Bit-tech best hardware of 2010

[scaryjim]

watercooled: yeah, I'd be happy with calling that near enough 10kppd on that type of work unit (which I assume is a 500k) - be aware that you're likely to get less points than that with other types of work units, but frankly if you keep folding like that you'll be shooting up the team rankings in no time

CAT: I think the issue with the 6950 1GB is that it was officially announced today but the cards haven't been sampled to review sites by AMD, so it'll be down to vendors to get the cards round. Hexus also don't have a 6950 1GB in their review...

[CAT-THE-FIFTH]

watercooled: yeah, I'd be happy with calling that near enough 10kppd on that type of work unit (which I assume is a 500k) - be aware that you're likely to get less points than that with other types of work units, but frankly if you keep folding like that you'll be shooting up the team rankings in no time

CAT: I think the issue with the 6950 1GB is that it was officially announced today but the cards haven't been sampled to review sites by AMD, so it'll be down to vendors to get the cards round. Hexus also don't have a 6950 1GB in their review...

Maybe review samples of the HD6950 1GB were only given to some websites like Anandtech.

However,Pixmania is already selling the HD6950 1GB.

[watercooled]

watercooled: yeah, I'd be happy with calling that near enough 10kppd on that type of work unit (which I assume is a 500k) - be aware that you're likely to get less points than that with other types of work units, but frankly if you keep folding like that you'll be shooting up the team rankings in no time

CAT: I think the issue with the 6950 1GB is that it was officially announced today but the cards haven't been sampled to review sites by AMD, so it'll be down to vendors to get the cards round. Hexus also don't have a 6950 1GB in their review...

I *think* it's a 500k WU, it says 'Completed x out of 500000 steps'.

[watercooled]

OK, so that's the first WU done.
Should appear on the tables soon.

Edit: http://fah-web.stanford.edu/cgi-bin/...evvEwB%40jwQSX

It's stripped the '@Hexus' bit from my username but I've double checked it's me with the passkey. I'm guessing I only have <500 points because I'm not at 10 WU's yet.

[scaryjim]

I *think* it's a 500k WU, it says 'Completed x out of 500000 steps'.

Yup, that's the one.

It's stripped the '@Hexus' bit from my username but I've double checked it's me with the passkey. I'm guessing I only have <500 points because I'm not at 10 WU's yet.

Yup, that's right. I think a 500K unit with the bonus points is typically worth ~ 2500 points. Since you appear to be getting through them in ~ 7 hours it shouldn't take too long to get through those first 10 units and really start bringing the points home!

Good to have you on board

[watercooled]

Great stuff.
I don't think I'll be able to run it 24/7, 1-2 WU per day is more likely ATM.

And FYI, here is what the benchmark page looked like at 95%:

Min. Time / Frame : 4mn 22s - 10224.66 ppd
Avg. Time / Frame : 4mn 23s - 10166.40 ppd
Cur. Time / Frame : 4mn 25s - 10051.53 ppd
R3F. Time / Frame : 4mn 25s - 10051.53 ppd
Eff. Time / Frame : 4mn 30s - 9773.62 ppd

And the PPD in the bottom right of the window was 10307.54

Hope this is of some use.
Bear in mind the system wasn't otherwise completely idle throughout the calculation, I was checking on apps like CoreTemp, HWMonitor etc and opening up Firefox from time to time to post the benchmark numbers, but I don't think that would have noticeably changed the results.

[Tattysnuc]

That's brilliant.

The reason I asked for the stats over a number is because it depends on the "core" (read: software) that is being used, and the data within the workunit itself.

10k is damned good from a price perspective as my i7 920 @ 3.5Ghz did about 7-8k running on all 8 threads (that was in conjunction with 2 GPU clients which can need CPU resources, but I never bothered/found out how to assign the GPU client to a specific core)

[Tattysnuc]

Great stuff.
I don't think I'll be able to run it 24/7, 1-2 WU per day is more likely ATM.

And FYI, here is what the benchmark page looked like at 95%:



And the PPD in the bottom right of the window was 10307.54

Hope this is of some use.
Bear in mind the system wasn't otherwise completely idle throughout the calculation, I was checking on apps like CoreTemp, HWMonitor etc and opening up Firefox from time to time to post the benchmark numbers, but I don't think that would have noticeably changed the results.

We're hi-jacking this thread as little - Could you pop over to the distributed computer threads and pop your results over there in the future. I'll start up a thread as soon as I've re-plumbed my system and got it back up and running...

[watercooled]

Oops, I got a bit carried away. Sorry, CAT.

Edit: Hey, at least we got some positive benchmark numbers in for Thuban!

[CAT-THE-FIFTH]

Oops, I got a bit carried away. Sorry, CAT.

Edit: Hey, at least we got some positive benchmark numbers in for Thuban!

No problems as you are proving a point. Bit-tech did not bother to test the X6 1055T because they probably knew it was much better value for money than a Core i7 in certain tasks.

It also seems that Bit-tech has kept quiet about the Intel chipset problems so far. It will be interesting to see when they decide to put in on their website.

They already have put a review up today too.

There is a thread on their forums about this problem too.

[CAT-THE-FIFTH]

I like their latest budget build:

http://www.bit-tech.net/hardware/buy...ebruary-2011/2

The reason that the 770-C45 is hard to find is because it has been replaced at the same price point by the 870-C45.

[watercooled]

Oh my word how much bullcrap can they come out with? You should just stop checking, you know what's going to be on there. I just try not to let stuff like that annoy me so much now.

[Tattysnuc]

I like their latest budget build:

http://www.bit-tech.net/hardware/buy...ebruary-2011/2

The reason that the 770-C45 is hard to find is because it has been replaced at the same price point by the 870-C45.

Now I didn't know that, so thank you! Now that you guys are pointing out the Intel bias, it really is blatant.

I was a bit thrown also with their choice of budget PSU - the Corsair CX430 I think they recommended, which I haven't even see a review of. Now I could be wrong, but why are they allegedly not sending them out to review, so how can they recommend something that they haven't reviewed?

Stinks of lazy to me....

[watercooled]

JonnyGURU bought one and reviewed it, it was OK - in line with a budget PSU but it used cheap caps and build quality isn't up to the Corsair standard. So basically they're using their name+reputation to sell rubbish, it's not what the CX400 was. Also CustomPC/Bit-tech's idea of a PSU review is to look at the marketing and check it works for 5 minutes on a PC. They did a test with a load tester + the help of FSP a while back but they only checked voltages and got the wattage of a few models wrong and gave them a bad review when they are in fact not bad units. For example the Arctic Cooling 550 is a 500w model, yeah the name is misleading but it proves they just can't be bothered doing things properly, they tested it as a 550w model but it didn't perform well at that load so gave it a terrible score. The same PSU has been given good reviews by proper reviewers who know what they're doing...

[CAT-THE-FIFTH]

Another great "review" from Bit-tech of an AMD Zacate motherboard:

http://www.bit-tech.net/hardware/mot...eview/comments

No power consumption figures at all supposedly because the reviewer is using a 110V PSU and is not located in the UK.

It seems that they think a G6950 undervolted would have similar power consumption and this is what they have recommended.

[watercooled]

Oh no, I'm not even clicking that one! xD