AMD's Radeon HD 6790: Coming Up Short At $150
by Ryan Smith on April 5, 2011 12:01 AM ESTThe last couple of weeks after the recent GeForce GTX 550 Ti launch have been more eventful than I had initially been expecting. As you may recall the GTX 550 Ti launched at $150, a price tag too high for its sub-6850 performance. I’m not sure in what order things happened – whether it was a price change or a competitive card that came first – but GTX 550 Ti prices have finally come down for some of the cards. The average price of the cheaper cards is now around $130, a more fitting price given the card’s performance.
The timing for this leads into today’s launch. AMD is launching a new card, the Radeon HD 6790, at that same $150 price point. Based on the same Barts GPU that powers the Radeon HD 6800 series, this is AMD’s customary 3rd tier product that we’ve come to expect after the 4830 and 5830. As we’ll see NVIDIA had good reason to drop the price on the GTX 550 if they didn’t already, but at the same time AMD must still deal with the rest of the competition: NVIDIA’s GTX 460 lineup, and of course AMD itself. So just how well does the 6790 stack up in the crowded $150 price segment? Let’s find out.
| AMD Radeon HD 6870 | AMD Radeon HD 6850 | AMD Radeon HD 5830 | AMD Radeon HD 6790 | AMD Radeon HD 5770 | |
| Stream Processors | 1120 | 960 | 1120 | 800 | 800 |
| Texture Units | 56 | 48 | 56 | 40 | 40 |
| ROPs | 32 | 32 | 16 | 16 | 16 |
| Core Clock | 900MHz | 775MHz | 800MHz | 840MHz | 850MHz |
| Memory Clock | 1.05GHz (4.2GHz data rate) GDDR5 | 1GHz (4GHz data rate) GDDR5 | 1GHz (4GHz data rate) GDDR5 | 1050MHz (4.2GHz data rate) GDDR5 | 1.2GHz (4.8GHz data rate) GDDR5 |
| Memory Bus Width | 256-bit | 256-bit | 256-bit | 256-bit | 128-bit |
| VRAM | 1GB | 1GB | 1GB | 1GB | 1GB |
| FP64 | N/A | N/A | 1/5 | N/A | N/A |
| Transistor Count | 1.7B | 1.7B | 2.15B | 1.7B | 956M |
| Manufacturing Process | TSMC 40nm | TSMC 40nm | TSMC 40nm | TSMC 40nm | TSMC 40nm |
| Price Point | ~$200 | ~$160 | N/A | $149 | ~$110 |
3rd tier products didn’t get a great reputation last year. AMD and NVIDIA both launched such products based on their high-end GPUs – Cypress and GF100 respectively – and the resulting Radeon HD 5830 and GeForce GTX 465 were eventually eclipsed by the GeForce GTX 460 that was cooler, quieter, and better performing at the same if not lower price. The problem with 3rd tier products is that they’re difficult to balance; 1st tier products are fully enabled parts that are the performance kings, and 2nd tier products are the budget minded parts that trade some performance for lower power consumption and all that follows.
While 2nd tier products are largely composed of salvaged GPUs that couldn’t make it as a 1st tier product, the lower power requirements and prices make the resulting video card a solid product. But where do 3rd tier products come from? It’s everything that couldn’t pass muster as a 2nd tier product – more damaged units functional units that won’t operate at lower voltages like a 2nd tier product. The GTX 465 and Radeon HD 5830 embodied this with power consumption of a 1st tier card and the performance of a last generation card, which made them difficult to recommend. This does not mean that a 3rd tier card can’t be good – the Radeon HD 4830 and GTX 260 C216 were fairly well received – but it’s a difficult hurdle to overcome.
Launching today is the Radeon HD 6790, the 3rd tier Barts part and like the rest of the Barts-based lineup, the direct descendent of its 5800 series counterpart, in this case the Radeon HD 5830. As is to be expected, the 6790 is further cut-down from the 6850, losing 2 SIMD units and half of its ROPs; mitigating this some are higher clockspeeds for both the core and the memory. With 800 SPs and 16 ROPs operating at 840MHz, on paper the 6790 looks a lot like a Radeon HD 5770 with a 256bit bus, albeit one that’s clocked slower given the 6790’s 1050MHz (4.2GHz data rate) memory clock.
