Radeon HD 9000 Series Arrives This October: Report

[Casecutter]

3870->4870->5870 Ah good days.

3870->4870 had even held to the same 55nm process but it really was like 36% increase. But that I believe was more what GDDR5 brought, AMD even surprised themselves... The good old days when those who first jumped on the GTX260 found their AIB’s feeling obligated to print reimbursement checks!

Historically it has been up and down. 80% would be fantastic for sure, but it won't happen and it can't happen.

True, one might figure CuracaoXL offering about 25% above a Tahiti Ghz, while holding to a similar if better efficiency. Figure 10% from increase Sp and TMU’s, while there's conceivably something from alleged 50% increase in ROP’s. Weigh in low leakage production (Malta) parts curtail power usage, possibly even while presenting 10% increase in clocks. All that while on the same or close to the same 28nm die size, so conceivably for a $500 MSRP. That places it about par with a 780 working from a chip that is 30-35% smaller than the GK110, for like 28% less money.

But seems it just won't come soon enough... Had AMD had them last May such printers might have ran agian.

 

[Prima.Vera]

3870->4870 had even held to the same 55nm process but it really was like 36% increase. But that I believe was more what GDDR5 brought, AMD even surprised themselves... The good old days when those who first jumped on the GTX260 found their AIB’s feeling obligated to print reimbursement checks!

Dude, my 5870 was a little faster than 2x4870 in CrossFire.
I would love to see the 9970 having the same performance level as two 7970 cards. One can only dream right?

 

[Casecutter]

Dude, my 5870 was a little faster than 2x4870 in CrossFire.

Well a 4870-> 5870 saw a 39% increase @1920x (see W1zzards' release review; Sept) and that was an actual die shrink 55->40Nm, while in that same review the 4870 X2 was shown to be 4% above the 5870. Included in that you also should consider some release driver (8.66 RC6) bump.

While even some 7 month later in the MSI 5870 Lighting review, the 4870 X2 still best that OC Lightening by 5%. Some difference in test set-up like Win7 64-Bit and matured drivers.

I would love to see the 9970 having the same performance level as two 7970 cards. One can only dream right?

We're not going to see anything like 39% in just a re-spin on 28Nm. I figure 25% in hardware, what they find in new release driver who knows. If they gave us 7990 performance which is 22% more than a Titan @2650x, while holding to the alleged 250W... in a $550 card. Then where's that leave Volcanic Island 4-5 months ?

 

[Prima.Vera]

Well a 4870-> 5870 saw a 39% increase @1920x (see W1zzards' release review; Sept) and that was an actual die shrink 55->40Nm, while in that same review the 4870 X2 was shown to be 4% above the 5870. Included in that you also should consider some release driver (8.66 RC6) bump.

I donno man, check those site with the reviews. There are completely different numbers than the ones you said.:
http://www.guru3d.com/articles_pages/radeon_hd_5870_review_test,15.html
http://www.tomshardware.com/reviews/radeon-hd-5870,2422-13.html

I agree tho, they were almost the same performance, to bad nowadays you will never see a next generation top card to have the performance of 2 top previews generation cards... Did they got lazy or what??:shadedshu

 

[xvi]

Dude, my 5870 was a little faster than 2x4870 in CrossFire.
I would love to see the 9970 having the same performance level as two 7970 cards. One can only dream right?

Did the same here and saw about the same performance for half the wattage. Just replaced the 5870 with two 6950s though. Not sure why I did that.


I wonder how much my old Radeon 9800 Pro would go for on eBay when these new cards drop.

 

[Casecutter]

you will never see a next generation top card to have the performance of 2 top previews generation cards

Well this isn't truly Next Gen… that normally entails a significant change in architecture and on a die shrink (ie Volcanic Island). This is a re-spin improvement of the GCN and 28Nm process.

I say we'd be lucky if AMD further enhances this re-spin with a move from the TSMC "HPL" (lower power) process, and use the more performance driven "HP" process. There had been speculation that AMD went with the HPL process to save on power (especially for Tahiti), and because TSMC HP process was having trouble... It wasn't showing ready for prime-time when AMD wanted to start 28Nm production about the beginning of the 4th quarter of 2011 (though TSMC kept a brave/poker face in the press). Although, it was later found the problem wasn't process, but production issues that TSMC finally fixed with end Jan-Feb shut-down. Such problems effected pretty much everyone's 28Nm node to one degree or another. It's why Nvidia held out with GK104 production and didn't release the GTX680 until mid-March, while any real volumes didn't come till mid-end of April.

Hard to know if AMD persistent with the "HPL" process after the TSMC production fix, but I think they did and just released GHz models. That's why see even with introduction of Boost, they still had an un-nerving 25% power usage jump for only 10% performance.

So where does that leave us... unless AMD found a good amount efficiencies in this architectural re-spin they might not go to the HP process. If they found enough ways to re-vamp architecturally power consumption could they go with the HP process and pick-up the performance benefits, while not giving too much of a compromise to higher power?

 

[makwy2]

Looking forward to the next step in the ATI/AMD - Nvdia war. Can AMD gain market share? Let's find out!