• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel's Haswell - Leaked Specs (Due Q2 2013)

Based on the assumption Haswell is +10% over Ivy - Would it be worth upgrading from my 1055T to an i5 Haswell at this point, given my CPU is 3 years old and possibly holding back my 7950?

Additionally, could anyone give me any idea in real world terms what sort of performance increase I would receive. For example, if 1055T gives me X FPS in BF3, what would the Haswell give?...something a long those lines :p
 
I don't think Haswell, if its really on 10-15% faster than an Ivy, is going to be THE reason to upgrade. If you want to upgrade for gaming, a Sandy, Ivy, or Haswell are going to be fine.

If you're not really struggling now, I'm not sure there will be many big performance games out before the next consoles hit.

Unless my mobo/cpu dies before the new consoles are out, I'm most likely to see where AMD console optimisation goes, and how steamroller does before deciding if a Haswell or Steamroller is the way to go.

Steamroller is likely to be 28nm, increase clocks, drastically increase IPC, and be better value... if its completely optimised for(due to console tie in) I think people are going to be surprised how good it is. Also personally, I find the IGP a complete waste when I won't need it and as discrete idle power usage drops each gen + the issues apparent when using switching graphics.

If Haswell came with hex cores without the IGP, I think I'd jump all over that the second it was available, another quad core, with an even bigger IGP I don't need whatsoever, with a marginal CPU performance bump :(
 
If AMD does manage to get some of their upcoming CPU's to be competitive with whatever top end Intel has out at the time I predict that we would see much more comparability in the prices from AMD.
If the new AMD GTi spec R chip from AMD matches Intel GTi spec R chip than they would both be priced similar.
 
Hmmm I will wait for reviews and prices to come out before planning the leap. It will be interesting to see how integrated voltage regulators affect overclocking.

Although my current rig is fairly balanced and can handle anything I play, the chipset is beginning to fall behind. Things like:
-Not having S-ATA 3 has held me back from upgrading to the new generations of SSDs.
-No USB 3.0 support cripples my USB3 pen drivers and external HDDs, this will only get worse in the future.
-Lack of PCIe 3 will probably also have an effect in the future. If Haswell lasts me as long as Nehalem did, GPUs 3 or 4 generations down the line are likely to take advantage of the extra bandwidth and CPU power.

Lastly, it's nice to be warm in the winter but the power draw of an "old" X58 system doesn't reflect well on the electricity bill. :eek:

Decisions decisions...
 
I think people were expecting close to 4 @ stock which would probably mean an easy path up to that magical 5 with Ivy. When that didn't happen, they were expecting it with Haswell, and it hasn't happened again (if the list is true)

My 2500K aint struggling yet, at all. No point in upgrading.
 
Why would that have you concerned? It's progress.. I really hope the next gen consoles do push things along and don't mind upgrading to play the latest games. PC gaming has stagnated. Roll on the next gen, can't come soon enough or powerful enough :p.

You read that completely wrong. My complaint is that PC hardware will fall behind. I think console progression is a good thing. In general I want everything to progress as fast as is physically possible. This inspires new ideas, new technologies and is beneficial for everyone.

I'm annoyed at Intel deliberately "stagnating" hardware when they know they can release hardware twice as fast.

For example SB-E could physically have 8-cores, however they just disabled 2. IB-E could have 10-12 cores, however ther are disabling 4-6 and keeping 6 core with the same specs, and will charge an even higher price.

Obviously Intel is a monopoly, but from a consumer perspective I simply won't pay more for something which is only 1-2% better than the last cycle, and once the new consoles are released I fear that PC's will be left behind unless hardware manufacturers keep up the momentum of progression.

Would you be happy if Nvidia/AMD released graphics cards which were only 1-3% faster than before and had exactly the same specs?

These are the effects of a monopoly I guess. I won't part with my money until Intel start progressing as they did from 2006-2008. Where we saw CPU progression boom in Cores, Cache, 10%~ speed boost in gaming.

No doubt PC gaming will "boom" again. Once everyone starts getting bored of Tablets, and realise that they are completely useless for any practical application.
 
You read that completely wrong. My complaint is that PC hardware will fall behind. I think console progression is a good thing. In general I want everything to progress as fast as is physically possible. This inspires new ideas, new technologies and is beneficial for everyone.

I'm annoyed at Intel deliberately "stagnating" hardware when they know they can release hardware twice as fast.

For example SB-E could physically have 8-cores, however they just disabled 2. IB-E could have 10-12 cores, however ther are disabling 4-6 and keeping 6 core with the same specs, and will charge an even higher price.

Obviously Intel is a monopoly, but from a consumer perspective I simply won't pay more for something which is only 1-2% better than the last cycle, and once the new consoles are released I fear that PC's will be left behind unless hardware manufacturers keep up the momentum of progression.

Would you be happy if Nvidia/AMD released graphics cards which were only 1-3% faster than before and had exactly the same specs?

These are the effects of a monopoly I guess. I won't part with my money until Intel start progressing as they did from 2006-2008. Where we saw CPU progression boom in Cores, Cache, 10%~ speed boost in gaming.

No doubt PC gaming will "boom" again. Once everyone starts getting bored of Tablets, and realise that they are completely useless for any practical application.

PC hardware will never fall behind Console hardware!
 
