NVIDIA’s Annual GPU Rebadge Begins: GeForce GT 730M and GeForce 710M Partial Specs Published
by Ryan Smith on January 8, 2013 2:00 PM ESTAs we discussed yesterday with AMD’s latest round of GPU rebadges, both AMD and NVIDIA are locked into playing the OEM rebadge game in order to fulfill their OEM partner’s calendar driven schedules. OEMs want to do yearly updates (regardless of where the technical product cycle really is), so when the calendar doesn’t line up with the technology this is achieved through rebadges of existing products. In turn these OEMs put pressure on component suppliers to rebadge too, so that when consumers compare the specs of this year’s “new” model to last year’s model the former look newer. The end result is that both AMD and NVIDIA need to play this game or find themselves locked out of the OEM market.
In any case, the bulk of these rebadges coincide with CES, which is where the OEMs announce their calendar-refresh products. We often see the specs for these systems leak out a couple of months in advance – and accordingly see the product numbers for the rebadged components they contain – but it’s not until CES that AMD and NVIDIA publish the specs of these products. So we’ve known these products were coming, we just haven’t had any solid details about them until now.
Jumping right into things, this morning NVIDIA updated their GeForce product page with a link to a PDF with the specifications for two new mobile products: GeForce GT 730M and GeForce 710M. NVIDIA’s PDF doesn’t go into great detail – in particular they aren’t listing the clockspeeds at this time – but from the specs provided we can divine some more information about these first members of the 700M family.
NVIDIA GeForce 700M Series GPU Specification Comparison | ||||
NVIDIA GeForce GT 730M | NVIDIA GeForce 710M | |||
Was | 640M? | 620M? | ||
Stream Processors | 384? | 96? | ||
Texture Units | 32? | 16? | ||
ROPs | 16? | 4? | ||
Core Clock | ? | ? | ||
Boost Clock | ? | ? | ||
Memory Clock | ? GDDR5 / DDR3 | ? DDR3 | ||
Memory Bus Width | 128-bit? | 128-bit? | ||
VRAM | ? | ? | ||
Transistor Count | 1.17B | 585M | ||
GPU | GK107 | GF117 | ||
Manufacturing Process | TSMC 28nm | TSMC 28nm | ||
Architecture | Kepler | Fermi |
For the time being we have a few unknowns and a few assumptions on our parts, but based on NVIDIA’s specs and naming scheme we are very confident about which GPUs are actually behind these products. The 730M is some kind of GK107 rebadge – almost certainly 640M – as evidenced by its support for Kepler family features such as TXAA, PCIe 3.0, and DisplayPort 1.2. We beiieve this to be 640M in particular based on the use of both GDDR5 and DDR3 and the product name, though a 650M rebadge is also a possibility due to the very similar features of those parts.
The other part on NVIDIA’s current 700M series list is the GeForce 710M. This we believe to be a GF117 rebadge – almost certainly 620M – as evidenced by its lack of support for Kepler family features such as PCIe 3.0, TXAA, or support for resolutions over 2560x1600. This means that yes, just like the 600M series, the 700M series will contain some last-generation Fermi parts too, so any hope of a unified mobile family have been dashed by this product. Like the 620M this is a DDR3-only part, and exists as NVIDIA’s entry-level part over Intel’s iGPUs.
It’s interesting to note that in lieu of clockspeeds (or really any other hard details) NVIDIA is listing something called the “GeForce Performance Score”, which is defined as the performance of the part relative to Intel’s HD4000 iGPU. The 730M and 710M are 4.8x and 3.0x respectively, and while NVIDIA is almost certainly being overgenerous in their performance estimations here, it does lend further proof to these being GK107 and GF117 rebadges.
Finally, although NVIDIA has only published information on the 730M and 710M so far, based on previous experience we believe that this is only the tip of the iceberg. In the coming weeks (if not days) we would expect to see more mobile rebadges, along with some kind of desktop rebadge. We’ll keep our eyes peeled, so until they stay tuned.
Source: NVIDIA GeForce Product Page
25 Comments
View All Comments
silverblue - Wednesday, January 9, 2013 - link
One of the other clever stunts they try to pull is adding far more memory than the card can realistically use in the hopes that people will think more RAM = better.As it stands, NV and AMD are generally as bad as each other... but if you know what you're looking for, you really shouldn't be buying the wrong product anyway.
CeriseCogburn - Wednesday, January 9, 2013 - link
LOLThe amd fanboy true believer, still.
Yes if only others, especially nVidia realized how many of you there are !
ROFL
JonnyDough - Tuesday, January 8, 2013 - link
Ugh. There's a reason I quite buying Nvidia products after the 8800 series.TheJian - Tuesday, January 8, 2013 - link
Didn't you just post the same crap already once? Not enough to make ignorant/biased statements once?Work for AMD?...LOL. Let me know when AMD stops doing it, then you have a leg to stand on saying this junk.
"Ugh. There's a reason I quite buying Nvidia/AMD products."
Fixed it for ya...ROFL
silverblue - Wednesday, January 9, 2013 - link
Homeles had the right idea; if you're savvy enough to research the parts then you'll never get stung. I'd prefer the retail products to be a single generation instead of two (or even three), however I guess we don't always get what we want.CeriseCogburn - Wednesday, January 9, 2013 - link
Maybe you're just incompetent. Have a bad memory. Illiterate. A control freak.Wanting to limit others choices.
Lazy. An amd fanboy.
Not savvy.
GOOD GOD
" I'd prefer the retail products to be a single generation instead of two (or even three) "
The Egg has an AMD HD2000 card on sale there right now. ROFL
silverblue - Wednesday, January 9, 2013 - link
And what has that got to do with anything?CeriseCogburn - Wednesday, January 9, 2013 - link
One more thing there before you run off into I sure am smart fantasy land.... boy we gotsa help those poor braindead average consumers...If the companies did what you wanted, a single generation, then we wouldn't have lower priced items for people like penny pinching crybaby amd fanboys.
Maybe you should become CEO and then you can throw away your last generation GPU cores in the trash, since the production has moved on from the former lines...
Hey then we can pay a lot more for all the systems coming out. Great idea there mr I'd prefer.
silverblue - Wednesday, January 9, 2013 - link
Actually, I defended the reuse of old parts to a certain point the other day. What I took exception to was parts being identically named. Where's that comment you made about illiteracy, again?Also, I said I'd PREFER one generation, not that it shouldn't happen and not that I'd want to control how a company releases its products.
Sponging like a lemming without thinking indeed. The only reason I reply to you is for the inevitable lulz I'll have when you dribble all over the keyboard in response.
CeriseCogburn - Friday, January 11, 2013 - link
What you prefer is being stupid, getting caught, then changing everything you said.