From the 5830 we learned that losing the ROPs hurts far more than the SPs, and we’re expecting much of the same here; total pixel pushing power is halved, and MSAA performance also takes a dive in this situation. Overall the 6790 has 90% of the shading/texturing, 54% of the ROP capacity, half the L2 cache, and 105% of the memory bandwidth of the 6850. Or to compare it to the 5770, it has 98% of the shading/texturing capacity, 98% of the ROP capacity, and 175% of the memory bandwidth, not accounting for the architectural differences between Barts and Juniper.
Further extending the 5830 comparison, as with the 5830 AMD is leaving the design of the card in the hands of their partners. The card being sampled to the press is based on the 6870’s cooler and PCB, as the 6790’s 150W TDP is almost identical to the 151W TDP of the 6870, however like the 5830 no one will be shipping a card using this design. Instead all of AMD’s partners will be using their own in-house designs, so we’ll be seeing a variety of coolers and PCBs in use. Accordingly while we can still take a look at the performance of the card, our power, temperature, and noise data will not match any retail card – power consumption should be very close however.
At 150W AMD is skirting the requirement for 2 PCIe power sockets. Being based on a 6870 our sample uses 2 sockets and any other design using a 6870 PCB verbatim should be similar, but some cards will ship with only a single socket. This doesn’t impact the power requirements of the card – it’s roughly 150W either way – but it makes the card more compatible with lower-wattage PSUs that only come with 1 PCIe power plug.
As we mentioned previously, AMD is launching the 6790 at $150. With the GTX 550’s price drop its direct competitor is no longer the GTX 550, but rather the closest competitor is now cheap GTX 460 768MB cards, which on average are about the same $150. AMD’s internal competition is the 6850, which averages closer to $160. Technically the Radeon HD 5770 is also competition, but with it going for around $110 after rebate, it’s far more value priced than the 6790 is.
Meanwhile the 6790 name also marks the first time we’ve seen the 6700 series in the retail market. In the OEM market AMD has rebadged the 5700 series as the 6700 series, however that change won’t ever be coming to the retail market, making this the only 6700 series card we’ll see. It’s a bit odd to see one series shared by two GPUs so significantly different, but AMD bases this on the fact that the 5770/6770 and the 6790 are so close in terms of specs; they want to frame the 6790 in terms of the 5770/6770, rather than in terms of the 6800 series. If nothing else it’s a nice correction for the poor naming of the 6800 series; a 6830 would have been the 5830 but slower.
| April 2011 Video Card MSRPs | ||
| NVIDIA | Price | AMD |
| $700 | Radeon HD 6990 | |
| $480 | ||
| $320 | Radeon HD 6970 | |
| $240 | Radeon HD 6950 1GB | |
| $200 | Radeon HD 6870 | |
| $160 | Radeon HD 6850 | |
| $150 | Radeon HD 6790 | |
| $130 | ||
|
|
$110 | Radeon HD 5770 |
Facebook
Twitter
RSS
69 Comments
View All Comments
geniekid - Tuesday, April 5, 2011 - link
Tom's used a higher end card paired with the 6790 to test the Crossfire performance of this thing. The results suggest there might be some value to this card if used in Xfire configuration compared to single cards around the same price. It would be nice to explore that possibility!BPB - Tuesday, April 5, 2011 - link
I have been happily running with two HD4850's for a few years now, and want to upgrade. It seems to me that if I stick to a 24" 1920x1200 monitor practically any card will do. Still, I don't like the idea of getting a 6800 series since it's practically 2+ year old technology. Wondering if I should go 6900 series, or wait till 7000 series. Come AMD, man up, put some real upgrades out there that make it easy for me to decide.richardginn - Tuesday, April 5, 2011 - link
it is nice to see a review the AMD 6790 video card, but how about a review of the AMD 6450, AMD 6550, and AMD 6670 OEM video cards???jabber - Tuesday, April 5, 2011 - link
Arent they just the 5XXX cards re-branded with a 6? If so a waste of time.Plus as they are OEM we wont be buying them.
richardginn - Wednesday, April 6, 2011 - link
Actually no. These OEM cards are very different.The 6450 video card which I have only seen sold as an option at HP.com is supposed to be like twice as fast as the 5450 based on the specs listed from the AMD website.
If you are talking just a rebranded video card you have to be talking about the OEM 6770 and 6750 video cards which have no performance boost in the FPS area.
Will these cards go off OEM status when Bulldozer CPU'S are released or just move on to something like a 7450 video card????