You read that completely wrong. My complaint is that PC hardware will fall behind. I think console progression is a good thing. In general I want everything to progress as fast as is physically possible. This inspires new ideas, new technologies and is beneficial for everyone.

I'm annoyed at Intel deliberately "stagnating" hardware when they know they can release hardware twice as fast.

For example SB-E could physically have 8-cores, however they just disabled 2. IB-E could have 10-12 cores, however ther are disabling 4-6 and keeping 6 core with the same specs, and will charge an even higher price.

Obviously Intel is a monopoly, but from a consumer perspective I simply won't pay more for something which is only 1-2% better than the last cycle, and once the new consoles are released I fear that PC's will be left behind unless hardware manufacturers keep up the momentum of progression.

Would you be happy if Nvidia/AMD released graphics cards which were only 1-3% faster than before and had exactly the same specs?

These are the effects of a monopoly I guess. I won't part with my money until Intel start progressing as they did from 2006-2008. Where we saw CPU progression boom in Cores, Cache, 10%~ speed boost in gaming.

No doubt PC gaming will "boom" again. Once everyone starts getting bored of Tablets, and realise that they are completely useless for any practical application.

Consoles can only use hardware that's available at the time, by the time said console gets to launch that hardware is typically out of date. So on that bases it's impossible for consoles to be above the curve of PC hardware.

I'm not too sure where you're getting 1-2% increases in performance though, was it a number you plucked from thin air?
 
-Not having S-ATA 3 has held me back from upgrading to the new generations of SSDs.

Do it anyway you will get the benfits, SATA II will stop you hitting 500 odd MB/s transfer speeds with a SATA III drive but unless your copying large files between the SSD's those numbers are just marketing gimmicks anyway, the IOPS of newer drives is what gives them the edge over the older ones and IIRC SATA II can handle the IOPS of even the newest drives ok.


I'm annoyed at Intel deliberately "stagnating" hardware when they know they can release hardware twice as fast.

For example SB-E could physically have 8-cores, however they just disabled 2. IB-E could have 10-12 cores, however ther are disabling 4-6 and keeping 6 core with the same specs, and will charge an even higher price.

I understand what your saying but bear in mind that SB-E is not the only high end SB chips on LGA2011, the Xeon E5's come in 6 and 8 core flavours and judging by them if Intel had released 8 core SB-E's the would have been an incredibly limited market as over clocking would have been highly limited due to thermals. You would have ended up with an 8 core CPU costing about £1000 that would have trouble hitting 4.5Ghz without a cooling solution that sounded like a chainsaw and single core performance that barely outclasses an IB i3, I just can't see Intel making money off that and I doubt they could either.
 
Just my 2p because this always annoys me :p

1) No card on the planet can take full advantage of PCI-E 3.0 x16 or even 8x bandwidth. It will remain as such for long into the future

2) SSD drive performance degrades rapidly as the drive fills up. A 75% full drive will have a lower sequential transfer rate than SATA-II maximum. The important thing is the IOPS, not the sequential transfer.
 
I was going to start buying parts for a new gaming rig I may hold off buying a Ivy and get one of these new cpus. I hear they use a new socket so buying a mobo and cpu now will be a waste of money if I have to buy an a new mobo as well as cpu in the near future.
 
1) No card on the planet can take full advantage of PCI-E 3.0 x16 or even 8x bandwidth. It will remain as such for long into the future

While I agree that PCI-E 3.0 is useless right now, lets not forget that a lot of people (admittedly myself included) once said that the was no point to buy a PCI-E board as no card came close to saturating the AGPx8 bus. It happened less than two graphics card generations after PCI-E's launch.
 
Haswell is probably the point that I will jump ship from my trusty 920. I haven't been overly impressed with individual generations since x58, but the cumulative performance increase is starting to make it more attractive now (~35%?). Not to mention having an up to date chip-set & lower power requirements, cooler running, etc.
 
Haswell is probably the point that I will jump ship from my trusty 920. I haven't been overly impressed with individual generations since x58, but the cumulative performance increase is starting to make it more attractive now (~35%?).

% increases from gen to gen don't stack cumulativly they stack... word I don't know... -ivly.

What I mean is its not a X% + X% + X% upgrade, its a (100% + X%) * X% * X% upgrade.

I doubt I'm making any sense but stock for stock a 3770K is already more than a 35% over a 920.

An example (as I still don't think I'm making any sense) using random %'s is, its not 15%+10%+10%=35% giving 135% of original. Its 100%+15% = 115% then 115%+10% = 126.5%, then 126.5%+10% = 139.15% of the original.
 
Last edited:
% increases from gen to gen don't stack cumulativly they stack... word I don't know... -ivly.

What I mean is its not a X% + X% + X% upgrade, its a (100% + X%) * X% * X% upgrade.

I doubt I'm making any sense but stock for stock a 3770K is already more than a 35% over a 920.

An example (as I still don't think I'm making any sense) using random %'s is, its not 15%+10%+10%=35% giving 135% of original. Its 100%+15% = 115% then 115%+10% = 126.5%, then 126.5%+10% = 139.15% of the original.

You make sense, but I didn't post any calculations for you to correct me on. It was an off the cuff guess without any thought put into it. Is cumulative the wrong term then?
 
Back
Top Bottom