5a3+
HS: Vocanno 7+
RAM: XMS PC3200
Video: Albatron GF4 Ti4200 128M Turbo
PS: Enermax 431W
CF: 4x 75CFM 80MM

Max clock I can finish 3dmark so far is 175x11@1925.
Voltage and temp is low, 1.65v vcore and cpu 27c system 19c.
 
Originally posted by: Daovonnaex
Originally posted by: chinoloco
Quick question, are the XP Tbreds either revision A or B MP enabled?
All Thorougbreds are MP enabled (unless I'm terribly wrong...).
can anyone confirm this, or can they at least be mp enabled by bridging the l5 like on the palominos? if so i may pick up a couple today and see what kind of oc i can get in a dually.

 
Well, mine is not as good as I expected.

After more testing and I found:

1. While 1.65v is the optimal, 1.75v is the max, setting higher will actually decrease the oc'ing ability.
2. When CPU clock goes up, GPU clock must goes down (I have no idea why). Within ~1.85g I can push my GF4 4200 all the way to 315/690. Once clocked to ~1.9g, it wouldn't even do 300/600, crashed right away in 3dmark.
3. Once it crashed in 3dmark, it would never re-boot again???
4. Setting memory timing too aggressive and hd is screwed. Have to reinstall windows.
 
Newbie, you have to lower your clock speed on the GPU because as u raise your FSB, the AGP/PCI bus isn't locked. So 166/5 = 33; but 18x/5 = around 36-37 for PCI, and for AGP its running outta spec. On my motherboard, it says the ratio is 5:2:1, so at 166/5 * 2 it would be 66 (spec), but at 18x it would be around 75ish, pretty high outta spec. Depending on your cards sensitivity, it could handle different AGP buses. When the card is running overclocked, its even more sensitive to the out of spec buses.
 
Once it crashed in 3dmark, it would never re-boot again???

Yah I noticed this a few times...i think its the mobo. After a few overclocking mishaps I could not get the system to POST and had to make several tries before it would. Quite hair-raising.
 
Just curious newbie you didnt mention your stepping for your t-bred.

The same as yours, AIUGA 0245.

BigJ2078,

The situation is not that simple. Since my TBred is unlocked, I have tried literately all kinds of oc'ing combination. Which is even though I used 11.5x167@1920MHz (33PCI/66AGP), I still couldn't clock my GF4 any higher than 280?
 
w00t w00t.. my 1700+ came in last night... put it into the chaintech 7NJS board (I know it doesn't have a multiplier option in bios... but they're putting it in with new update ^^). I was able to get 166fsb on default core... now I'm on 1.7volt and running @ 180fsb(2.0ghz) =) . I'm using a Swiftech 462 and idle is 49C and full is 53C, right now til I get my 120mm and adapter to slap onto it.

Specs
Chaintech 7NJS nForce2
XP 1700+ TBred (aiuga 0245) from Newegg
512MB Single stick Samsung Original Pc-2700
Chaintech GF4 4200 Sp Ed.
WD 80gig 8mb

^^, I can't wait for the new bios update =), I'm gonna run some 200fsb x 10 hehe... and samsung probably makes the best low priced oc'ing ram ever =)
*PCMark thinks it's a Duron =S... lol <shrug> I don't care.. as long is it passes prime and I've had UD running 24/7 =). w00t w00t again =)
Join UD today and support cancer research!! http://www.ud.com/home.htm
 
Currently running mine at 1.933 (A stepping at 1.65V (actual 1.696). Not yet fully stressed. A7N8X. Looped 3dMark2001 for a bit, Ran the benches in UT2003 - no problems so far. Prime 95 ran for 1 hour before I stopped it.
 
Back
Top