Ryan Smith - Tuesday, April 5, 2011 - link
Funny you should mention that...BoFox - Tuesday, April 5, 2011 - link
Is the memory bus really 128-bit instead of 256 bits wide? I'm wondering why Anandtech put a lot of effort into checking up on GTX 550 Ti's 192-bit bandwidth with an odd number of chips, but not on either the 5830 or 6790 that claims 256-bit bus while the ROPs are cut in half.We all know that the number of ROPs is tied with the memory bus for a given architecture design. This is why NV's GTX 550 Ti seems much more valid, as it is linear with 24 ROPs.
If we look at the 5830 here:
http://techreport.com/articles.x/18521/5
3D Mark Vantage color fill test is strongly correlated with the memory bandwidth. If the 5830 were 256-bit, it would have had identical bandwidth with 5850. However, the performance shows that it is not the case. It is barely half of 5850's performance, and also much slower than HD 4870 which has only 16 ROPs at a lower clock than that of 5830.
Next, if we look at BeHardware's ultimate scrutiny (just as respectable as Anandtech's examination of GTX 550 Ti): http://www.behardware.com/articles/783-3/preview-r...
We see that the 5830 has far lower FP16 and FP32 GPixel/s writes than not only the 5770 that has a slightly higher fill rate, but also the 4890 to a far greater degree. The test is directly linear to the available bandwidth as the 4890 is so much faster than the 5770, let alone 5830 in that respect.
One more thing is that as with Barts architecture, we should all know that it is based on VLIW4 architecture, not the traditional VLIW5 one. It seems that AMD wanted to save the "thunder" for Cayman's launch by reserving the announcement for what desperately needed as much thunder as possible. Just look at how close 6870's performance is to 5870 while comparing 6870's 1120sp and 4.2Gbps bandwidth to 5870's 1600sp and 4.8Gbps bandwidth.
Hope you guys enjoyed a little bit of exposure! Is AMD deliberately giving us wrong information? That's not my problem, but if I were the one reviewing the product, I would definitely point these things out in my article.
Ryan Smith - Tuesday, April 5, 2011 - link
From a graphics point of view it's not possible to separate the performance of the ROPs from memory bandwidth. Color fill, etc are equally impacted by both. To analyze bandwidth you'd have to work from a compute point of view. However with that said I don't have any reason to believe AMD doesn't have a 256-bit; achieving identical performance with half the L2 cache will be harder though.And Barts is VLIW5, not VLIW4. Only Cayman is VLIW4.
BoFox - Tuesday, April 5, 2011 - link
Barts is also Northern Islands--the keynote of the architecture being VLIW4.See, the 6790 wouldn't come within 2-3% of 5830 according to all of the benchmarks at the review here.
If it were VLIW5 like the 5830, the 6790 would've been MUCH slower (or the 5830 much faster).
This is because HD 5830 has 1120sp, which is 40% more than 6790's 800sp. It also has 56 TMU's, which is 40% more than 6790's 40 TMU's.
All of the other specs are shockingly similar, with only 5% difference in core and memory clock speeds. Both the 5830 and 6790 have the same "alleged 256-bit bus".
In spite of the whopping 40% shader and TMU difference, the 6790 comes SO close to the 5830--close enough that it's only possible if the 800sp were VLIW4 (multiply it by 5/4 and you get performance like 1000sp). It would only make sense there, as 1000sp is about 10% less than 1120sp, but with 5% higher clock, it comes to within 2-3% of 5830's performance.
If not for VLIW4, what would it be? Tell me.
BoFox - Tuesday, April 5, 2011 - link
Typo: I forgot to add "if it weren't for VLIW4" before the second sentence above. Sorry if it's confusing.Shouldn't we all already know that Barts XT was of the Northern Islands VLIW4 architecture from how close it was to the 5870 (1120sp vs 1600sp which is 43% higher)? Even after adjusting for the clock differences, the shader/TMU operations per second is still 35% higher for the 5870, yet the 5870 turned out to be only 9% faster overall. It would've made perfect sense if Barts XT had 1400 VLIW5 shaders (using the 5:4 ratio over 1120sp).
Using the math: 1600sp x 850Mhz is ... 8% faster than 1400sp x 900MHz. The memory bandwidth does not affect the performance too much since the proportion of bandwidth to GPU muscle is not changed by much. So, it's really close to the overall 9% actual performance difference between 5870 and 6870. Is that coincidental? The reason the difference is actually greater than the calculation is because 1120sp VLIW4 does not exactly translate to 1400sp VLIW5. VLIW4 is not perfectly efficient as to scale 100% at a 5/4 ratio, but it's pretty close.
What else would it be, sincerely?